CN111092454B - Unit combination rapid calculation method based on characteristic scheduling points - Google Patents

Unit combination rapid calculation method based on characteristic scheduling points Download PDF

Info

Publication number
CN111092454B
CN111092454B CN201911064846.0A CN201911064846A CN111092454B CN 111092454 B CN111092454 B CN 111092454B CN 201911064846 A CN201911064846 A CN 201911064846A CN 111092454 B CN111092454 B CN 111092454B
Authority
CN
China
Prior art keywords
scheduling
generator
point
characteristic
points
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
Application number
CN201911064846.0A
Other languages
Chinese (zh)
Other versions
CN111092454A (en
Inventor
杨知方
张梦晗
余娟
林伟
代伟
杜尔顺
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing University
Original Assignee
Chongqing University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Chongqing University filed Critical Chongqing University
Priority to CN201911064846.0A priority Critical patent/CN111092454B/en
Publication of CN111092454A publication Critical patent/CN111092454A/en
Application granted granted Critical
Publication of CN111092454B publication Critical patent/CN111092454B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/46—Controlling of the sharing of output between the generators, converters, or transformers
    • 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
    • H02J3/008—Circuit arrangements for ac mains or ac distribution networks involving trading of energy or energy transmission rights
    • 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
    • H02J3/04—Circuit arrangements for ac mains or ac distribution networks for connecting networks of the same frequency but supplied from different sources
    • H02J3/06—Controlling transfer of power between connected networks; Controlling sharing of load between connected networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The invention discloses a unit rapid combination method based on characteristic scheduling points, which mainly comprises the following steps: 1) establishing a combined operation optimization model of the generator set, and selecting a power network characteristic scheduling point; 2) determining the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point; 3) and carrying out iterative correction on the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point to obtain a combined operation result of the generator set of the power network. The invention solves the problem of overlarge calculated amount caused by the increase of the granularity of the scheduling time interval in the existing unit combination problem in the day-ahead scheduling.

Description

Unit combination rapid calculation method based on characteristic scheduling points
Technical Field
The invention relates to the field of power systems, in particular to a unit combination rapid calculation method based on characteristic scheduling points.
Background
Renewable energy sources are increasingly large in power generation scale at present, and a lot of uncertainty is brought to operation optimization of a power system. In order to ensure safer and more economic operation of the power system, the granularity of the scheduling time interval of the power system is increased. The united states alliance energy management committee, et al, has proposed to increase the slot granularity of existing crew combinations and economic schedules, i.e., to improve from every 60 minutes of scheduling to every 15 minutes of scheduling. The MISO in the united states is considering improving the scheduling period granularity, which has been directly applied in scheduling in guangdong province of our country. However, the unit combination problem is essentially a mixed integer programming problem, integer variables increase with increasing granularity, and solution time increases exponentially with adjustment. In fact, in the Guangdong high-granularity scheduling scheme, the convergence gap setting is usually large, and the calculation is accelerated by sacrificing certain economy. Even so, this scheme may fail to obtain the scheduling result for a long time. Therefore, how to rapidly combine computer sets and economically schedule under high granularity is an urgent problem to be solved in industrial application.
The integral variable is required to be reduced for essentially reducing the unit combination calculation amount. Integer variable is the number of generators IgAnd the number of scheduling periods TNAnd (4) forming. The existing fast calculation method based on the aggregation generator or the aggregation scheduling time interval has the defects of insufficient extraction of operation characteristics and difficulty in obtaining the original scheduling scheme under high granularity, so that the method is usually applied to long-term planning with low precision requirement and cannot be directly applied to day-ahead scheduling.
Disclosure of Invention
The present invention is directed to solving the problems of the prior art.
The technical scheme adopted for realizing the purpose of the invention is that the method for quickly combining the units based on the characteristic scheduling points mainly comprises the following steps:
1) and establishing a combined operation optimization model of the generator set in groups, and selecting a power network characteristic scheduling point.
Further, the main steps of selecting the characteristic scheduling points are as follows:
1.1) mixing TNDividing original scheduling points of each unit into TNAnd m groups, wherein each group has m machine group scheduling points.
1.2) establishing a combined operation optimization model of any group of scheduling point generator sets, which mainly comprises the following steps:
1.2.1) establishing an objective function with the aim of minimizing the total operation cost, namely:
Figure GDA0002854219660000021
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000022
representing the primary power generation cost.
Figure GDA0002854219660000023
Representing the cost of idling.
Figure GDA0002854219660000024
Is a continuous variable and is characterized in that,
Figure GDA0002854219660000025
is an integer variable. A. the1,iAnd A2,iRespectively representing a primary power generation cost coefficient and an idle load cost coefficient; i is the serial number of the generator; i is the total number of the generators,
Figure GDA0002854219660000026
to the total number of scheduling points
Figure GDA0002854219660000027
1.2.2) determining the constraint conditions of the objective function, as shown in the formula (2) to the formula (6), respectively:
the upper and lower limits of the unit output force are restricted as follows:
Figure GDA0002854219660000028
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000029
*maximum and minimum values of the parameter, respectively; combined operation optimization model for representing m scheduling point correspondences by superscript mp
The unit climbing upper and lower limit constraints are respectively shown in formula (3) and formula (4):
Figure GDA00028542196600000210
Figure GDA00028542196600000211
in the formula, RUi、RDiAnd the upper limit and the lower limit of the slope climbing of the ith generator are respectively restricted. Sui、SdiRespectively start/stop ramp-up constraints for the ith generator.
The network security constraints are as follows:
Figure GDA00028542196600000212
in the formula, PLine,tFor the flow at the t-th dispatch point, θtThe phase angle of the t-th scheduling point; B. b isfRespectively a node parameter matrix and a branch parameter matrix;
the power balance constraint is as follows:
Figure GDA00028542196600000213
in the formula, PD,tRefers to the node power load matrix at the t-th scheduling point. A. theGTo convert the generator output to a transfer matrix on each node; a. theDA matrix for associating loads to each node;
1.3) establishing a characteristic scheduling point evaluation model, which mainly comprises an extreme value evaluation model, a climbing capability evaluation model and a generator state change evaluation model.
The extremum evaluation model is as follows:
Figure GDA00028542196600000214
wherein a is a preset evaluation parameter. S1，tAnd evaluating the parameter for the extreme value of the t-th scheduling point.
The climbing capability evaluation model is as follows:
Figure GDA0002854219660000031
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000032
and evaluating coefficients for the climbing capacity based on the combined result of the m-point units.
Figure GDA0002854219660000033
The coefficients are evaluated for the climbing ability based on the load curve characteristics.
Climbing capability evaluation coefficient based on m-point unit combination result
Figure GDA0002854219660000034
As follows:
Figure GDA0002854219660000035
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000036
and characterizing the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t +1 moment.
Figure GDA0002854219660000037
And characterizing the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t-1 moment.
the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t +1 moment
Figure GDA0002854219660000038
As follows:
Figure GDA0002854219660000039
in the formula (I), the compound is shown in the specification,
Figure GDA00028542196600000310
is the output result obtained by the m-point unit combination. RU (RU)i、RDiThe climbing capacity limit of the ith generator in unit time is the climbing capacity limit of the ith generator, namely the climbing upper and lower limits of the ith generator are restricted.
the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t-1 moment
Figure GDA00028542196600000311
As follows:
Figure GDA00028542196600000312
climbing capability evaluation coefficient based on load curve characteristics
Figure GDA00028542196600000313
As follows:
Figure GDA00028542196600000314
in the formula (I), the compound is shown in the specification,
Figure GDA00028542196600000315
the ratio of the load change between time t and time t +1 to the maximum load change during the entire scheduling period is characterized.
Figure GDA00028542196600000316
The ratio of the load change between time t and time t-1 to the maximum load change during the entire scheduling period is characterized.
Ratio of load change between time t and time t +1 to maximum load change throughout scheduling
Figure GDA00028542196600000317
As follows:
Figure GDA00028542196600000318
ratio of load change between time t and time t-1 to maximum load change throughout scheduling
Figure GDA00028542196600000319
As follows:
Figure GDA00028542196600000320
in the formula (I), the compound is shown in the specification,
Figure GDA00028542196600000321
Figure GDA00028542196600000322
PD,tis the total load demand of the t-th dispatch point. Parameter (P)D,t+1-PD,t) And parameter-PD,t+1+PD,tAs a function of time t.
The generator state change evaluation model is as follows:
Figure GDA0002854219660000041
1.4) calculating an evaluation coefficient S of the t-th scheduling point by utilizing a characteristic scheduling point evaluation modeltotal,t＝S1,t+S2,t+S3,t。
1.5) according to the evaluation factor Stotal,tThe scheduling points are sorted in descending order, and the first scheduling points are selected as characteristic scheduling points.
1.6) interpolation characteristic scheduling points:
judging whether the time interval of the two characteristic scheduling points is greater than a time threshold value tmaxIf yes, inserting a plurality of characteristic scheduling points in the two characteristic scheduling points to ensure that the time interval of the two characteristic scheduling points is less than or equal to a time threshold tmax. The characteristic scheduling points are written into the set N.
2) And determining the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point.
Further, the main steps of determining the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point are as follows:
2.1) establishing a unit combination operation model based on the characteristic scheduling points, which mainly comprises the following steps:
2.1.1) determining the objective function, namely:
Figure GDA0002854219660000042
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000043
which is the primary power generation cost.
Figure GDA0002854219660000044
Is the idle cost.
Figure GDA0002854219660000045
For startup costs.
Figure GDA0002854219660000046
Is the cost of shutdown. Wherein
Figure GDA0002854219660000047
And
Figure GDA0002854219660000048
is a continuous variable and is characterized in that,
Figure GDA0002854219660000049
is an integer variable.
2.1.2) establishing constraints as shown in equations 17 to 28, respectively:
the upper and lower limits of the unit output force are restricted as follows:
Figure GDA00028542196600000410
the upper and lower limits of the unit climbing are constrained as follows:
Figure GDA00028542196600000411
Figure GDA00028542196600000412
the minimum start-stop time constraint of the unit is as follows:
Figure GDA00028542196600000413
Figure GDA00028542196600000414
Figure GDA0002854219660000051
Figure GDA0002854219660000052
Figure GDA0002854219660000053
Figure GDA0002854219660000054
the unit start-stop cost constraints are as follows:
Figure GDA0002854219660000055
the network security constraints are as follows:
Figure GDA0002854219660000056
the power balance constraint is as follows:
Figure GDA0002854219660000057
in the formula, xRSPThe x is a parameter vector related to a unit combination model based on a characteristic scheduling point, PD,tRefers to the node power load matrix, G, at the t-th scheduling pointi,JiSet of scheduling periods, H, during which the ith generator must be on-line or off-line in order to be in the initial starting phaseU,i,HD,iThe cost of starting or stopping an ith generator once,
Figure GDA0002854219660000058
for minimum startup time or minimum shutdown time of the ith generator, A/B refers to removing elements in set B from set A.
2.1.3) based on the characteristic scheduling point set N, carrying out climbing capacity upper limit on the ith generator in the nth characteristic scheduling point in unit time
Figure GDA0002854219660000059
The lower limit of the climbing capacity of the ith generator in the nth characteristic scheduling point in unit time
Figure GDA00028542196600000510
Starting climbing constraint of ith generator in nth characteristic scheduling point
Figure GDA00028542196600000511
Shutdown and hill climbing restraint of ith generator in nth characteristic scheduling point
Figure GDA00028542196600000512
Scheduling period set of ith generator that must be online
Figure GDA00028542196600000513
Scheduling period set of i-th generator that must be offline
Figure GDA00028542196600000514
Updating is performed as shown in equations (29) to (34), respectively:
Figure GDA00028542196600000515
Figure GDA00028542196600000516
Figure GDA00028542196600000517
Figure GDA00028542196600000518
Figure GDA00028542196600000519
Figure GDA0002854219660000061
wherein, g (N) is a mapping function for mapping the nth element in the characteristic scheduling point set N to the corresponding tth element in the original scheduling point set T.
And 2.2) determining the starting/stopping state of the generator set on the characteristic scheduling point based on the unit combination operation model.
2.3) determining the starting/stopping state of the generator set on the non-characteristic scheduling point, which mainly comprises the following steps:
2.3.1) dividing the time interval between two adjacent characteristic scheduling points according to the characteristic scheduling points in the characteristic scheduling point set N.
2.3.2) for each time interval, a non-characteristic scheduling point aggregation model for a single time interval is established.
Establishing a non-characteristic scheduling point aggregation model objective function by taking the minimum sum of the difference value of the load value on each non-characteristic scheduling point and the load values on the two characteristic scheduling points in a single time interval as a target, namely:
Figure GDA0002854219660000062
and (3) adding a non-characteristic scheduling point aggregation model constraint condition by taking the characteristic scheduling point of each aggregation cluster to have time continuity as a target, namely:
xj≤xj+1 x∈{0,1} (36)
wherein k represents the number of non-specific scheduling points in a single time interval, xjAn integer variable indicating a target characteristic scheduling point to which the jth non-characteristic scheduling point should be aggregated.
2.3.3) separately combining x in a single time interval according to the results of the non-characteristic scheduling point set modeljThe non-characteristic scheduling points with the same value are aggregated to two characteristic scheduling points. The operation of a single period is repeated in each time interval until a scheduling point aggregation cluster containing all scheduling points is formed. On each scheduling point cluster, each generator state is the same. The generator state is divided into start and stop states.
3) And carrying out iterative correction on the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point to obtain a combined operation result of the generator set of the power network.
Further, the main steps of carrying out iterative correction on the combined running state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point are as follows:
and 3.1) taking the scheduling point where the generator with the changed state is positioned and h scheduling points adjacent to the scheduling point as candidate scheduling points. The start-stop variables of the generators with the corresponding state change at the scheduling point are regarded as candidate integer variables.
3.2) aiming at the candidate integer variables, establishing a unit combination model containing the original granularity of the indicator variables, wherein the objective function is as follows:
Figure GDA0002854219660000063
the constraints are as follows:
Figure GDA0002854219660000071
Figure GDA0002854219660000072
Figure GDA0002854219660000073
Figure GDA0002854219660000074
Figure GDA0002854219660000075
Figure GDA0002854219660000076
Figure GDA0002854219660000077
Figure GDA0002854219660000078
Figure GDA0002854219660000079
Figure GDA00028542196600000710
Figure GDA00028542196600000711
Figure GDA00028542196600000712
Figure GDA00028542196600000713
in the formula, a penalty factor R1Penalty factor R2Penalty factor R3Penalty factor R4And penalty factor R5Is a positive number; r5>>R1,R5>>R2,R5>>R3,R5>>R4；xcoAnd x is a parameter vector related to the unit combination model with the original granularity in the correction strategy. PD,tRefers to the node power load matrix at the t-th scheduling point. Gi,JiTo be at an initial start-up stage, the ith generator must be on-line or off-line for a set of scheduling periods. HU,i,HD,iThe cost of starting or stopping the ith generator once.
Figure GDA00028542196600000714
For minimum startup time or minimum shutdown time of the ith generator, A/B refers to removing elements in set B from set A.
3.3) judging the feasibility of the updated generator set combined operation result, wherein the feasibility is mainly divided into the following 2 conditions:
I) xi is a1,t、ξ2,t、ξ3,t、ξ4,tAnd xi5,tAnd if the current generator set combination is zero, judging that the updated generator set combination operation result is feasible, ending iteration, and outputting the current generator set combination.
II) if xi1,t、ξ2,t、ξ3,t、ξ4,tOr xi5,tNot equal to zero, step 3.4 is entered.
3.4) expanding the correction range and returning to the step 3.2.
The method for expanding the correction range comprises the following steps: and 3.3, taking the scheduling point corresponding to the non-zero indication variable in the step 3.3 as a new candidate scheduling point. And if the new candidate dispatching point is determined by the constraint condition (46) and the constraint condition (47), taking the state of the generator violating the climbing constraint on the new candidate dispatching point as a candidate integer variable in the next iteration solution. And if the new candidate dispatching point is determined by the constraint condition (48) and the constraint condition (49), taking all the generator states of the new candidate dispatching point as candidate integer variables in the next iteration solution.
The technical effect of the present invention is undoubted. The invention solves the problem of overlarge calculated amount caused by the increase of the granularity of the scheduling time interval in the existing unit combination problem in the day-ahead scheduling. The correction strategy provided by the invention can ensure a set of feasible unit combination scheduling solutions, and when the result is not feasible, the correction strategy continuously increases the candidate integer variables until a set of feasible solutions is obtained. The worst condition is to solve the unit combination model of the original granularity scale; the correction strategy provided by the invention has high solving efficiency, the main calculation amount of the correction strategy is iteratively corrected for many times, but the candidate integer variables needing to be corrected are usually less, and the solution is quicker.
Drawings
FIG. 1 is a scheduling period aggregation cluster partitioning diagram;
FIG. 2 is a comparison of maximum output capacity for different methods;
FIG. 3 is a comparison LMP in node 83 of the IEEE118 node Standard test System;
FIG. 4 is a comparison LMP in node 573 of a test system for nodes 661 of a province;
Detailed Description
The present invention is further illustrated by the following examples, but it should not be construed that the scope of the above-described subject matter is limited to the following examples. Various substitutions and alterations can be made without departing from the technical idea of the invention and the scope of the invention is covered by the present invention according to the common technical knowledge and the conventional means in the field.
Example 1:
referring to fig. 1, a method for quickly combining units based on feature scheduling points mainly includes the following steps:
1) and establishing a combined operation optimization model of the generator set in groups, and selecting a power network characteristic scheduling point.
Further, the main steps of selecting the characteristic scheduling points are as follows:
1.1) mixing TNDividing original scheduling points of each unit into TNAnd m groups, wherein each group has m machine group scheduling points.
1.2) establishing a combined operation optimization model of any group of scheduling point generator sets, which mainly comprises the following steps:
1.2.1) establishing an objective function with the aim of minimizing the total operation cost, namely:
Figure GDA0002854219660000081
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000091
represents the primary power generation cost;
Figure GDA0002854219660000092
represents the unloaded cost;
Figure GDA0002854219660000093
is a continuous variable representing the output of the generator,
Figure GDA0002854219660000094
the integral variable represents the starting and stopping state of the generator; a. the1,iAnd A2,iRespectively representing a primary power generation cost coefficient and a no-load cost coefficient, wherein i is a generator serial number; i is the total number of the generators,
Figure GDA0002854219660000095
to the total number of scheduling points
Figure GDA0002854219660000096
1.2.2) determining the constraint conditions of the objective function, as shown in the formula (2) to the formula (6), respectively:
the upper and lower limits of the unit output force are restricted as follows:
Figure GDA0002854219660000097
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000098
*maximum and minimum values of the parameter, respectively; and the superscript mp represents a combined operation optimization model corresponding to the m scheduling points.
The unit climbing upper and lower limit constraints are respectively shown in formula (3) and formula (4):
Figure GDA0002854219660000099
Figure GDA00028542196600000910
in the formula, RUi、RDiAnd the upper limit and the lower limit of the slope climbing of the ith generator are respectively restricted. Sui、SdiRespectively start/stop ramp-up constraints for the ith generator.
The network security constraints are as follows:
Figure GDA00028542196600000911
in the formula, PLine,tFor the flow at the t-th dispatch point, θtThe phase angle of the t-th scheduling point; B. b isfRespectively a node parameter matrix and a branch parameter matrix;
the power balance constraint is as follows:
Figure GDA00028542196600000912
in the formula, PD,tA node power load matrix on the t-th scheduling point is pointed; a. theGTo convert the generator output to a transfer matrix on each node; a. theDA matrix for associating loads to each node;
1.3) establishing a characteristic scheduling point evaluation model, which mainly comprises an extreme value evaluation model, a climbing capability evaluation model and a generator state change evaluation model.
The extremum evaluation model is as follows:
Figure GDA00028542196600000913
wherein a is a preset evaluation parameter. S1，tAnd evaluating the parameter for the extreme value of the t-th scheduling point.
The climbing capability evaluation model is as follows:
Figure GDA0002854219660000101
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000102
and evaluating coefficients for the climbing capacity based on the combined result of the m-point units.
Figure GDA0002854219660000103
The coefficients are evaluated for the climbing ability based on the load curve characteristics.
Climbing capability evaluation coefficient based on m-point unit combination result
Figure GDA0002854219660000104
As follows:
Figure GDA0002854219660000105
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000106
and characterizing the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t +1 moment.
Figure GDA0002854219660000107
And characterizing the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t-1 moment.
the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t +1 moment
Figure GDA0002854219660000108
As follows:
Figure GDA0002854219660000109
in the formula (I), the compound is shown in the specification,
Figure GDA00028542196600001010
is the output result obtained by the m-point unit combination. RU (RU)i、RDiThe climbing capacity limit of the ith generator in unit time is the climbing capacity limit of the ith generator, namely the climbing upper and lower limits of the ith generator are restricted.
the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t-1 moment
Figure GDA00028542196600001011
As follows:
Figure GDA00028542196600001012
climbing capability evaluation coefficient based on load curve characteristics
Figure GDA00028542196600001013
As follows:
Figure GDA00028542196600001014
in the formula (I), the compound is shown in the specification,
Figure GDA00028542196600001015
the ratio of the load change between time t and time t +1 to the maximum load change during the entire scheduling period is characterized.
Figure GDA00028542196600001016
The ratio of the load change between time t and time t-1 to the maximum load change during the entire scheduling period is characterized.
Ratio of load change between time t and time t +1 to maximum load change throughout scheduling
Figure GDA00028542196600001017
As follows:
Figure GDA00028542196600001018
ratio of load change between time t and time t-1 to maximum load change throughout scheduling
Figure GDA00028542196600001019
As follows:
Figure GDA00028542196600001020
in the formula, PD,tIs the total load demand of the t-th dispatch point.
Figure GDA00028542196600001021
Finger PD,t+1-PD,tMaximum value of (2).
Figure GDA0002854219660000111
finger-PD,t+1+PD,tMaximum value of (2). t is varied, thus-PD,t+1+PD,tThe calculation result of (2) varies with the variation of t, formula14 selects a certain t among all t such that-PD,t+1+PD,tAnd PD,t+1-PD,tThe maximum value is selected as the constant value
Figure GDA0002854219660000112
Or
Figure GDA0002854219660000113
The generator state change evaluation model is as follows:
Figure GDA0002854219660000114
1.4) calculating an evaluation coefficient S of the t-th scheduling point by utilizing a characteristic scheduling point evaluation modeltotal,t＝S1,t+S2,t+S3,t。
1.5) according to the evaluation factor Stotal,tThe scheduling points are sorted in descending order, and the first scheduling points are selected as characteristic scheduling points.
1.6) interpolation characteristic scheduling points:
judging whether the time interval of the two characteristic scheduling points is greater than a time threshold value tmaxIf yes, inserting a plurality of characteristic scheduling points in the two characteristic scheduling points to ensure that the time interval of the two characteristic scheduling points is less than or equal to a time threshold tmax. The characteristic scheduling points are written into the set N.
Referring to fig. 1, a corresponding time interval is divided between every two characteristic scheduling points, specifically, a time interval (t)mAnd tn) For example, the non-characteristic point in each time interval is divided into a left part and a right part according to the optimization model, and the non-characteristic scheduling point on the left side is aggregated to the characteristic scheduling point tmRight non-characteristic scheduling point is aggregated to characteristic scheduling point tn. Fig. 1(a) shows the non-specific scheduling point division in a single time interval, and fig. 1(b) shows the non-specific scheduling point division result in all time intervals.
2) And determining the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point.
Further, the main steps of determining the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point are as follows:
2.1) establishing a unit combination operation model based on the characteristic scheduling points, which mainly comprises the following steps:
2.1.1) determining the objective function, namely:
Figure GDA0002854219660000115
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000116
which is the primary power generation cost.
Figure GDA0002854219660000117
Is the idle cost.
Figure GDA0002854219660000118
For startup costs.
Figure GDA0002854219660000119
Is the cost of shutdown. Wherein
Figure GDA00028542196600001110
And
Figure GDA00028542196600001111
is a continuous variable and is characterized in that,
Figure GDA00028542196600001112
is an integer variable.
2.1.2) establishing constraints as shown in equations 17 to 28, respectively:
the upper and lower limits of the unit output force are restricted as follows:
Figure GDA00028542196600001113
the upper and lower limits of the unit climbing are constrained as follows:
Figure GDA0002854219660000121
Figure GDA0002854219660000122
the minimum start-stop time constraint of the unit is as follows:
Figure GDA0002854219660000123
Figure GDA0002854219660000124
Figure GDA0002854219660000125
Figure GDA0002854219660000126
Figure GDA0002854219660000127
Figure GDA0002854219660000128
the unit start-stop cost constraints are as follows:
Figure GDA0002854219660000129
the network security constraints are as follows:
Figure GDA00028542196600001210
the power balance constraint is as follows:
Figure GDA00028542196600001211
in the formula, xRSPThe x is a parameter vector related to a unit combination model based on a characteristic scheduling point, PD,tRefers to the node power load matrix, G, at the t-th scheduling pointi,JiSet of scheduling periods, H, during which the ith generator must be on-line or off-line in order to be in the initial starting phaseU,i,HD,iThe cost of starting or stopping an ith generator once,
Figure GDA00028542196600001212
the minimum start-up time or the minimum stop time of the ith generator, A/B refers to removing elements in the set B from the set A;
2.1.3) based on the characteristic scheduling point set N, carrying out climbing capacity upper limit on the ith generator in the nth characteristic scheduling point in unit time
Figure GDA00028542196600001213
The lower limit of the climbing capacity of the ith generator in the nth characteristic scheduling point in unit time
Figure GDA00028542196600001214
Starting climbing constraint of ith generator in nth characteristic scheduling point
Figure GDA00028542196600001215
Shutdown and hill climbing restraint of ith generator in nth characteristic scheduling point
Figure GDA00028542196600001216
Scheduling period set of ith generator that must be online
Figure GDA00028542196600001217
Scheduling period set of i-th generator that must be offline
Figure GDA00028542196600001218
Updating is performed as shown in equations (29) to (34), respectively:
Figure GDA00028542196600001219
Figure GDA00028542196600001220
Figure GDA0002854219660000131
Figure GDA0002854219660000132
Figure GDA0002854219660000133
Figure GDA0002854219660000134
wherein, g (N) is a mapping function for mapping the nth element in the characteristic scheduling point set N to the corresponding tth element in the original scheduling point set T.
And 2.2) determining the starting/stopping state of the generator set on the characteristic scheduling point based on the unit combination operation model.
2.3) determining the starting/stopping state of the generator set on the non-characteristic scheduling point, which mainly comprises the following steps:
2.3.1) dividing the time interval between two adjacent characteristic scheduling points according to the characteristic scheduling points in the characteristic scheduling point set N.
2.3.2) for each time interval, a non-characteristic scheduling point aggregation model for a single time interval is established.
Establishing a non-characteristic scheduling point aggregation model objective function by taking the minimum sum of the difference value of the load value on each non-characteristic scheduling point and the load values on the two characteristic scheduling points in a single time interval as a target, namely:
Figure GDA0002854219660000135
and (3) adding a non-characteristic scheduling point aggregation model constraint condition by taking the characteristic scheduling point of each aggregation cluster to have time continuity as a target, namely:
xj≤xj+1 x∈{0,1} (36)
wherein k represents the number of non-specific scheduling points in a single time interval, xjAn integer variable indicating a target characteristic scheduling point to which the jth non-characteristic scheduling point should be aggregated.
2.3.3) separately combining x in a single time interval according to the results of the non-characteristic scheduling point set modeljThe non-characteristic scheduling points with the same value are aggregated to two characteristic scheduling points. Repeating the operation of a single period in each time interval until a scheduling point cluster containing all scheduling points is formed; on each scheduling point aggregation cluster, the state of each generator is the same; the generator state is divided into start and stop states.
3) And carrying out iterative correction on the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point to obtain a combined operation result of the generator set of the power network.
Further, the main steps of carrying out iterative correction on the combined running state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point are as follows:
and 3.1) taking the scheduling point where the generator with the changed state is positioned and h scheduling points adjacent to the scheduling point as candidate scheduling points. The start-stop variables of the generators with the corresponding state change at the scheduling point are regarded as candidate integer variables.
3.2) aiming at the candidate integer variables, establishing a unit combination model containing the original granularity of the indicator variables, wherein the objective function is as follows:
Figure GDA0002854219660000141
the constraints are as follows:
Figure GDA0002854219660000142
Figure GDA0002854219660000143
Figure GDA0002854219660000145
Figure GDA0002854219660000146
Figure GDA0002854219660000147
Figure GDA0002854219660000148
Figure GDA0002854219660000149
Figure GDA00028542196600001410
Figure GDA00028542196600001411
Figure GDA00028542196600001412
Figure GDA00028542196600001413
Figure GDA00028542196600001414
in the formula, a penalty factor R1Penalty factor R2Penalty factor R3Penalty factor R4And penalty factor R5Is a positive number; r5>>R1,R5>>R2,R5>>R3,R5>>R4；xcoThe index x is a parameter vector related to a unit combination model of the original granularity in the correction strategy; pD,tA node power load matrix on the t-th scheduling point is pointed; gi,JiA set of scheduling periods during which the ith generator must be on-line or off-line for the initial start-up phase; hU,i,HD,iThe cost of starting or stopping the ith generator once;
Figure GDA00028542196600001415
the minimum start-up time or the minimum stop time of the ith generator, A/B refers to removing elements in the set B from the set A;
3.3) judging the feasibility of the updated generator set combined operation result, wherein the feasibility is mainly divided into the following 2 conditions:
I) xi is a1,t、ξ2,t、ξ3,t、ξ4,tAnd xi5,tIf the values are zero, the updated generator set combined operation result is judged to be feasible and finishedAnd (5) iterating the beams, and outputting the current generator set combination.
II) if xi1,t、ξ2,t、ξ3,t、ξ4,tOr xi5,tNot equal to zero, step 3.4 is entered.
3.4) expanding the correction range and returning to the step 3.2.
The method for expanding the correction range comprises the following steps: and 3.3, taking the scheduling point corresponding to the non-zero indication variable in the step 3.3 as a new candidate scheduling point. And if the new candidate dispatching point is determined by the constraint condition (46) and the constraint condition (47), taking the state of the generator violating the climbing constraint on the new candidate dispatching point as a candidate integer variable in the next iteration solution. And if the new candidate dispatching point is determined by the constraint condition (48) and the constraint condition (49), taking all the generator states of the new candidate dispatching point as candidate integer variables in the next iteration solution.
Example 2:
a unit fast combination method based on characteristic scheduling points mainly comprises the following steps:
1) and establishing a combined operation optimization model of the generator set in groups, and selecting a power network characteristic scheduling point.
2) And determining the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point.
3) And carrying out iterative correction on the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point to obtain a combined operation result of the generator set of the power network.
Example 3:
a quick unit combination method based on characteristic scheduling points mainly comprises the following steps of embodiment 2, wherein the characteristic scheduling points are selected mainly as follows:
1) will TNDividing original scheduling points of each unit into TNAnd m groups, wherein each group has m machine group scheduling points.
2) The method comprises the following steps of establishing a combined operation optimization model of any group of scheduling point generator sets, and mainly comprising the following steps:
2.1) establishing an objective function with the aim of minimizing the total operation cost, namely:
Figure GDA0002854219660000151
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000152
represents the primary power generation cost;
Figure GDA0002854219660000153
represents the unloaded cost;
Figure GDA0002854219660000154
is a continuous variable representing the output of the generator,
Figure GDA0002854219660000155
the integral variable represents the starting and stopping state of the generator; a. the1,iAnd A2,iRespectively representing a primary power generation cost coefficient and a no-load cost coefficient, wherein i is a generator serial number; i is the total number of the generators,
Figure GDA0002854219660000156
to the total number of scheduling points
Figure GDA0002854219660000157
2.2) determining the constraint conditions of the objective function, which are respectively shown in the formula (2) to the formula (6):
the upper and lower limits of the unit output force are restricted as follows:
Figure GDA0002854219660000161
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000162
*maximum and minimum values of the parameter, respectively; combined operation optimization model for representing m scheduling point correspondences by superscript mp
The unit climbing upper and lower limit constraints are respectively shown in formula (3) and formula (4):
Figure GDA0002854219660000163
Figure GDA0002854219660000164
in the formula, RUi、RDiAnd the upper limit and the lower limit of the slope climbing of the ith generator are respectively restricted. Sui、SdiRespectively start/stop ramp-up constraints for the ith generator.
The network security constraints are as follows:
Figure GDA0002854219660000165
in the formula, PLine,tFor the flow at the t-th dispatch point, θtThe phase angle of the t-th scheduling point; B. b isfRespectively a node parameter matrix and a branch parameter matrix;
the power balance constraint is as follows:
Figure GDA0002854219660000166
in the formula, PD,tA node power load matrix on the t-th scheduling point is pointed; a. theGTo convert the generator output to a transfer matrix on each node; a. theDA matrix for associating loads to each node;
3) and establishing a characteristic scheduling point evaluation model which mainly comprises an extreme value evaluation model, a climbing capability evaluation model and a generator state change evaluation model.
The extremum evaluation model is as follows:
Figure GDA0002854219660000167
wherein a is preSetting an evaluation parameter. S1，tAnd evaluating the parameter for the extreme value of the t-th scheduling point.
The climbing capability evaluation model is as follows:
Figure GDA0002854219660000168
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000169
and evaluating coefficients for the climbing capacity based on the combined result of the m-point units.
Figure GDA00028542196600001610
The coefficients are evaluated for the climbing ability based on the load curve characteristics.
Climbing capability evaluation coefficient based on m-point unit combination result
Figure GDA00028542196600001611
As follows:
Figure GDA00028542196600001612
in the formula (I), the compound is shown in the specification,
Figure GDA00028542196600001613
and characterizing the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t +1 moment.
Figure GDA00028542196600001614
And characterizing the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t-1 moment.
the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t +1 moment
Figure GDA0002854219660000171
As follows:
Figure GDA0002854219660000172
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000173
is the output result obtained by the m-point unit combination. RU (RU)i、RDiThe climbing capacity limit of the ith generator in unit time is the climbing capacity limit of the ith generator, namely the climbing upper and lower limits of the ith generator are restricted.
the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t-1 moment
Figure GDA0002854219660000174
As follows:
Figure GDA0002854219660000175
climbing capability evaluation coefficient based on load curve characteristics
Figure GDA0002854219660000176
As follows:
Figure GDA0002854219660000177
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000178
the ratio of the load change between time t and time t +1 to the maximum load change during the entire scheduling period is characterized.
Figure GDA0002854219660000179
The ratio of the load change between time t and time t-1 to the maximum load change during the entire scheduling period is characterized.
Ratio of load change between time t and time t +1 to maximum load change throughout scheduling
Figure GDA00028542196600001710
As follows:
Figure GDA00028542196600001711
ratio of load change between time t and time t-1 to maximum load change throughout scheduling
Figure GDA00028542196600001712
As follows:
Figure GDA00028542196600001713
in the formula, PD,tIs the total load demand of the t-th dispatch point.
Figure GDA00028542196600001714
Finger PD,t+1-PD,tMaximum value of (2).
Figure GDA00028542196600001715
finger-PD,t+1+PD,tMaximum value of (2).
The generator state change evaluation model is as follows:
Figure GDA00028542196600001716
4) calculating an evaluation coefficient S of the t-th scheduling point by using a characteristic scheduling point evaluation modeltotal,t＝S1,t+S2,t+S3,t。
5) According to the evaluation coefficient Stotal,tThe scheduling points are sorted in descending order, and the first scheduling points are selected as characteristic scheduling points.
6) Interpolation characteristic scheduling points:
judging whether the time interval of the two characteristic scheduling points is larger than a time threshold valuetmaxIf yes, inserting a plurality of characteristic scheduling points in the two characteristic scheduling points to ensure that the time interval of the two characteristic scheduling points is less than or equal to a time threshold tmax. The characteristic scheduling points are written into the set N.
Example 4:
a quick combination method of a generator set based on characteristic scheduling points mainly comprises the following steps of embodiment 2, wherein the main steps of determining the combined operation state of the generator set on the characteristic scheduling points and the generator set on the non-characteristic scheduling points are as follows:
2.1) establishing a unit combination operation model based on the characteristic scheduling points, which mainly comprises the following steps:
2.1.1) determining the objective function, namely:
Figure GDA0002854219660000181
in the formula (I), the compound is shown in the specification,
Figure GDA0002854219660000182
which is the primary power generation cost.
Figure GDA0002854219660000183
Is the idle cost.
Figure GDA0002854219660000184
For startup costs.
Figure GDA0002854219660000185
Is the cost of shutdown. Wherein
Figure GDA0002854219660000186
And
Figure GDA0002854219660000187
is a continuous variable and is characterized in that,
Figure GDA0002854219660000188
is an integer variable.
2.1.2) establishing constraints as shown in equations 17 to 28, respectively:
the upper and lower limits of the unit output force are restricted as follows:
Figure GDA0002854219660000189
the upper and lower limits of the unit climbing are constrained as follows:
Figure GDA00028542196600001810
Figure GDA00028542196600001811
the minimum start-stop time constraint of the unit is as follows:
Figure GDA00028542196600001812
Figure GDA00028542196600001813
Figure GDA00028542196600001814
Figure GDA00028542196600001815
Figure GDA00028542196600001816
Figure GDA00028542196600001817
the unit start-stop cost constraints are as follows:
Figure GDA00028542196600001818
the network security constraints are as follows:
Figure GDA00028542196600001819
the power balance constraint is as follows:
Figure GDA0002854219660000191
in the formula, xRSPThe x is a parameter vector related to a unit combination model based on a characteristic scheduling point, PD,tRefers to the node power load matrix, G, at the t-th scheduling pointi,JiSet of scheduling periods, H, during which the ith generator must be on-line or off-line in order to be in the initial starting phaseU,i,HD,iThe cost of starting or stopping an ith generator once,
Figure GDA0002854219660000192
the minimum start-up time or the minimum stop time of the ith generator, A/B refers to removing elements in the set B from the set A;
2.1.3) based on the characteristic scheduling point set N, carrying out climbing capacity upper limit on the ith generator in the nth characteristic scheduling point in unit time
Figure GDA0002854219660000193
The lower limit of the climbing capacity of the ith generator in the nth characteristic scheduling point in unit time
Figure GDA0002854219660000194
Starting climbing constraint of ith generator in nth characteristic scheduling point
Figure GDA0002854219660000195
Shutdown and hill climbing restraint of ith generator in nth characteristic scheduling point
Figure GDA0002854219660000196
Scheduling period set of ith generator that must be online
Figure GDA0002854219660000197
Scheduling period set of i-th generator that must be offline
Figure GDA0002854219660000198
The updating is performed as shown in equations (14) to (19), respectively:
Figure GDA0002854219660000199
Figure GDA00028542196600001910
Figure GDA00028542196600001911
Figure GDA00028542196600001912
Figure GDA00028542196600001913
Figure GDA00028542196600001914
wherein, g (N) is a mapping function for mapping the nth element in the characteristic scheduling point set N to the corresponding tth element in the original scheduling point set T.
And 2.2) determining the starting/stopping state of the generator set on the characteristic scheduling point based on the unit combination operation model.
2.3) determining the starting/stopping state of the generator set on the non-characteristic scheduling point, which mainly comprises the following steps:
2.3.1) dividing the time interval between two adjacent characteristic scheduling points according to the characteristic scheduling points in the characteristic scheduling point set N.
2.3.2) for each time interval, a non-characteristic scheduling point aggregation model for a single time interval is established.
Establishing a non-characteristic scheduling point aggregation model objective function by taking the minimum sum of the difference value of the load value on each non-characteristic scheduling point and the load values on the two characteristic scheduling points in a single time interval as a target, namely:
Figure GDA0002854219660000201
and (3) adding a non-characteristic scheduling point aggregation model constraint condition by taking the characteristic scheduling point of each aggregation cluster to have time continuity as a target, namely:
xj≤xj+1 x∈{0,1} (36)
wherein k represents the number of non-specific scheduling points in a single time interval, xjAn integer variable indicating a target characteristic scheduling point to which the jth non-characteristic scheduling point should be aggregated.
2.3.3) separately combining x in a single time interval according to the results of the non-characteristic scheduling point set modeljThe non-characteristic scheduling points with the same value are aggregated to two characteristic scheduling points. Repeating the operation of a single period in each time interval until a scheduling point cluster containing all scheduling points is formed; on each scheduling point aggregation cluster, the state of each generator is the same; the generator state is divided into start and stop states.
Example 5:
a quick combination method of a set based on characteristic scheduling points mainly comprises the following steps of embodiment 2, wherein the main steps of carrying out iterative correction on the combined running state of the set on the characteristic scheduling points and the set on the non-characteristic scheduling points are as follows:
1) and taking a scheduling point where the generator with the changed state is located and h scheduling points adjacent to the scheduling point as candidate scheduling points. The start-stop variables of the generators with the corresponding state change at the scheduling point are regarded as candidate integer variables.
2) Aiming at the candidate integer variables, a unit combination model of the original granularity containing the indicator variables is established, and the objective function is as follows:
Figure GDA0002854219660000202
the constraints are as follows:
Figure GDA0002854219660000203
Figure GDA0002854219660000204
Figure GDA0002854219660000205
Figure GDA0002854219660000206
Figure GDA0002854219660000207
Figure GDA0002854219660000208
Figure GDA0002854219660000209
Figure GDA0002854219660000211
Figure GDA0002854219660000212
Figure GDA0002854219660000213
Figure GDA0002854219660000214
Figure GDA0002854219660000215
Figure GDA0002854219660000216
in the formula, a penalty factor R1Penalty factor R2Penalty factor R3Penalty factor R4And penalty factor R5Is a positive number. R5>>R1,R5>>R2,R5>>R3,R5>>R4。xcoAnd x is a parameter vector related to the unit combination model with the original granularity in the correction strategy. PD,tRefers to the node power load matrix at the t-th scheduling point. Gi,JiTo be at an initial start-up stage, the ith generator must be on-line or off-line for a set of scheduling periods. HU,i,HD,iThe cost of starting or stopping the ith generator once.
Figure GDA0002854219660000217
For minimum startup time or minimum shutdown time of the ith generator, A/B refers to removing elements in set B from set A.
3) The feasibility of the updated generator set combined operation result is judged mainly according to the following 2 conditions:
I) xi is a1,t、ξ2,t、ξ3,t、ξ4,tAnd xi5,tAnd if the current generator set combination is zero, judging that the updated generator set combination operation result is feasible, ending iteration, and outputting the current generator set combination.
II) if xi1,t、ξ2,t、ξ3,t、ξ4,tOr xi5,tIf not, go to step 4.
4) The correction range is expanded and the procedure returns to step 2.
The method for expanding the correction range comprises the following steps: and 3, taking the scheduling point corresponding to the non-zero indication variable in the step 3 as a new candidate scheduling point. And if the new candidate dispatching point is determined by the constraint condition (10) and the constraint condition (11), taking the state of the generator violating the climbing constraint on the new candidate dispatching point as a candidate integer variable in the next iteration solution. And if the new candidate dispatching point is determined by the constraint condition (12) and the constraint condition (13), taking all the generator states of the new candidate dispatching point as candidate integer variables in the next iteration solution.
Example 6:
referring to fig. 1 to 4, an experiment for verifying a unit combination fast algorithm based on feature scheduling points mainly includes the following steps:
1) and building an IEEE30 node, an IEEE118 node standard test system and a 661 node test system in a province.
2) Determining a unit combination solving model:
UC-RSP: solution method using characteristic scheduling point-based unit combination, i.e. text method
UC-24: and based on a 24-point unit combination model. Each scheduling period is 60 minutes in length
UC-48: based on a 48-point unit combination model. Each scheduling period is 30 minutes in length
UC-96: and based on a 96-point unit combination model. Each scheduling period is 15 minutes in length
3) The convergence gap settings in the different systems are as follows:
TABLE 1 Convergence gap settings in each test System
Figure GDA0002854219660000221
A. Calculating speed contrast
TABLE 2 comparison of calculated speeds in various test systems
Figure GDA0002854219660000222
Note: total time t of UC-RSPtotalMainly composed of three parts, 1) point selection time tse: solving time of the m-point unit combination model; 2) solving time trsp: solving time of the unit combination based on the characteristic scheduling points; 3) correcting time tco: total time of iterative correction
The proposed method UC-RSP is always used 247.94s in IEEE118 node standard test system, whereas the benchmark method takes 13151.01 s. The calculation speed of the UC-RSP can be more than 50 times faster than the benchmark test. Meanwhile, the total time of UC-RSP is only slightly increased, but still within an acceptable range, compared to UC-24 and UC-48.
B. Comparison of computational accuracy
In the precision comparison, the total operation cost and the energy cost loss (EPD) are used as the index of the two comparisons, respectively. Wherein, the energy cost loss reflects the influence of the unit scheduling result in the electric power market settlement. The energy cost loss is calculated as follows:
Figure GDA0002854219660000223
wherein the content of the first and second substances,
Figure GDA0002854219660000224
refers to LMPs calculated by different methods,
Figure GDA0002854219660000225
refer to LMP obtained by Benchmark. The LMP is the node marginal electricity price obtained by a subsequent economic dispatching model after the start-stop state is fixed.
The results are shown in the following table:
TABLE 3 operating cost and energy cost penalty in IEEE30 node Standard test line
Figure GDA0002854219660000231
TABLE 4 operating cost and energy cost penalty in IEEE118 node Standard test frame
Figure GDA0002854219660000232
TABLE 5 operating expenses and energy expense losses in 661 node test system of a province
Figure GDA0002854219660000233
a.X (Y), X is a state different from benchmark after the application of the correction strategy, and Y is a state different from benchmark before the application of the correction strategy.
As shown in tables 3-5, the loss of accuracy of the operating costs of the proposed method is below 0.01%. Compared with UC-24 and UC-48, the calculation precision of the method is higher. For example, in 661 testing system of a province, compared with UC-24 and UC-48, the proposed method UC-RSP can reduce the total cost error by 97.91% and 93.91%, respectively. As for the error of energy cost loss, the UC-RSP of the proposed method can be reduced by 75.80% and 75.48% compared with UC-24 and UC-48, respectively.
C. Maximum output capacity contrast
Fig. 2 shows the maximum output capacity of the unit combination scheme obtained by different methods in a 661 node testing system in a province. It can be seen from the results that in the dotted circles, the maximum output capacity obtained by UC-24 and UC-48 is insufficient, whereas the maximum output capacity of Benchmark and the proposed method UC-RSP is sufficient for the load demand.
D. Comparison of calculation results under different quantity of characteristic scheduling points
TABLE 6 comparison of different numbers of characteristic scheduling point results in IEEE118 node standard test system
Figure GDA0002854219660000234
Figure GDA0002854219660000241
TABLE 7 comparison of different numbers of characteristic scheduling point results in 661 node testing system of a province
Figure GDA0002854219660000242
a.tse,trspAnd tcoSame as in table 2.
Tables 6 to 7 respectively list the calculation results of different numbers of characteristic scheduling points selected by the proposed method UC-RSP in the IEEE118 node standard test system and the 661 node test system of a certain province. As shown in the table, even if the number of selected characteristic scheduling points varies widely (for example, in a 661 node testing system of a province, the number of the characteristic scheduling points varies from 24 to 41), the proposed UC-RSP always provides a feasible unit combination result with less precision loss.
Furthermore, as the characteristic scheduling points increase, t required for the corrective policycoThe time is reduced. For example, in a 661-node test system in a province, the correction time t of 24 characteristic scheduling pointscoCorrection time t of 41 characteristic scheduling points for 405.21scoIs 73.93 s. This is because more feature scheduling points can improve the accuracy of the unit combination result solved based on the feature scheduling points before correction, thereby reducing the number of iterations in the correction strategy, i.e., reducing the correction time.
E. Comparison of calculation results under different convergence gaps
Tables 8 and 9 compare the results of UC-96 and UC-RSP under different convergence gaps in the IEEE118 node standard test system and the 661 node standard test system of a certain province. As shown in the table, on the one hand, the calculation time of UC-RSP is significantly smaller than UC-96 at similar convergence gaps. On the other hand, the convergence gap of UC-RSP can be much smaller than UC-96 at similar computation times. Therefore, the proposed method UC-RSP has the potential to achieve higher computational accuracy in the real industry with less time.
TABLE 8 comparison of results for different convergence gaps in IEEE118 node Standard test System
Figure GDA0002854219660000243
TABLE 9 comparison of results of different convergence gaps in 661-node standard test system of a certain province
Figure GDA0002854219660000251
F. Comparison of results of different markets
Fig. 3 and 4 show LMPs obtained by different methods in an IEEE118 node standard test system and a 661 node test system of a certain province. As can be seen from the two representative nodes in fig. 3 and fig. 4, LMPs obtained by UC-RSP is mostly closer to Benchmark, while LMPs obtained by other methods obviously deviate from Benchmark.
Error analysis of LMP obtained by different methods is given in table 10 and table 11. The maximum error and the average error of the UC-RSP are much smaller than those of other methods. In 661 node testing system of a province, the maximum error of UC-RSP is only 3.2411$/MWh, while the maximum errors of UC-24, UC-48 and UC-96 (0.1%) are 4.9366$/MWh, 4.9366$/MWh and 4.9366$/MWh, respectively. In terms of average error, the maximum average error of UC-RSP is only 0.0694$/MWh, while the maximum average errors of UC-24, UC-48 and UC-96 (0.1%) are 0.2735$/MWh, 0.2629$/MWh and 0.1514$/MWh, respectively. Therefore, the electricity price obtained by the method can reflect the electricity price signal more accurately.
TABLE 10 LMP error analysis in EEE118 node standard test system
Figure GDA0002854219660000252
TABLE 11 LMP error analysis in 661 node testing system of a certain province
Figure GDA0002854219660000253

Claims (3)

1. A unit combination rapid calculation method based on characteristic scheduling points is characterized by mainly comprising the following steps:
1) establishing a combined operation optimization model of the generator set in groups, and selecting a power network characteristic scheduling point;
the main steps for selecting the characteristic scheduling points are as follows:
1.1) mixing TNDividing original scheduling points of each unit into TNThe device comprises a/m group, wherein each group is provided with m machine group scheduling points;
1.2) establishing a combined operation optimization model of any group of scheduling point generator sets, which mainly comprises the following steps:
1.2.1) establishing an objective function with the aim of minimizing the total operation cost, namely:
Figure FDA0002981103910000011
in the formula (I), the compound is shown in the specification,
Figure FDA0002981103910000012
represents the primary power generation cost;
Figure FDA0002981103910000013
represents the unloaded cost;
Figure FDA0002981103910000014
is a continuous variable representing the output of the generator,
Figure FDA0002981103910000015
the integral variable represents the starting and stopping state of the generator; a. the1,iAnd A2,iRespectively representing a primary power generation cost coefficient and a no-load cost coefficient, wherein i is a generator serial number; i is the total number of the generators,
Figure FDA0002981103910000016
to the total number of scheduling points
Figure FDA0002981103910000017
1.2.2) determining the constraint conditions of the objective function, as shown in the formula (2) to the formula (6), respectively:
the upper and lower limits of the unit output force are restricted as follows:
Figure FDA0002981103910000018
in the formula (I), the compound is shown in the specification,
Figure FDA0002981103910000019
*maximum and minimum values of the parameter, respectively; combined operation optimization model for representing m scheduling point correspondences by superscript mp
The unit climbing upper and lower limit constraints are respectively shown in formula (3) and formula (4):
Figure FDA00029811039100000110
Figure FDA00029811039100000111
in the formula, RUi、RDiThe first generator is respectively constrained by the climbing upper limit and the climbing lower limit; sui、SdiAre respectively asStart/stop ramp-up constraints of the ith generator;
the network security constraints are as follows:
Figure FDA00029811039100000112
in the formula, PLine,tFor the flow at the t-th dispatch point, θtThe phase angle of the t-th scheduling point; B. b isfRespectively a node parameter matrix and a branch parameter matrix;
the power balance constraint is as follows:
Figure FDA0002981103910000021
in the formula, PD,tA node power load matrix on the t-th scheduling point is pointed; a. theGTo convert the generator output to a transfer matrix on each node; a. theDA matrix for associating loads to each node;
1.3) establishing a characteristic scheduling point evaluation model which mainly comprises an extreme value evaluation model, a climbing capability evaluation model and a generator state change evaluation model;
the extremum evaluation model is as follows:
Figure FDA0002981103910000022
in the formula, a is a preset evaluation parameter; s1，tEvaluating parameters for the extremum of the t-th scheduling point;
the climbing capability evaluation model is as follows:
Figure FDA0002981103910000023
in the formula (I), the compound is shown in the specification,
Figure FDA0002981103910000024
evaluating coefficients for the climbing capability based on the combination result of the m-point unit;
Figure FDA0002981103910000025
evaluating a coefficient for the climbing capability based on the load curve characteristics;
climbing capability evaluation coefficient based on m-point unit combination result
Figure FDA0002981103910000026
As follows:
Figure FDA0002981103910000027
in the formula (I), the compound is shown in the specification,
Figure FDA0002981103910000028
representing the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the time t and the time t + 1;
Figure FDA0002981103910000029
representing the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t-1 moment;
the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t +1 moment
Figure FDA00029811039100000210
As follows:
Figure FDA00029811039100000211
in the formula (I), the compound is shown in the specification,
Figure FDA00029811039100000212
is the output result obtained by the m-point unit combination;RUi、RDithe climbing capacity limit of the ith generator in unit time is defined as the climbing upper and lower limit constraints of the ith generator;
the ratio of the output of the ith generator to the climbing capacity limit of the ith generator between the t moment and the t-1 moment
Figure FDA00029811039100000213
As follows:
Figure FDA00029811039100000214
climbing capability evaluation coefficient based on load curve characteristics
Figure FDA00029811039100000215
As follows:
Figure FDA00029811039100000216
in the formula, Ft t+1Representing the ratio of the load change between the time t and the time t +1 to the maximum load change in the whole scheduling period; ft t-1Representing the ratio of the load change between the t moment and the t-1 moment to the maximum load change in the whole scheduling period;
ratio F of load change between time t and time t +1 to maximum load change during the entire schedulingt t+1As follows:
Figure FDA0002981103910000032
ratio F of load change between time t and time t-1 to maximum load change during the entire schedulingt t-1As follows:
Figure FDA0002981103910000033
in the formula (I), the compound is shown in the specification,
Figure FDA0002981103910000034
PD,ta node power load matrix on the t-th scheduling point; parameter (P)D,t+1-PD,t) And parameter-PD,t+1+PD,tVaries with time t;
the generator state change evaluation model is as follows:
Figure FDA0002981103910000035
1.4) calculating an evaluation coefficient S of the t-th scheduling point by utilizing a characteristic scheduling point evaluation modeltotal,t＝S1,t+S2,t+S3,t；
1.5) according to the evaluation factor Stotal,tThe scheduling points are sorted in descending order, and the first scheduling points are selected as characteristic scheduling points;
1.6) interpolation characteristic scheduling points:
judging whether the time interval of the two characteristic scheduling points is greater than a time threshold value tmaxIf yes, inserting a plurality of characteristic scheduling points in the two characteristic scheduling points to ensure that the time interval of the two characteristic scheduling points is less than or equal to a time threshold tmax(ii) a Writing the characteristic scheduling points into a set N;
2) determining the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point;
3) and carrying out iterative correction on the combined operation state of the generator set at the characteristic scheduling point and the non-characteristic scheduling point to obtain a combined operation result of the generator set of the power network.
2. The method for fast calculating the unit combination based on the characteristic scheduling points as claimed in claim 1, wherein the main steps for determining the combined operation state of the generator units at the characteristic scheduling points and the non-characteristic scheduling points are as follows:
1) the method comprises the following steps of establishing a unit combination operation model based on characteristic scheduling points:
1.1) determining an objective function, namely:
Figure FDA0002981103910000036
in the formula (I), the compound is shown in the specification,
Figure FDA0002981103910000037
the cost of primary power generation;
Figure FDA0002981103910000038
no-load cost;
Figure FDA0002981103910000039
for startup costs;
Figure FDA00029811039100000310
the cost for shutdown; wherein
Figure FDA00029811039100000311
And
Figure FDA00029811039100000312
is a continuous variable and is characterized in that,
Figure FDA0002981103910000041
is an integer variable;
1.2) establishing constraint conditions as shown in the formula 17 to the formula 28 respectively:
the upper and lower limits of the unit output force are restricted as follows:
Figure FDA0002981103910000042
the upper and lower limits of the unit climbing are constrained as follows:
Figure FDA0002981103910000043
Figure FDA0002981103910000044
the minimum start-stop time constraint of the unit is as follows:
Figure FDA0002981103910000045
Figure FDA0002981103910000046
Figure FDA0002981103910000047
Figure FDA0002981103910000048
Figure FDA0002981103910000049
Figure FDA00029811039100000410
the unit start-stop cost constraints are as follows:
Figure FDA00029811039100000411
the network security constraints are as follows:
Figure FDA00029811039100000412
the power balance constraint is as follows:
Figure FDA00029811039100000413
in the formula, xRSPThe x is a parameter vector related to a unit combination model based on the characteristic scheduling point; pD,tA node power load matrix on the t-th scheduling point is pointed; gi,JiA set of scheduling periods during which the ith generator must be on-line or off-line for the initial start-up phase; hU,i,HD,iThe cost of starting or stopping an ith generator once,
Figure FDA00029811039100000414
the minimum starting time or the minimum stopping time of the ith generator; A/B refers to removing elements in set B from set A;
1.3) based on the characteristic scheduling point set N, carrying out climbing capacity upper limit on the ith generator in the nth characteristic scheduling point in unit time
Figure FDA00029811039100000415
The lower limit of the climbing capacity of the ith generator in the nth characteristic scheduling point in unit time
Figure FDA00029811039100000416
Starting climbing constraint of ith generator in nth characteristic scheduling point
Figure FDA00029811039100000417
Shutdown and hill climbing restraint of ith generator in nth characteristic scheduling point
Figure FDA00029811039100000418
Scheduling period set of ith generator that must be online
Figure FDA00029811039100000419
Scheduling period set of i-th generator that must be offline
Figure FDA0002981103910000051
Updating is performed as shown in equations (29) to (34), respectively:
Figure FDA0002981103910000052
Figure FDA0002981103910000053
Figure FDA0002981103910000054
Figure FDA0002981103910000055
Figure FDA0002981103910000056
Figure FDA0002981103910000057
wherein, g (N) is a mapping function for mapping the nth element in the characteristic scheduling point set N to the corresponding tth element in the original scheduling point set T;
2) determining the starting/stopping state of the generator set on the characteristic scheduling point based on the unit combination operation model;
3) determining the starting/stopping state of the generator set on the non-characteristic scheduling point, and mainly comprising the following steps of:
3.1) dividing a time interval between two adjacent characteristic scheduling points according to the characteristic scheduling points in the characteristic scheduling point set N;
3.2) establishing a non-characteristic scheduling point aggregation model for a single time interval for each time interval;
establishing a non-characteristic scheduling point aggregation model objective function by taking the minimum sum of the difference value of the load value on each non-characteristic scheduling point and the load values on the two characteristic scheduling points in a single time interval as a target, namely:
Figure FDA0002981103910000058
and (3) adding a non-characteristic scheduling point aggregation model constraint condition by taking the characteristic scheduling point of each aggregation cluster to have time continuity as a target, namely:
xj≤xj+1 x∈{0,1} (36)
wherein k represents the number of non-specific scheduling points in a single time interval, xjAn integer variable indicating a target characteristic scheduling point to which the jth non-characteristic scheduling point should be aggregated;
3.3) separately combining x in a single time interval according to the result of the non-specific scheduling point set modeljAggregating the non-characteristic scheduling points with the same value to two characteristic scheduling points; repeating the operation of a single period in each time interval until a scheduling point cluster containing all scheduling points is formed; on each scheduling point aggregation cluster, the state of each generator is the same; the generator state is divided into start and stop states.
3. The method for rapidly calculating the unit combination based on the characteristic scheduling points as claimed in claim 1, wherein the main steps of iteratively correcting the combined operation state of the generator units at the characteristic scheduling points and the non-characteristic scheduling points are as follows:
1) taking a scheduling point where a generator with changed state is located and h scheduling points adjacent to the scheduling point as candidate scheduling points; starting and stopping variables of the generator with corresponding state change at the dispatching point are regarded as candidate integer variables;
2) aiming at the candidate integer variables, a unit combination model of the original granularity containing the indicator variables is established, and the objective function is as follows:
Figure FDA0002981103910000061
the constraints are as follows:
Figure FDA0002981103910000062
Figure FDA0002981103910000063
Figure FDA0002981103910000064
Figure FDA0002981103910000065
Figure FDA0002981103910000066
Figure FDA0002981103910000067
Figure FDA0002981103910000068
Figure FDA0002981103910000069
Figure FDA00029811039100000610
Figure FDA00029811039100000614
Figure FDA00029811039100000611
Figure FDA00029811039100000612
Figure FDA00029811039100000613
in the formula, a penalty factor R1Penalty factor R2Penalty factor R3Penalty factor R4And penalty factor R5Is a positive number; r5>>R1,R5>>R2,R5>>R3,R5>>R4；xcoThe index x is a parameter vector related to a unit combination model of the original granularity in the correction strategy; pD,tA node power load matrix on the t-th scheduling point is pointed; gi,JiA set of scheduling periods during which the ith generator must be on-line or off-line for the initial start-up phase; hU,i,HD,iThe cost of starting or stopping the ith generator once;
Figure FDA0002981103910000071
the minimum start-up time or the minimum stop time of the ith generator, A/B refers to removing elements in the set B from the set A;
3) the feasibility of the updated generator set combined operation result is judged mainly according to the following 2 conditions:
I) xi is a1,t、ξ2,t、ξ3,t、ξ4,tAnd xi5,tIf the current generator set combination is zero, judging that the updated generator set combination operation result is feasible, ending iteration, and outputting the current generator set combination;
II) if xi1,t、ξ2,t、ξ3,t、ξ4,tOr xi5,tIf not, entering step 4;
4) enlarging the correction range and returning to the step 2;
the method for expanding the correction range comprises the following steps: taking the scheduling point corresponding to the non-zero indication variable in the step 3 as a new candidate scheduling point; if the new candidate dispatching point is determined by the constraint condition (46) and the constraint condition (47), taking the state of the generator violating the climbing constraint on the new candidate dispatching point as a candidate integer variable in the next iteration solution; and if the new candidate dispatching point is determined by the constraint condition (48) and the constraint condition (49), taking all the generator states of the new candidate dispatching point as candidate integer variables in the next iteration solution.
CN201911064846.0A 2019-11-04 2019-11-04 Unit combination rapid calculation method based on characteristic scheduling points Active CN111092454B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911064846.0A CN111092454B (en) 2019-11-04 2019-11-04 Unit combination rapid calculation method based on characteristic scheduling points

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911064846.0A CN111092454B (en) 2019-11-04 2019-11-04 Unit combination rapid calculation method based on characteristic scheduling points

Publications (2)

Publication Number Publication Date
CN111092454A CN111092454A (en) 2020-05-01
CN111092454B true CN111092454B (en) 2021-06-22

Family

ID=70393061

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911064846.0A Active CN111092454B (en) 2019-11-04 2019-11-04 Unit combination rapid calculation method based on characteristic scheduling points

Country Status (1)

Country Link
CN (1) CN111092454B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103956042A (en) * 2014-04-21 2014-07-30 南京师范大学 Public bike scheduling area intelligent partition method based on graph theory
CN105790265A (en) * 2016-04-21 2016-07-20 三峡大学 AC power flow constraint-based uncertainty unit commitment model and solving method
CN109902874A (en) * 2019-02-28 2019-06-18 武汉大学 A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning
CN110098638A (en) * 2019-06-04 2019-08-06 西安交通大学 A kind of quick unit combined method based on load condition transfer curve

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103956042A (en) * 2014-04-21 2014-07-30 南京师范大学 Public bike scheduling area intelligent partition method based on graph theory
CN105790265A (en) * 2016-04-21 2016-07-20 三峡大学 AC power flow constraint-based uncertainty unit commitment model and solving method
CN109902874A (en) * 2019-02-28 2019-06-18 武汉大学 A kind of micro-capacitance sensor photovoltaic power generation short term prediction method based on deep learning
CN110098638A (en) * 2019-06-04 2019-08-06 西安交通大学 A kind of quick unit combined method based on load condition transfer curve

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
采用功率预测信息的风电场有功优化控制方法;汤奕等;《中国电机工程学报》;20121205;第32卷(第34期);全文 *

Also Published As

Publication number Publication date
CN111092454A (en) 2020-05-01

Similar Documents

Publication Publication Date Title
Ke et al. A novel probabilistic optimal power flow model with uncertain wind power generation described by customized Gaussian mixture model
CN113193600B (en) Electric power system scheduling method and device considering wind power probability distribution uncertainty
CN105279707B (en) A kind of random production analog method considering load and wind-powered electricity generation temporal characteristics
CN109103924B (en) Analysis method of renewable energy consumption capability key factors based on flexible analysis
CN109830979B (en) Calculation based on time sequence simulation and power generation system flexibility evaluation method based on random faults of conventional unit
CN110429637B (en) Visualization method for probability static voltage stability domain
CN110429636B (en) Static voltage stabilization fault screening and sorting method
CN109002650A (en) A kind of Wind turbines power curve modeling method
CN106570790B (en) Wind power plant output data restoration method considering wind speed data segmentation characteristics
CN112054554B (en) Non-parameter statistics-based adaptive distribution robust unit combination method and system
CN111092454B (en) Unit combination rapid calculation method based on characteristic scheduling points
CN114282757A (en) Novel power system operation flexibility assessment method considering wind curtailment load means
CN114066257A (en) Electricity-gas comprehensive energy distribution robust optimization scheduling method and device
CN111797537B (en) Reliability assessment method and device for wind power-containing power distribution system
CN109787217B (en) Standby clearing method based on wind power multi-state model and opportunity cost correction
CN116228292A (en) Electric power spot market price prediction method considering inertia fluctuation
CN110098638A (en) A kind of quick unit combined method based on load condition transfer curve
CN112734158B (en) Thermoelectric load distribution method and device of generator set and electronic equipment
CN112653196B (en) Power system power generation side standby resource scheduling method, system, device and medium
CN113743806A (en) Method for searching non-convex dual-target optimal power flow global solution of power system
CN115296345A (en) Start-stop-output-standby combined optimization method and device for gas generator set
KR20220025511A (en) Method and apparatus for managing energy of smart factory
CN112215383A (en) Distributed photovoltaic power generation power prediction method and system
Kikusato et al. Method for determining line drop compensator control parameters of low-voltage regulator using random forest
CN110707695B (en) Transient power angle stability margin calculation method and system based on artificial intelligence

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