CN106779182B - Reservoir dispatching diagram optimization method based on dynamic planning - Google Patents

Reservoir dispatching diagram optimization method based on dynamic planning Download PDF

Info

Publication number
CN106779182B
CN106779182B CN201611074956.1A CN201611074956A CN106779182B CN 106779182 B CN106779182 B CN 106779182B CN 201611074956 A CN201611074956 A CN 201611074956A CN 106779182 B CN106779182 B CN 106779182B
Authority
CN
China
Prior art keywords
water level
period
dispatching
initial
time
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.)
Expired - Fee Related
Application number
CN201611074956.1A
Other languages
Chinese (zh)
Other versions
CN106779182A (en
Inventor
蒋志强
覃晖
陈璐
周建中
孙怀卫
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.)
Huazhong University of Science and Technology
Original Assignee
Huazhong University of Science and Technology
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 Huazhong University of Science and Technology filed Critical Huazhong University of Science and Technology
Priority to CN201611074956.1A priority Critical patent/CN106779182B/en
Publication of CN106779182A publication Critical patent/CN106779182A/en
Application granted granted Critical
Publication of CN106779182B publication Critical patent/CN106779182B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • E—FIXED CONSTRUCTIONS
    • E02—HYDRAULIC ENGINEERING; FOUNDATIONS; SOIL SHIFTING
    • E02B—HYDRAULIC ENGINEERING
    • E02B9/00—Water-power plants; Layout, construction or equipment, methods of, or apparatus for, making same
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06—Energy or water supply
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/20—Hydro energy

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Marketing (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Quality & Reliability (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • Development Economics (AREA)
  • Primary Health Care (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Mechanical Engineering (AREA)
  • Civil Engineering (AREA)
  • Structural Engineering (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a dynamic planning-based reservoir dispatching diagram optimization methodA method comprising the steps of: dividing the whole scheduling period into T time periods; separating the dispatching lines, and dispersing the dispatching lines in a feasible water level range of each time period to obtain water level dispersion points; establishing all water level discrete point combination spaces of the dispatching line
Figure DDA0001165654630000011
Determining a reservoir dispatching diagram recursive computation model based on dynamic planning: starting reverse time sequence recursive calculation from the T-th time period; the method comprises the following steps of enabling t to be t-1, entering next period calculation, and finishing the reverse time sequence recursive calculation process after the t-1 period calculation is finished; from the first time interval, recursion to the last time interval sequentially can determine the optimal discrete water level combination { CtAnd obtaining an optimal reservoir dispatching diagram. The reservoir dispatching graph drawing model is coupled with the dynamic planning model, and the optimal reservoir dispatching graph is calculated in a reverse time sequence recursion mode, so that the global optimality of the obtained dispatching graph is guaranteed by fully utilizing the global convergence of dynamic planning.

Description

Reservoir dispatching diagram optimization method based on dynamic planning
Technical Field
The invention belongs to the field of optimal operation of hydroelectric energy and optimal scheduling of power generation of a power system, and particularly relates to a reservoir scheduling graph optimization method based on dynamic planning.
Background
The reservoir dispatching diagram is a graphical representation of reservoir dispatching rules and mainly comprises a plurality of dispatching lines and corresponding dispatching areas. At present, two methods are mainly used for drawing a reservoir dispatching diagram, one method is a traditional method, the method is used for calculating and calculating the dispatching diagram by selecting a typical hydrological year runoff series through reverse time recursion, the reservoir dispatching diagram obtained by the method can ensure that the normal operation of a reservoir is not damaged in most time periods, and the method has good physical background significance and high reliability, but has some defects: firstly, the result is greatly influenced by the quality of selection in the typical year; secondly, repeated manual correction is needed in the drawing process, and randomness and experience are high.
Another method is an optimization method that optimizes the dispatch line using an optimization algorithm, such as a genetic algorithm, a particle swarm algorithm, a stepwise optimization algorithm, etc., based on a given initial solution. The reservoir dispatching graph obtained by the optimization method is relatively simple and convenient, but has certain defects: on one hand, the optimization algorithm needs an initial solution, the quality of the initial solution greatly affects the final result, and in addition, the optimization algorithm can only carry out simulation optimization on the initial line generally but cannot directly calculate the optimal scheduling graph; on the other hand, although many evolutionary algorithms have been proved to have global convergence, due to the influence of randomness, they usually cannot guarantee to obtain a global optimal solution under a limited number of iterations, so that the optimality of the scheduling graph obtained by the optimization method is difficult to guarantee.
Patent document CN102080366A discloses a method for making a cascade reservoir joint dispatching graph. The method comprises the steps of taking a long-series historical runoff process as input, taking ten days as a calculation time interval, ensuring output according to the design of each hydropower station, making an initial dispatching graph of each reservoir according to a single-reservoir dispatching graph making method step by step from upstream to downstream according to an equal-output method, carrying out hydropower calculation on each hydropower station according to the hydraulic sequence from top to bottom according to the initial dispatching graph, optimizing each dispatching line of a cascade reservoir combined dispatching graph, and finally outputting the cascade reservoir combined dispatching graph and dispatching rules. The invention takes the output indicated by the cascade joint dispatching diagram as coordination, can play the role of compensation dispatching of the cascade reservoir to a certain extent and improve the utilization of water energy. However, the method for making the cascade reservoir united dispatching graph disclosed by the document has the following defects:
(1) the invention sets the initial dispatching diagram of each reservoir according to the method for manufacturing the single-reservoir dispatching diagram from upstream to downstream by the equal-output method, which is a traditional calculation method of the typical year method, wherein the typical year and the typical year number are selected according to different people, and the randomness and the experience are high, so that the problems that the dispatching diagrams drawn by different designers are different are caused.
(2) According to the method, each power station hydraulic energy is calculated by adopting an enumeration trial calculation method from top to bottom according to an initial scheduling graph in a hydraulic sequence so as to optimize and adjust each scheduling line of the cascade power station scheduling graph, the final optimization result has high dependence on an initial solution, and the final optimization result is calculated in sequence from top to bottom according to the hydraulic sequence, so that the lower bank is influenced only by the upper bank, and therefore the overall optimality of a cascade system cannot be considered globally by the used optimization algorithm.
Disclosure of Invention
Aiming at the defects or improvement requirements in the prior art, the invention provides a dynamic programming-based reservoir dispatching diagram optimization method, which aims to couple a reservoir dispatching diagram drawing model with a dynamic programming model, and adopts a reverse time recursion mode to calculate an optimal reservoir dispatching diagram so as to fully utilize the global convergence of dynamic programming to ensure the global optimality of the obtained dispatching diagram.
In order to achieve the aim, the invention provides a reservoir dispatching diagram optimization method based on dynamic programming, which is characterized by comprising the following steps:
s1: dividing the whole scheduling period into T time periods;
s2: separating the dispatching lines, and dispersing the dispatching lines in a feasible water level range of each time period to obtain water level dispersion points of the dispatching lines at the initial and final time of the time period;
s3: establishing all discrete water level combination spaces of the dispatching line in the dispatching period
Figure GDA0002601021360000021
The discrete water level combination space
Figure GDA0002601021360000031
In a discrete water level combination
Figure GDA0002601021360000032
A time interval schedule chart forming the time interval, wherein k is a water level discrete point number, T is a time interval number, and T is 1,2, …, T;
s4: determining a reservoir dispatching diagram recursive computation model based on dynamic planning:
Figure GDA0002601021360000033
wherein the content of the first and second substances,
Figure GDA0002601021360000034
showing the combination of discrete water levels at the beginning of the t-th period
Figure GDA0002601021360000035
The optimal benefit of the remaining period of time,
Figure GDA0002601021360000036
represents the initial discrete water level combination in the t +1 th period
Figure GDA0002601021360000037
Optimal remaining period benefits; k1 represents a discrete water level combination number at the beginning of the time period, k2 represents a discrete water level combination number at the end of the time period, k1 is 1,2, …, Mn-1,Mn，k2＝1,2,…,Mn-1,MnC represents a discrete water level combination, M is the discrete number of points of the feasible water level range, n is the number of the dispatching lines,
Figure GDA0002601021360000038
j is the flow, j is the age, j is 1,2, …, Y;
s5: starting reverse time sequence recursive computation from the T-th time period, obtaining and storing the optimal remaining period benefit OCB (k, T-1), the optimal remaining period candidate path OCP (k, T-1) and the corresponding optimal time period initial water level Z (k, T-1) corresponding to the current discrete water level combination;
s6: the method comprises the following steps of enabling t to be t-1, entering next period calculation, and finishing the whole reverse time sequence recursive calculation process after the t-1 period calculation is finished;
s7: based on the saved optimal remaining period candidate path OCP (k, T-1), from the first time interval to the last time interval sequentially recursion, the optimal discrete water level combination { C can be determinedtAnd obtaining an optimal reservoir dispatching diagram.
Further, the dispatching lines comprise an upper basic dispatching line, a lower basic dispatching line, an increasing output line and a reducing output line.
Further, the discrete water level combination
Figure GDA0002601021360000039
The period initial water level of (a) is calculated by the following formula:
Figure GDA00026010213600000310
wherein the content of the first and second substances,
Figure GDA0002601021360000041
for discrete water level combination in t-th time period
Figure GDA0002601021360000042
The initial water level of the corresponding time interval,
Figure GDA0002601021360000043
the average power generation amount of Y year in the t period, optk2 is a discrete water level combination number at the end of the optimal period, ZdeadThe end water level of the final period of the scheduling period.
Further, the initial time interval water level trial calculation process in the inverse time sequence recursive calculation is as follows:
s11: assuming an initial water level in a time period;
s12: determining the output N of the reservoir in the time interval according to the initial water level of the time intervalt；
S13: flow according to the time period
Figure GDA0002601021360000044
And a final water level Z (k, t +1), and obtaining the initial water level Z (k, t) of the time interval through iterative calculation of the time interval scheduling graph;
s14: if the calculated initial time-interval water level Z (k, t) just falls in the scheduling area of the initial time-interval water level, assuming that the time-interval water level Z (k, t) is true, and obtaining a proper initial time-interval water level;
s15: if the calculated initial water level Z (k, t) of the time period does not fall within the schedule region of the initial water level, steps S11, S12, S13 and S14 are repeated assuming the initial water level of another time period until a suitable initial water level of the time period is obtained,
s16: and finally obtaining the optimal remaining period benefit OCB (k, t-1), the optimal remaining period candidate path OCP (k, t-1) and the corresponding optimal time period initial water level Z (k, t-1) corresponding to the current discrete water level combination.
In general, compared with the prior art, the above technical solution contemplated by the present invention can achieve the following beneficial effects:
(1) compared with the traditional drawing method, the method does not need to select typical years, the optimal reservoir dispatching diagram is directly optimized and calculated through a dynamic planning algorithm, the method is simpler, more convenient and more effective, and the problems of randomness and experience caused by manual correction in the traditional method can be effectively solved.
(2) The invention provides a dynamic programming-based reservoir dispatching diagram optimization method, which can ensure the global optimality of the obtained reservoir dispatching diagram to a certain extent because of the global convergence of dynamic programming.
Drawings
FIG. 1 is a schematic illustration of a conventional reservoir scheduling of an embodiment of the present invention;
fig. 2 is a feasible water level dispersion diagram of each scheduling line related to the reservoir scheduling diagram optimization method based on dynamic programming according to the embodiment of the invention;
FIG. 3 is a schematic diagram of a time interval dispatching diagram related to a dynamic programming-based reservoir dispatching diagram optimization method in the embodiment of the invention;
fig. 4 is a discrete combination space diagram formed by feasible discrete water levels of each scheduling line related to the reservoir scheduling diagram optimization method based on dynamic programming according to the embodiment of the invention;
FIG. 5 is a process diagram of the reservoir dispatching diagram optimization method based on dynamic programming according to the embodiment of the present invention, wherein the initial water level in the time interval is calculated and calculated by hypothesis;
FIG. 6 is a flow chart of reservoir dispatching diagram calculation based on dynamic programming according to the reservoir dispatching diagram optimization method based on dynamic programming of the embodiment of the present invention;
fig. 7 is an optimal scheduling diagram of the cliff-goat mountain reservoir in consideration of guaranteed output constraints, relating to a dynamic programming-based reservoir scheduling diagram optimization method in the embodiment of the invention;
FIG. 8 is an optimal cliff-goat reservoir dispatching diagram considering a guarantee rate constraint, relating to a reservoir dispatching diagram optimizing method based on dynamic planning in the embodiment of the invention;
fig. 9 is a water level process diagram of a cliff goat landscape reservoir in a multi-year average period, which relates to the reservoir scheduling diagram optimizing method based on dynamic programming according to the embodiment of the invention;
fig. 10 is a chart of a annual average period output process of a cliff goat landscape reservoir related to the dynamic programming-based reservoir scheduling graph optimization method in the embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The invention combines a dynamic planning model with a reservoir dispatching diagram drawing model, and provides a novel reservoir dispatching diagram drawing method which mainly comprises two parts of reservoir dispatching diagram drawing model construction and model recursion solving based on dynamic planning.
(1) Dynamic planning-based reservoir scheduling diagram drawing model
The conventional reservoir dispatching diagram is shown in fig. 1, wherein a plurality of increasing output lines and a plurality of reducing output lines can be drawn according to needs, and only one increasing output line and one reducing output line are provided for simplifying the expression. The method comprises the following concrete steps of constructing a dynamic planning-based reservoir dispatching diagram drawing model:
the method comprises the following steps: the scheduling lines in fig. 1 are separated and placed in independent coordinate systems, and the feasible water level range of each scheduling line in each time interval is discretized, as shown in fig. 2.
Step two: according to fig. 2, a scheduling line water level combination can be obtained corresponding to each water level discrete point of the upper and lower basic scheduling lines, the increasing output line and the decreasing output line at the beginning of a certain time period, another similar water level combination can be obtained corresponding to each water level discrete point of each scheduling line at the end of the time period, and the two water level combinations can form a time period scheduling graph under the condition that the scheduling lines are not crossed, as shown in fig. 3.
Step three: all possible discrete watermark combination spaces for each dispatch line during the dispatch period are established as shown in fig. 4. Where C (k, t) denotes the kth combination of the t-th period, k being 1,2, …, Mn-1,Mn(ii) a n represents the number of scheduling lines in the scheduling graph; m is the discrete number of points of the feasible water level range; t is the slot number, T ═ 1,2, …, T; t is the number of periods of the entire scheduling period.
Step four: on the basis, a reservoir dispatching diagram recursive computation model based on dynamic programming can be obtained according to the dynamic programming principle. The model comprises a reverse time sequence recursion process and a sequential time sequence recursion process, wherein the reverse time sequence recursion calculation process starts from the last time period of the scheduling period, the optimal remaining period benefits of all combinations of all time periods are sequentially calculated and stored from back to front, and the final optimal discrete water level combination process is obtained through sequential time sequence recursion calculation.
For the dispatch graph calculation with only one year of runoff data, the reverse-order recursion equation at the t-th period is as follows:
Figure GDA0002601021360000071
wherein, OCBt(Ck1 t-1) Denotes the initial combination C of the t-th periodk1 t-1Optimal remaining period benefits; OCBt+1(Ck2 t) Denotes the initial combination C in the t +1 th periodk2 tOptimal remaining period benefits; k1 represents a discrete water level combination number at the beginning of the time period, and k2 represents a discrete water level combination number at the end of the time period; k1 ═ 1,2, …, Mn-1,Mn；k2＝1,2,…,Mn-1,Mn(ii) a C represents a discrete water level combination, Ct-1 k1Equivalent to C (k1, t-1) in FIG. 4,
Figure GDA0002601021360000072
equivalent to C (k2, t) in fig. 4.
When the runoff data in Y years is used for calculating the dispatching graph, for each discrete water level combination in the back-push calculation, the output calculation is carried out once by utilizing each runoff data in the current time interval in the long series of Y years, the optimal remaining period benefit is taken as the average value of the generated energy or the Y years of the output, and the initial water level of the average time interval in Y years obtained by calculation is taken as the initial water level of the final time interval in the current time interval. The recursive equation at this time is expressed as:
Figure GDA0002601021360000073
in the reverse time sequence recursion calculation process, the evolution law of the initial water level of each time interval is expressed by a formula (3) as follows:
Figure GDA0002601021360000074
the formula (3) represents the corresponding discrete water level combination Ct-1 k1Time period initial water level Zt-1About the end of the time period water level ZtAnd average power generation for many years
Figure GDA0002601021360000075
A function of, and
Figure GDA0002601021360000076
the optk2 therein is determined by equation (2) corresponding to the optimal remaining period candidate path.
(2) Scheduling graph drawing model solving
When reverse time sequence recursion calculation is carried out, known data has a long series of historical runoff Qj t(j-1, 2, …, Y; T-1, 2, …, T) and a schedule end fixed water level (typically a dead water level). The specific steps for solving the dynamic programming-based reservoir scheduling graph drawing model can be represented as follows:
the method comprises the following steps: dividing the whole scheduling period into T periods within a feasible rangeObtaining discrete water level combination by dispersing each scheduling line, if there are M discrete points at the beginning (end) of each time interval for each scheduling line, then there are M at the beginning (end) of each time interval when the number of scheduling lines in the scheduling graph is nnAnd (4) combining the discrete water levels.
Step two: starting from the T-th time interval, performing reverse-time recursive computation, for any time interval schedule diagram in the current time interval, for example, the time interval schedule diagram formed by the combination C (3, T-1) and the combination C (1, T) in FIG. 4, computing the optimal remaining period benefit OCB (3, T-1), the optimal remaining period candidate path OCP (3, T-1) and the corresponding optimal initial time interval water level Z (3, T-1) corresponding to the combination C (3, T-1) through a trial computation process, and using the known traffic process Q in the trial computation processT j(j ═ 1,2, …, Y) and the fixed end water level (dead water level) of combination C (1, T). And storing the OCB (3, T-1), the OCP (3, T-1) and the Z (3, T-1) corresponding to the combination C (3, T-1), and entering the third step.
Step three: for other discrete water level combinations in this period, i.e. from C (1, T-1) to C (M) after C (3, T-1) is removednAnd T-1), performing the same calculation in the step two, and storing OCB (k, T-1), OCP (k, T-1) and Z (k, T-1) corresponding to each combination.
Step four: and (3) letting T be T-1, entering the next period calculation, wherein the optimal remaining period benefit OCB (k, T) represents the maximum value of the sum of the benefits from the T-th period to the last T-th period, and is calculated by using a formula (2), unlike the OCB (k, T-1) which only represents the benefit of the current period in the step two. Further, the water level at the end of the period of the t-th period of each combination is no longer the dead water level, but corresponds to the period initial water level Z (k, t +1) determined in the (t +1) -th period calculation.
Step five: when the t-th-1 period is calculated, the whole reverse-time recursive calculation process is finished. Finally, based on the saved optimal remaining period candidate paths, a final optimal discrete watermark combination process { C can be determined from the first time period to the last time period in a sequential recursion (backtracking) mannertAnd obtaining an optimal reservoir dispatching diagram.
In the calculation of the output of the scheduling graph in each time period, one is needed when the initial water level of the time period is calculatedThe trial calculation process, namely, firstly assuming an initial water level in a period of time, the initial water level is located in one of 5 scheduling regions (regions A, B, C, D and E) in the graph 3, and assuming that the initial water level is located in the region A, the output in the period of time of the reservoir can be determined as NAThen using the known flow process Qt j(j ═ 1,2, …, Y) and an end-of-period water level Z (k, t +1), and the initial-of-period water level Z (k, t) can be obtained through iterative calculation, and is assumed to be true if the obtained initial-of-period water level Z (k, t) is exactly located in the region a; otherwise, the trial calculation is continued on the assumption of the initial water level value (change of the output situation) in another time period until all possible output situations in the current time period are traversed, as shown in fig. 5. At this time, there may be more than one output situation satisfying the assumption condition, and therefore, these output situations need to be compared, and one output and the initial water level of the corresponding time period are finally retained according to the principle of maximizing the benefit of the remaining period. Of course, there may be no output condition that satisfies the assumed condition, and in this case, the output having the smallest difference between the calculated output and the assumed output is selected as the final output. In order to avoid the discrete water level combination of the latter situation from being selected in the calculation process as much as possible, a certain penalty should be set for the situation. The general flow of calculating the optimal reservoir scheduling map based on the dynamic programming model is shown in fig. 6, and the above-mentioned hypothetical trial calculation process can be represented by the left part in fig. 6.
In the preferred embodiment of the invention, the cliff sheep mountain reservoir station in the Lixianjiang river basin of China is taken as an example, the novel method is used for drawing the reservoir dispatching diagram, long-series simulation dispatching calculation is carried out, and the result is compared and analyzed with the traditional method to show the effect achieved by the invention. The cliff-sheep-mountain-water power station is a Lixianjiang basin faucet power station and has season regulation performance, the normal water storage level is 835m, the dead water level and the flood limiting water level are 818m, the design guarantee rate of the power station is 95%, and the output power is guaranteed to be 23.2 MW. The implementation steps of the invention are as follows:
the method comprises the following steps: and determining the number of the increased and reduced output lines of the reservoir dispatching diagram, the number of the dispatching periods and the state discrete number, and dispersing each dispatching line in the feasible water level range of each period to construct a dispatching line water level discrete combination space.
Step two: inputting data such as long series monthly runoff, boundary control conditions and the like.
Step three: and (3) performing reverse time recursive calculation from the T-th time interval according to the formulas (2) and (3), calculating the corresponding optimal remaining period benefit, optimal remaining period candidate path and optimal time interval initial water level of any one time interval scheduling graph in the current time interval by assuming a trial calculation process, and storing related data.
Step four: let t be t-1 and proceed to the calculation of the next time interval.
Step five: when the calculation of the time period t-1 is completed, according to the stored optimal remaining period candidate path, carrying out sequential recursive calculation from the first time period to the last time period, and determining the final optimal discrete water level combination process { CtAnd obtaining an optimal reservoir dispatching diagram.
The results after implementation of the technical scheme are as follows:
when 23200kW is taken as the minimum guaranteed output constraint of the power station, long-series simulation calculation obtains that the guaranteed rate is 94% and the power generation amount is 4.969 hundred million kWh at the moment, and a corresponding reservoir dispatching diagram is shown in FIG. 7. When 95% is taken as the minimum guarantee rate constraint of the reservoir, the guaranteed output at the moment is 22600kW and the generated energy is 4.972 hundred million kilowatts obtained by long-series simulation calculation, and the corresponding reservoir dispatching diagram is shown in FIG. 8.
And when the output is 23200kW, the long series simulation results of the reservoir dispatching diagram obtained by the traditional method are as follows: the guarantee rate is 93%, and the power generation amount is 4.963 hundred million kWh. When the guarantee rate is 95%, the long series simulation result of the reservoir dispatching diagram obtained by the conventional method is as follows: the output is 22400kW, and the power generation is 4.967 hundred million kWh.
Therefore, when the output is 23200kW, the method provided by the invention is improved in the guarantee rate and the power generation amount compared with the traditional method, and is respectively improved by 1.08% and 0.12%; when the guarantee rates are all 95%, the method provided by the invention is improved in both the guarantee output and the generated energy compared with the traditional method, and the guarantee output and the generated energy are respectively improved by 0.89% and 0.10%. Therefore, the obtained results verify that the dynamic programming-based reservoir dispatching diagram drawing method provided by the invention is superior to the traditional reservoir dispatching diagram drawing method in terms of guarantee rate, guaranteed output and generated energy, and the new method completely avoids the problems of randomness and experience caused by typical year selection and manual revision in the traditional method.
In addition, for comparison with the direct optimization result of the dynamic programming, in the embodiment, for two guarantee rates of 94% and 95%, direct optimization calculation is performed by using the dynamic programming respectively: (1) when the guarantee rate is not lower than 94% and the guarantee output is 23200kW, the annual average generated energy directly optimized by dynamic programming is 4.987 hundred million kilowatt hours; (2) when the guarantee rate is not lower than 95% and the guarantee output is 22600kW, the annual average generated energy directly optimized by dynamic programming is 4.989 billion kilowatt hours. It can be seen that, in the two guarantee rate cases, although the generated energy of the direct optimization calculation is slightly better than the reservoir dispatching diagram result drawn based on the dynamic programming, the difference is not large, and is only 0.018 hundred million kWh and 0.016 hundred million kWh respectively. Therefore, the dynamic planning-based reservoir dispatching diagram drawing method can effectively keep the global convergence of dynamic planning to a certain extent.
When the guarantee rate is 95%, the change process of the mean water level of the cliff mountain reservoir for many years is shown in fig. 9, and the change process of the mean output force for many years is shown in fig. 10. As can be seen from fig. 9, the reservoir is rapidly filled at the initial stage of water storage, and then basically maintains high water level operation to increase the head efficiency. In the later stage of the water supply period, the water level of the reservoir is uniformly reduced, so that the output is uniform, the concentrated damage or water abandonment is avoided, and correspondingly, as can be seen from fig. 10, the output in the water supply period is basically maintained at about the guaranteed output, and the output is increased only in the last period due to the need of emptying the reservoir capacity. Therefore, the simulation result of the scheduling graph is consistent with the actual scheduling condition, and the obtained annual average operation process conforms to the actual scheduling principle, so that the rationality of the method is further verified.
The invention takes a cliff goat mountain reservoir in the Yangxiang river basin of Li Xian of China as an example, the reservoir dispatching diagram is obtained by the proposed method and long-series simulation calculation is carried out, and the calculation result shows that the reservoir dispatching diagram obtained by the invention is superior to the traditional method in the aspects of total generated energy, guaranteed output and guaranteed rate, and particularly the increment is about 1% in the aspects of guaranteed rate and guaranteed output. Therefore, the simulation result well verifies the rationality and effectiveness of the method.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims (3)

1. A reservoir dispatching diagram optimization method based on dynamic programming is characterized by comprising the following steps:
s1: dividing the whole scheduling period into T time periods;
s2: separating the dispatching lines, and dispersing the dispatching lines in a feasible water level range of each time period to obtain water level dispersion points of the dispatching lines at the initial and final time of the time period;
s3: establishing all discrete water level combination spaces of the dispatching line in the dispatching period
Figure FDA0002601021350000011
The discrete water level combination space
Figure FDA0002601021350000012
In a discrete water level combination
Figure FDA0002601021350000013
A time interval schedule chart forming the time interval, wherein k is a water level discrete point number, T is a time interval number, and T is 1,2, …, T;
s4: determining a reservoir dispatching diagram recursive computation model based on dynamic planning:
Figure FDA0002601021350000014
wherein the content of the first and second substances,
Figure FDA0002601021350000015
is shown intime interval and time interval initial discrete water level combination
Figure FDA0002601021350000016
The optimal benefit of the remaining period of time,
Figure FDA0002601021350000017
represents the initial discrete water level combination in the t +1 th period
Figure FDA0002601021350000018
Optimal remaining period benefits; k1 represents a discrete water level combination number at the beginning of the time period, k2 represents a discrete water level combination number at the end of the time period, k1 is 1,2, …, Mn-1,Mn，k2＝1,2,…,Mn-1,MnM is the discrete number of points of the feasible water level range, n is the number of the dispatching lines,
Figure FDA0002601021350000019
j is the flow, j is the age, j is 1,2, …, Y;
s5: starting reverse time sequence recursive computation from the T-th time period, obtaining and storing the optimal remaining period benefit OCB (k, T-1), the optimal remaining period candidate path OCP (k, T-1) and the corresponding optimal time period initial water level Z (k, T-1) corresponding to the current discrete water level combination; the time interval initial water level trial calculation process in the reverse time sequence recursion calculation comprises the following steps:
s11: assuming an initial water level in a time period;
s12: determining the output N of the reservoir in the time interval according to the initial water level of the time intervalt；
S13: flow according to the time period
Figure FDA00026010213500000110
And a final water level Z (k, t +1), and obtaining the initial water level Z (k, t) of the time interval through iterative calculation of the time interval scheduling graph;
s14: if the calculated time period initial water level Z (k, t) just falls in the scheduling area of the time period initial water level, assuming that the time period initial water level Z (k, t) is true, and obtaining a proper time period initial water level;
s15: if the calculated initial time period water level Z (k, t) does not fall in the scheduling region of the initial time period water level, assuming the initial time period water level of another time period, repeating the steps S11, S12, S13 and S14 until a proper initial time period water level is obtained;
s16: finally, obtaining the optimal remaining period benefit OCB (k, t-1), the optimal remaining period candidate path OCP (k, t-1) and the corresponding optimal time period initial water level Z (k, t-1) corresponding to the current discrete water level combination;
s6: the method comprises the following steps of enabling t to be t-1, entering next period calculation, and finishing the whole reverse time sequence recursive calculation process after the t-1 period calculation is finished;
s7: based on the saved optimal remaining period candidate path OCP (k, T-1), from the first time interval to the last time interval sequentially recursion, the optimal discrete water level combination { C can be determinedtAnd obtaining an optimal reservoir dispatching diagram.
2. The dynamic programming-based reservoir dispatching diagram optimization method of claim 1, wherein the dispatching lines comprise an upper basic dispatching line, a lower basic dispatching line, an increasing output line and a decreasing output line.
3. The dynamic programming-based reservoir dispatching diagram optimization method of claim 1 or 2, wherein the discrete water level combination
Figure FDA0002601021350000021
The period initial water level of (a) is calculated by the following formula:
Figure FDA0002601021350000022
wherein the content of the first and second substances,
Figure FDA0002601021350000023
is the t time interval and the initial discrete water level combination
Figure FDA0002601021350000024
The initial water level of the corresponding time interval,
Figure FDA0002601021350000025
the average power generation amount of Y year in the t period, optk2 is a discrete water level combination number at the end of the optimal period, ZdeadThe end water level of the final period of the scheduling period.
CN201611074956.1A 2016-11-29 2016-11-29 Reservoir dispatching diagram optimization method based on dynamic planning Expired - Fee Related CN106779182B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201611074956.1A CN106779182B (en) 2016-11-29 2016-11-29 Reservoir dispatching diagram optimization method based on dynamic planning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201611074956.1A CN106779182B (en) 2016-11-29 2016-11-29 Reservoir dispatching diagram optimization method based on dynamic planning

Publications (2)

Publication Number Publication Date
CN106779182A CN106779182A (en) 2017-05-31
CN106779182B true CN106779182B (en) 2020-10-16

Family

ID=58900853

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201611074956.1A Expired - Fee Related CN106779182B (en) 2016-11-29 2016-11-29 Reservoir dispatching diagram optimization method based on dynamic planning

Country Status (1)

Country Link
CN (1) CN106779182B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115169825B (en) * 2022-06-15 2023-03-24 中国水利水电科学研究院 Method for determining drought alarm water levels of reservoirs with different adjusting capacities

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102080366A (en) * 2011-01-01 2011-06-01 国网电力科学研究院 Method for drawing joint scheduling graph of step reservoir
US20140081613A1 (en) * 2011-11-01 2014-03-20 Austin Geomodeling, Inc. Method, system and computer readable medium for scenario mangement of dynamic, three-dimensional geological interpretation and modeling
CN102867275A (en) * 2012-08-14 2013-01-09 贵州乌江水电开发有限责任公司 Medium-term and long-term combined power generation optimal scheduling method and system in cascade reservoir group

Also Published As

Publication number Publication date
CN106779182A (en) 2017-05-31

Similar Documents

Publication Publication Date Title
JP6736112B2 (en) A short-term practical scheduling method for ultra-large-scale hydropower stations
CN112633578B (en) Cascade reservoir group optimal scheduling method under influence of diversion and water diversion engineering
CN106682810B (en) Long-term operation method of cross-basin cascade hydropower station group under dynamic production of giant hydropower station
CN102518092B (en) Design method for optimized allocation of ice prevention storage capacity of step hydroelectric station reservoir
CN109636674B (en) Large-scale hydropower station group monthly transaction electric quantity decomposition and checking method
CN105631528B (en) Multi-target dynamic optimal power flow solving method based on NSGA-II and approximate dynamic programming
CN102855591A (en) Method and system for optimizing scheduling for short-term combined generation of cascade reservoir group
CN107609679A (en) The preferred method for drafting of multi-parameter and system of a kind of annual-storage reservoir power generation dispatching figure
CN110322056A (en) It is a kind of meter and central air conditioner system the random ADAPTIVE ROBUST Optimization Scheduling of virtual plant
Li et al. Hierarchical multi-reservoir optimization modeling for real-world complexity with application to the Three Gorges system
CN115471006B (en) Power supply planning method and system considering uncertainty of wind power output
CN111612292A (en) Cascade hydropower station dispatching control system and method based on key water level control
CN109657850B (en) Medium-and-long-term step hydropower optimization scheduling method and device
CN104504455B (en) A kind of lower GROUP OF HYDROPOWER STATIONS Long-term Optimal Dispatch method of step accumulation of energy control
CN115310797A (en) Robust planning method for park comprehensive energy system
Jiang et al. Research and application of key technologies in drawing energy storage operation chart by discriminant coefficient method
CN112581310A (en) Cascade hydropower station group power generation index distribution method
CN113435659B (en) Scene analysis-based two-stage optimized operation method and system for comprehensive energy system
CN109584099B (en) Hydropower station short-term multi-target power generation plan compilation method and system for power grid peak regulation
Ha et al. Electricity generation cost reduction for hydrothermal systems with the presence of pumped storage hydroelectric plants
CN113363976B (en) Scene graph-based wind-solar-water complementary power generation system medium-term optimization scheduling method
CN103679284A (en) Accommodated wind power accessed fixed interval rolling scheduling method
CN106779182B (en) Reservoir dispatching diagram optimization method based on dynamic planning
CN106815656B (en) Method for acquiring cascade reservoir energy storage dispatching diagram
CN109635999B (en) Hydropower station scheduling method and system based on particle swarm-bacterial foraging

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
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20201016

Termination date: 20211129

CF01 Termination of patent right due to non-payment of annual fee