CN1971601A - Feasible searching method of optimizing scheduling of reservoir - Google Patents
Feasible searching method of optimizing scheduling of reservoir Download PDFInfo
- Publication number
- CN1971601A CN1971601A CNA2006101253455A CN200610125345A CN1971601A CN 1971601 A CN1971601 A CN 1971601A CN A2006101253455 A CNA2006101253455 A CN A2006101253455A CN 200610125345 A CN200610125345 A CN 200610125345A CN 1971601 A CN1971601 A CN 1971601A
- Authority
- CN
- China
- Prior art keywords
- reservoir
- feasible
- period
- water level
- level
- 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.)
- Pending
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a feasible search method for reservoir optimum dispatching, all collected data is compressed by the computer, forms the feasible search method for single reservoir optimum dispatching with complex constraint conditions or reservoir clusters optimum dispatching and the reservoir optimum dispatching with water resources optimal allocation; the steps of invention contain forward recur the feasible extent of water level operating trace, backward recur the feasible extent of water level operating trace, determine the largest feasible extent of water level operating trace, choose the feasible extent of water level operating trace and choose the optimum answer of reservoir optimum dispatching. The invention can define the feasible trace extent in the dispatching period of reservoir; guarantees all answers are feasible solutions; convenient for man-machine interaction, help the decision maker make decision; the thinking route of invention is intelligible, simplicity and utility, and high executing efficiency, and it fits for single reservoir optimum dispatching with complex constraint conditions, and also fits for reservoir clusters optimum dispatching and the water resources optimal allocation.
Description
Technical field
The present invention relates to the optimizing scheduling of reservoir technical field, particularly relate to a kind of Feasible searching method of optimizing scheduling of reservoir.
Background technology
Optimizing scheduling of reservoir is under the situation of the requirements of comprehensive utilization of given given period initial water level and termination water level and each calculation interval, reservoir inflow according to weather report, the problem of carrying out the optimal selection of reservoir operation strategy.Because of it can obtain the common concern that huge economic benefit and social benefit are subjected to industry under the condition that does not increase any engineering cost.
From the forties in 20th century, American Mases will optimize notion the earliest and be incorporated into after the reservoir operation, the method that optimizing scheduling of reservoir adopted is also more and more, generally can be divided into linear programming, nonlinear programming, dynamic programming and improve one's methods, genetic algorithm, artificial neural network, ant group algorithm, chaos optimization algorithm and particle swarm optimization etc. (see document
[1] [5]).
Linear programming (LP) is a simpler and widely used planing method in the optimizing scheduling of reservoir.But this method counting yield is low, because the non-linear and randomness of optimizing scheduling of reservoir, when the objective function of scheduling and constraint condition are very complicated, needs with other method problem linearization is found the solution again earlier.
Optimizing process slow, relative computing time of long problem appears in nonlinear programming (NLP) method usually.Than linear programming complexity, and do not have general method for solving and program, thereby the popularity of the application of nonlinear programming is less than linear programming and dynamic programming.
Dynamic programming (DP) is to solve the most frequently used a kind of mathematical method of multistage decision process optimization problem, can solve this problem of non-linear and randomness of water resource system well, but this method is still perfect inadequately, its maximum problem is to work as the reservoir number for a long time, tend to produce inevitably " dimension calamity ", amount of calculation is also very big in addition.Therefore, relevant scholars have proposed a large amount of innovative approachs, mainly contain the coarse grid interpositioning, approach dynamic programming (DPSA), increment dynamic programming (IDP) and discrete differential dynamic programming (DDDP), optimized Algorithm (POA) etc. progressively one by one continuously.
Genetic algorithm (GA) provides a kind of general framework for separating the complication system optimization problem, and it does not rely on the specific field of problem, and the kind of finding the solution problem is had very strong robustness, and its weak point is that there is " precocity " phenomenon in its computation process.
Artificial neural network (ANN) is a kind of information handling system with height nonlinear adaptive parallel distributed, for many reservoirs, multivariate, when multiple goal is carried out the system decision-making, have fast operation, structure flexibly, the advantage that manual intervention is few.
Ant group algorithm (ACO) is a kind of randomization heuristic search algorithm, advantage with parallelization, strong robustness, positive feedback, in optimizing scheduling of reservoir, can in search, find acceptable preferably separating, (see document but stagnation behavior appears in this method easily with ant group algorithm
[3]).
Chaos optimization algorithm (CA) is when being applied to optimizing scheduling of reservoir, adopt Chaos Variable allowing solution space to search for, search procedure is undertaken by the chaotic motion self-law, the easier locally optimal solution of jumping out, and the search efficiency height is a kind of very effective optimization method.
(PSO) is simple to operate for particle swarm optimization, the few and fast convergence rate of the empirical parameter of dependence, but also exist search precision shortcoming not high and that be easy to be absorbed in locally optimal solution (to see document
[5]).
In addition, people such as Cui Ruihong, Dong Zengchuan is " reservoir optimizing and dispatching method is researched and analysed "
[1]In the document, analyzed the progress of research of domestic and international optimizing scheduling of reservoir.To the list of application analysis of several representational methods in optimizing scheduling of reservoir relatively, at last prospect has been made in the research and development of reservoir optimizing and dispatching method from now on.
Said method has the following disadvantages in solving optimizing scheduling of reservoir:
The discretize problem.Reservoir level is a continuous variable, above-mentioned a lot of algorithms such as DP and improvement algorithm, genetic algorithm and ant group algorithm etc. all are based on discrete water level or the storage capacity point, so the gained optimum solution is not an optimum solution truly, real optimum solution should be sought in continuous space, and reservoir level or storage capacity are carried out the discretize processing, reduce the precision of understanding.
Initial feasible solution is selected problem, has the reservoir operation problem of complicated constraint condition, and these methods all fail to propose the feasible solution that initial feasible solution is selected, and a kind of method is to use penalty function method, with penalty factor restricted problem is converted into unconstrained problem; Another kind method is that each feasibility of separating of structure is checked, and removes non-feasible solution, and amount of calculation greatly increases, and efficient is very low.As the document " application of self-adapted genetic algorithm in optimizing scheduling of reservoir " that the water conservancy journal is delivered in April, 2006
[2]In, the selection of having introduced initial solution also is to adopt the method for removing non-feasible solution;
Local extremum problem, a lot of algorithms have susceptibility to initial feasible solution, and dynamic programming, genetic algorithm, ant group algorithm and chaos optimization algorithm etc. easily are absorbed in local extremum problem and (see document
[2] [4]).
In addition, in relevant document: introduced ant group algorithm, PSO algorithm and two kinds of several methods that algorithm can merge, they have carried out more detailed argumentation to this problem of ant group algorithm.For storehouse group's problem, the optimization problem that just solves single storehouse or two storehouses that many optimization methods provide then seldom relates to for many storehouses problem.
In sum, the method that provides in the prior art about solution optimizing scheduling of reservoir problem, exist many defectives, particularly select can not effectively handle aspect problem, the local extremum problem at discretize problem, initial feasible solution, especially outstanding to the problem of the optimizing scheduling of reservoir under the complicated constraint condition situation of tool.
Summary of the invention
Technical matters to be solved by this invention is: a kind of Feasible searching method of optimizing scheduling of reservoir is provided, to overcome the defective of prior art, especially makes positive contribution in the optimizing scheduling of reservoir problem of effectively handling the complicated constraint condition of tool.
The present invention solves its technical matters and adopts following technical scheme:
The Feasible searching method of optimizing scheduling of reservoir provided by the invention is: the data of being gathered are by Computer Processing, constitute the single optimizing scheduling of reservoir be applicable to the complicated constraint condition of tool, or the Feasible searching method of multi-reservoir Optimization Dispatching and the water resource optimizing scheduling of reservoir of distributing rationally.The step that this method adopts comprises: the determining of forward recursive reservoir level running orbit feasible region, backward induction method reservoir level running orbit feasible region, reservoir level running orbit maximum feasible scope, reservoir level moves feasible track chooses selection with the optimum solution of optimizing scheduling of reservoir.
The Feasible searching method of optimizing scheduling of reservoir provided by the invention, its purposes has: can carry out single optimizing scheduling of reservoir and use, the multi-reservoir Optimization Dispatching is used, and can carry out water resource and distribute application rationally.
The present invention compared with prior art has following marked improvement and outstanding effect:
One. by determining the bound of reservoir operation phase day part water level, can be clearly in schedule periods reservoir level running orbit maximum feasible scope;
They are two years old. in reservoir level running orbit feasible region, determine the water level of reservoir can guarantee to obtain feasible track by the period by the order that begins from schedule periods;
They are three years old. and Deterministic Methods and random device are combined, can guarantee the continuity of the of overall importance and water level searched for;
They are four years old. and the algorithm in conjunction with DDDP and GA etc. can obtain better approximate optimal solution;
They are five years old. be convenient to man-machine interaction, aid decision making person's decision-making.
They are six years old. this method clear thinking, simple and practical, execution efficient height, be applicable to the single optimizing scheduling of reservoir of the complicated constraint condition of tool, be applicable to that also multi-reservoir Optimization Dispatching and water resource distribute rationally, and can effectively improve the work efficiency of optimizing scheduling of reservoir.
Description of drawings
Fig. 1: forward recursive reservoir level running orbit maximum feasible scope synoptic diagram;
Fig. 2: backward induction method reservoir level running orbit maximum feasible scope synoptic diagram;
Fig. 3: determine reservoir level running orbit maximum feasible scope synoptic diagram;
Fig. 4: reservoir level running orbit method of adjustment synoptic diagram during forward recursive;
Fig. 5: reservoir level running orbit method of adjustment synoptic diagram during backward induction method;
Fig. 6: the flow chart of optimizing scheduling of reservoir Feasible searching method.
Embodiment
The present invention is a kind of Feasible searching method of optimizing scheduling of reservoir, it has the purposes in carrying out single optimizing scheduling of reservoir application, purposes during the multi-reservoir Optimization Dispatching is used is distributed purposes in the application rationally carrying out water resource, and the purposes of others in using.
Be described further below in conjunction with embodiment and accompanying drawing Feasible searching method optimizing scheduling of reservoir of the present invention:
1. reservoir level running orbit maximum feasible scope determines
Reservoir level has different requirements in the difference operation period, usually require reservoir level to be controlled between normal pool level and the level of dead water in non-flood season, be controlled at flood season between flood control and the level of dead water, sometimes because some special circumstances also have the policymaker to control the requirement of some period reservoir level as the case may be.In reservoir operation in the phase, the day part reservoir also has requirement such as comprehensive utilization grade, and as the minimum and maximum letdown flow of day part, minimum and maximum generated output etc., what add upper storage reservoir comes water and reservoir self characteristics, therefore, optimizing scheduling of reservoir has the optimization problem of complicated constraint condition often.All there is a limit range in reservoir level in the variation of each period of schedule periods, promptly has reservoir level running orbit maximum feasible scope.
(1) forward recursive reservoir level running orbit feasible region:
According to the basic data of given reservoir and the comprehensive utilization and the constraint condition of day part, inquire into by computing machine:
As shown in Figure 1: from initial water level, according to this period promptly the maximum letdown flow, the maximum generation that allow of the 1st period exert oneself and period Mo lowest water level require calculating, in conjunction with reservoir following ability of letting out and the unit output restriction of this moment, obtain the water level lower limit that the reservoir at the beginning of i.e. the 2nd period of the 1st period Mo may reach, calculate water level lower limit at the beginning of the 3rd period according to identical method based on this again, till calculating the end of term, draw the feasible region of the day part reservoir level running orbit lower limit that the forward recursive at the beginning of the given period obtains.
As shown in Figure 1: from initial water level, calculate according to the minimum letdown flow of i.e. the 1st period permission of this period, minimum generated output and period Mo peak level requirement, in conjunction with reservoir following ability of letting out and the unit output restriction of this moment, obtain the water level higher limit that the reservoir at the beginning of i.e. the 2nd period of the 1st period Mo may reach, calculate water level higher limit at the beginning of the 3rd period according to identical method based on this again, till calculating the end of term, draw the feasible region of the day part reservoir level running orbit upper limit that the forward recursive at the beginning of the given period obtains.
For guaranteeing that reservoir level all can reach the water level of setting in each period, can adopt following method to carry out the water level lower limit of subrange and the adjustment of the upper limit:
If begin to calculate with the water level upper limit at the beginning of certain period, even if do not reach the water level upper limit requirement of this period end according to letting out ability (spacious let out) under the maximum yet, shown in Fig. 4 n-1 period, period end requires water level to be up to the G point, and the D point maximum possible at the beginning of the period arrives the F point.At this moment, the water level upper limit that the reservoir level running orbit upper limit of period end is adjusted into this period end requires the G point, and the water level upper limit at the beginning of the period is adjusted as follows: the water level upper limit based on period end requires the G point, retrodict to draw on the water level at the beginning of this period and be limited to the H point, at the beginning of this period that water level running orbit upper limit A point went out for basic calculation at the beginning of a period before if this upper limit was positioned within water level coverage pattern D point and the C point, be the reservoir level running orbit upper limit at the beginning of this period then with this H point, otherwise, serve as that the basis is to the last period retrodicts again with the water level upper limit H point at the beginning of this period of falling back out again, till at the beginning of the water level upper limit that obtains is positioned at the more last period, being limited within reached at the water level range of basic calculation gained on the water level, adjust the water level running orbit upper limit at the beginning of all periods of retrodicting.
If begin to calculate with the water level lower limit at the beginning of certain period, even if also do not reach the water level lower limit requirement of this period end according to minimum letdown flow, shown in Fig. 4 n+1 period, period end requires to be limited under the water level N point, the K point maximum possible arrival L point at the beginning of the period.At this moment, the water level lower limit that the reservoir level running orbit lower limit of period end is adjusted into this period end requires the N point, and the water level lower limit at the beginning of the period is adjusted as follows: the water level lower limit based on this period end requires the N point, retrodict to draw under the water level at the beginning of this period and be limited to the O point, at the beginning of this period that a period water level running orbit lower limit Q point went out for basic calculation before if this lower limit was positioned within water level coverage pattern R point and the K point, then with the O point as reservoir level running orbit lower limit at the beginning of the period; Otherwise, serve as that the basis is to the last period retrodicts again with the water level lower limit O point at the beginning of this period of falling back out again, till at the beginning of the water level lower limit that obtains is positioned at the more last period, being limited within reached at the water level range of basic calculation gained under the water level, adjust the water level running orbit lower limit at the beginning of all periods of retrodicting.
When various requirements of comprehensive utilization are conflicting, determine priority from high to low with water level, flow, the order of exerting oneself.
2. backward induction method reservoir level running orbit maximum feasible scope:
According to the basic data of given reservoir and the comprehensive utilization and the constraint condition of day part, inquire into by computing machine:
As shown in Figure 2: from calculating end of term water level, according to the most last period is that minimum flow, lowest water level or the minimum generated output of letting out under the permission of N period requires reverse calculating, in conjunction with reservoir following ability of letting out and the unit output restriction of this moment, inquire into the possible water level lower limit of reservoir that obtains at the beginning of the N period, inquire into the water level lower limit of N-1 period more based on this according to identical method, till at the beginning of the given period, draw from calculating the feasible region that the end of term begins the day part reservoir level running orbit lower limit that backward induction method obtains.
As shown in Figure 2: from calculating end of term water level, according to maximum flow, peak level or the maximum generation of letting out under the i.e. permission of N period of the most last period requirement of exerting oneself, in conjunction with reservoir following ability of letting out and the unit output restriction of this moment, reverse calculating, inquire into the possible water level higher limit of reservoir that obtains at the beginning of the N period, inquire into the water level higher limit of N-1 period more based on this according to identical method, till at the beginning of the given period, draw from calculating the feasible region that the end of term begins the day part reservoir level running orbit upper limit that backward induction method obtains.
If the water level upper limit with certain period art begins to calculate, even if should the period according to letting out the water level upper limit requirement that does not also reach at the beginning of this period under the minimum letdown flow, shown in Fig. 5 n+1 period, be limited to the F point on the water level at the beginning of the requirement period, and when being the basis with the B point of period end, lowest water level also need be at the E point at the beginning of the period.At this moment, the reservoir level running orbit upper limit at the beginning of period is adjusted into the F point, and the water level upper limit of period end is adjusted as follows: based on the water level upper limit F point at the beginning of this period, inquire on the water level of this period end and be limited to the G point, if the G point is positioned at the beginning of next period that a following period Mo water level running orbit upper limit A point goes out for basic calculation within the water level coverage pattern B point and C point, be the period art reservoir level running orbit upper limit then with this G point, otherwise, be to inquire into a downward again period of basis with the G point again, till within the water level upper limit that obtains is positioned at reached at the water level range that is limited to the basic calculation gained on next water level more period Mo, adjust the water level running orbit upper limit that all are just pushing away the period art.
If the water level lower limit with certain period end begins to calculate, even if should the period do not reach the water level lower limit requirement of this period end yet, shown in Fig. 5 n period, require to be limited under the water level P point at the beginning of the period according to letting out ability (spacious letting out) under the maximum, and when being the basis, do not reach the P point at the beginning of the period with the M point of period art.At this moment, reservoir level running orbit lower limit at the beginning of period is adjusted into the P point, and the water level lower limit of period end is adjusted as follows: based on the water level lower limit P point at the beginning of this period, inquire under the water level of this period end and be limited to the Q point, if the Q point is positioned at the beginning of next period that a following period Mo water level running orbit lower limit K point goes out for basic calculation within the water level coverage pattern L point and M point, then with this Q point as this period end reservoir level running orbit lower limit, otherwise, serve as that the basis is to next period inquires into again with the Q point again, till within the water level lower limit that obtains is positioned at reached at the water level range that is limited to the basic calculation gained under next period art water level more, adjust the water level running orbit lower limit that all are just pushing away period end.
When various requirements of comprehensive utilization are conflicting, determine priority from high to low with water level, flow, the order of exerting oneself.
3. reservoir level running orbit maximum feasible scope is definite:
The result of comprehensive day part forward recursive and backward induction method draws schedule periods reservoir level running orbit maximum feasible scope, and its method is:
The forward recursive that obtains according to above-mentioned steps and the feasible region of backward induction method reservoir level running orbit lower limit, according to getting big principle with the period, draw from the beginning of the given period to the maximum feasible scope of the reservoir level running orbit lower limit that calculates the end of term; As shown in Figure 3.
The forward recursive that obtains according to above-mentioned steps and the feasible region of the backward induction method reservoir level running orbit upper limit, according to getting little principle with the period, draw from the beginning of the given period to the maximum feasible scope of the reservoir level running orbit upper limit of calculating the end of term; As shown in Figure 3.
4. reservoir level moves choosing of feasible track:
After having determined reservoir level running orbit maximum feasible scope, next select the possible strategy of reservoir operation.Because this maximum feasible scope is to change continuously under the maximum or minimum continuously situation at reservoir level to try to achieve, consideration be opposite extreme situations, therefore, be not that any track in this scope all is feasible track.
Seek feasible track in reservoir level running orbit maximum feasible scope, need to be undertaken by the period according to the period order.Initial water level at the beginning of given period is known, according to the minimum and maximum letdown flow of reservoir in this period, minimum and maximum generated output requires to calculate, in conjunction with reservoir following ability of letting out and the unit output restriction of this moment, calculate water level maximal value and the minimum value that reservoir may reach in period end by computing machine, if this value exceeds reservoir level running orbit maximum feasible scope, then adjust the maximal value and the minimum value of period art water level by the maximum feasible scope, obtain the upper and lower bound of period Mo reservoir level, in this scope, select a water level value as this water level by Deterministic Methods or random device period Mo, reservoir level at the beginning of i.e. second period is finished the calculating of a period.By period repetition above-mentioned steps, till calculating the end of term, the water level sequence that obtains promptly is the feasible track of a reservoir operation.
In this step, resulting every strategy is possible strategy; When day part end water level is selected, adopted Deterministic Methods and random device in selectable scope, obtained selecting, guaranteed the diversity of reservoir level operation possible strategy to guarantee the reservoir level operation strategy under border condition.
The concrete grammar of choosing that reservoir level moves feasible track is: reservoir level is Z at the beginning of the t period
t, according to the minimum outbound flow Q of reservoir in this period
T, minWith maximum outbound flow Q
T, maxRequirement, and the minimum generated output N of period
T, minWith the maximum generation N that exerts oneself
T, maxRequire etc., can calculate the water level maximal value Z that reservoir may reach in period end
T+1, maxWith minimum value Z
T+1, min, if water level maximal value and minimum value have exceeded the maximum feasible scope of SEA LEVEL VARIATION, then adjust the maximal value and the minimum value of period Mo water level by the maximum feasible scope, obtain the upper limit Z of period Mo water level
T+1, up, lower limit Z
T+1, low, the water level of next period of reservoir is only selected to be only feasible in this scope, can be by following Deterministic Methods or water level value of random device selection as this water level period Mo.
Deterministic Methods: day part is selected water level by same scale-up factor, and last water level can be chosen with following formula: to certain given α ∈ [0,1], Z
T+1=Z
T+1, low+ α (Z
T+1, up-Z
T+1, low), when producing numerous initial feasible solution, this scale-up factor can be chosen by from 0 to 1 arithmetic progression, and quantity is the bigger the better, and guarantees that the reservoir level operation strategy under limiting case and numerous intermediate state thereof can both obtain selecting; Also can choose by the different proportion coefficient, the scale-up factor of day part changes, can select from the beginning of the given period to calculating the end of term at times by from 0 to 1 scale-up factor, or by from 1 to 0 scale-up factor, data that also can various intermediatenesses are as scale-up factor, as from 0.5 to 1, or from 0.2 to 0.8 etc.
Random device: day part is all selected a scale-up factor between 0 and 1 at random in the optional scope of period Mo water level, try to achieve period Mo water level by following formula: to α ∈ [0,1], Z
T+1=Z
T+1, low+ α (Z
T+1, up-Z
T+1, low);
Comprehensive method: Deterministic Methods and random device combine, and the part period determines that with Deterministic Methods the part period is determined with random device.
5. the selection of the optimum solution of optimizing scheduling of reservoir:
Repeating step 4 can obtain many feasible tracks arbitrarily, calculates the target function value of every track, can adopt following method to carry out the selection of optimum solution:
1. from all feasible tracks that obtain select target functional value the maximum as approximate optimal solution;
2. several tracks are as the initial feasible solution of discrete differential dynamic programming (DDDP) preferably for the select target functional value, and the DDDP that carries out variable discrete steps again calculates, and obtains better approximate optimal solution, and this method is called the FS-DDDP method;
3. with the initial population of a large amount of feasible tracks as genetic algorithm (GA), carry out the GA computing again, the chromosome in each generation still is feasible solution at the new chromosome that intersects, will guarantee during mutation operation gained in the GA computing, obtains better approximate optimal solution, and this method is called the FS-GA method.
The Feasible searching method of optimizing scheduling of reservoir provided by the invention, it possesses: for the man-machine interaction on the man-machine interactive platform in the optimizing scheduling of reservoir system provides the purposes in the application.
In the man-machine interaction process, the decision maker can adjust constraint condition at any time according to the situation of approximate optimal solution, recomputates; If problem does not have feasible solution, judge when beginning by computing machine earlier, provide prompting then.
List of references
1. " reservoir optimizing and dispatching method is researched and analysed "
[1](people such as Cui Ruihong, Dong Zengchuan), download by following website:
http://www.paper.edu.cn/process/download.jsp？file＝200605-150
2. the few ripple of king is separated and is opened a position, Kong Ke, " application of self-adapted genetic algorithm in optimizing scheduling of reservoir " [2], water conservancy journal, in April, 2006
3. Xu Nin, Li Chunguang, Zhang Jian, Yu Juebang, the comparative studies of several modern optimization algorithms, systems engineering and electronic technology, 2002.12
4. Li Jin shields, Han Yanbin, Sun Zhisheng, the performance evaluation of chaos optimization algorithm, small-sized microcomputer system, 2005.08
5. Li Chong is great, and it is flourishing to record, and Li Wenwu improves particle swarm optimization and the application in optimizing scheduling of reservoir thereof, Chinese countryside water conservancy and hydropower, 2006 the 2nd phases
Claims (10)
1. the Feasible searching method of an optimizing scheduling of reservoir, the data that it is characterized in that being gathered are by Computer Processing, constitute the single optimizing scheduling of reservoir that is applicable to the complicated constraint condition of tool, or the Feasible searching method of multi-reservoir Optimization Dispatching and the water resource optimizing scheduling of reservoir of distributing rationally, but this method adopts and comprises the line search that following step is carried out optimizing scheduling of reservoir:
(1) forward recursive reservoir level running orbit feasible region:
According to the basic data of given reservoir and the comprehensive utilization and the constraint condition of day part, inquire into by computing machine:
From initial water level, according to this period is the minimum and maximum letdown flow of first time slot request, minimum and the peak level requirement calculating of minimum and maximum generated output and period Mo, in conjunction with reservoir following ability of letting out and the unit output restriction of this moment, obtain the water level lower limit that the reservoir at the beginning of i.e. second period of the first period Mo may reach, again to be limited to the basis under the reservoir level at the beginning of i.e. second period of the first period Mo, calculate water level lower limit at the beginning of i.e. the 3rd period of the second period Mo according to identical method, till calculating the end of term, draw the feasible region of the forward recursive reservoir level running orbit lower limit at the beginning of the given period
In like manner can release the feasible region of the forward recursive reservoir level running orbit upper limit;
(2) backward induction method reservoir level running orbit feasible region:
From calculating end of term water level, according to the minimum and the maximum flow of letting out under the most last period permission, minimum and maximum generation is exerted oneself and minimum and peak level requirement at the beginning of the period, in conjunction with reservoir following ability of letting out and the unit output restriction of this moment, reverse calculating, inquire into the possible water level lower limit of reservoir that obtains at the beginning of this period, serve as the last period is inquired on the basis according to identical method water level lower limit with the possible water level lower limit of the reservoir at the beginning of this period again, till at the beginning of the given period, draw from calculating the feasible region of the backward induction method reservoir level running orbit lower limit that the end of term begins
In like manner can draw the feasible region of the backward induction method reservoir level running orbit upper limit;
(3) determining of reservoir level running orbit maximum feasible scope:
The forward recursive that obtains according to above-mentioned steps and the feasible region of backward induction method reservoir level running orbit lower limit, according to getting big principle with the period, draw from the beginning of the given period to the maximum feasible scope of the reservoir level running orbit lower limit that calculates the end of term,
The forward recursive that obtains according to above-mentioned steps and the feasible region of the backward induction method reservoir level running orbit upper limit, according to getting little principle with the period, draw from the beginning of the given period to the maximum feasible scope of the reservoir level running orbit upper limit of calculating the end of term;
(4) reservoir level moves choosing of feasible track:
According to the period order, given reservoir level at the beginning of first period, according to the minimum and maximum letdown flow of reservoir in this period, minimum and the peak level requirement calculating of minimum and maximum generated output and period Mo, in conjunction with reservoir following ability of letting out and the unit output restriction of this moment, calculate water level maximal value and the minimum value that reservoir may reach in period end by computing machine, if this value exceeds reservoir level running orbit maximum feasible scope, then adjust the maximal value and the minimum value of period Mo water level by the maximum feasible scope, obtain the upper and lower bound of period Mo reservoir level, in this scope, select a water level value as this water level by Deterministic Methods or random device period Mo, reservoir level at the beginning of i.e. second period, finish the calculating of a period
Repeat said process, when calculating, till the end of term, so just obtain the feasible track of a reservoir level operation;
(5) selection of the optimum solution of optimizing scheduling of reservoir:
Repeating step (4) can obtain many feasible tracks arbitrarily, calculates the target function value of every track, adopts several different methods to carry out the selection of optimum solution then: direct select target functional value the maximum is as approximate optimal solution from all feasible tracks that obtain; The feasible track of being tried to achieve as initial feasible solution, is asked approximate optimal solution in conjunction with other optimized Algorithm; Other optimized Algorithm comprises the FS-DDDP method that constitutes with DDDP, and with the FS-GA method that GA constitutes, DDDP is the discrete differential dynamic programming, and GA is a genetic algorithm.
2. the Feasible searching method of optimizing scheduling of reservoir according to claim 1 is characterized in that: in the step (1), reach the water level of setting at the calculation interval end for guaranteeing reservoir level, adopt the method for backward induction method reservoir level running orbit feasible region.
3. the Feasible searching method of optimizing scheduling of reservoir according to claim 1 is characterized in that: the result of comprehensive day part forward recursive and backward induction method draws schedule periods reservoir level running orbit maximum feasible scope.
4. the Feasible searching method of optimizing scheduling of reservoir according to claim 1, it is characterized in that: in the backward induction method process of the forward recursive of step (1) and step (2), feature according to reservoir, satisfy various requirements of comprehensive utilization as much as possible, if various requirement is conflicting, then determine priority from high to low with water level, flow, the order of exerting oneself.
5. the Feasible searching method of optimizing scheduling of reservoir according to claim 1, it is characterized in that: in the step (4), resulting every strategy is possible strategy; When day part end water level is selected, adopted Deterministic Methods and random device in selectable scope, obtained selecting, guaranteed the diversity of reservoir level operation possible strategy to guarantee the reservoir level operation strategy under border condition.
6. the Feasible searching method of optimizing scheduling of reservoir according to claim 1 is characterized in that: adopt the optimum solution of the feasible track of combined method acquisition reservoir level in the step (5), i.e. FS-DDDP method, FS-GA method.
7. the Feasible searching method of optimizing scheduling of reservoir according to claim 1 is characterized in that: for the man-machine interaction on the man-machine interactive platform in the optimizing scheduling of reservoir system provides the purposes in the application.
8. the Feasible searching method of optimizing scheduling of reservoir according to claim 7, it is characterized in that: in the man-machine interaction process, the decision maker can adjust constraint condition at any time according to the situation of approximate optimal solution, recomputates; If problem does not have feasible solution, judge when beginning by computing machine earlier, provide prompting then.
9. the purposes during the purposes of the Feasible searching method of optimizing scheduling of reservoir in carrying out single optimizing scheduling of reservoir application, or multi-reservoir Optimization Dispatching used.
10. the Feasible searching method of optimizing scheduling of reservoir is distributed purposes in the application rationally carrying out water resource.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101253455A CN1971601A (en) | 2006-12-07 | 2006-12-07 | Feasible searching method of optimizing scheduling of reservoir |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101253455A CN1971601A (en) | 2006-12-07 | 2006-12-07 | Feasible searching method of optimizing scheduling of reservoir |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1971601A true CN1971601A (en) | 2007-05-30 |
Family
ID=38112407
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2006101253455A Pending CN1971601A (en) | 2006-12-07 | 2006-12-07 | Feasible searching method of optimizing scheduling of reservoir |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1971601A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102004835A (en) * | 2010-11-26 | 2011-04-06 | 武汉大学 | Power generation risk-oriented hydropower station optimal operation chart drawing method |
| CN101714186B (en) * | 2009-12-29 | 2011-05-18 | 北京师范大学 | Method of optimizing and determining water supply type reservoir dispatching diagram considering human and ecological needs |
| CN102780235A (en) * | 2012-08-02 | 2012-11-14 | 南通大学 | Pumped storage power station dispatching method on basis of genetic algorithm |
| CN102867226A (en) * | 2012-10-16 | 2013-01-09 | 国网电力科学研究院 | Optimizing and dispatching method of hydropower station |
| CN103544533A (en) * | 2012-07-12 | 2014-01-29 | 上海宝信软件股份有限公司 | Diversion project water resource optimal allocation dynamic model establishing system |
| CN104049532A (en) * | 2013-03-15 | 2014-09-17 | 洛克威尔自动控制技术股份有限公司 | Sequential deterministic optimization based control system and method |
| CN104063808A (en) * | 2014-06-27 | 2014-09-24 | 大连理工大学 | Trans-provincial power transmission cascade hydropower station group peak-shaving dispatching two-phase search method |
| CN105354630A (en) * | 2015-10-20 | 2016-02-24 | 华中科技大学 | Variable-stage gradual optimization based joint optimization scheduling method for cascaded hydropower stations |
| CN106845712A (en) * | 2017-01-20 | 2017-06-13 | 中国水利水电科学研究院 | The extracting method of scheduling rule is decomposed in polymerization towards big system multi-reservoir |
| CN112633578A (en) * | 2020-12-24 | 2021-04-09 | 国电电力发展股份有限公司和禹水电开发公司 | Optimized dispatching method for lower-grade reservoir group under influence of diversion project |
-
2006
- 2006-12-07 CN CNA2006101253455A patent/CN1971601A/en active Pending
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101714186B (en) * | 2009-12-29 | 2011-05-18 | 北京师范大学 | Method of optimizing and determining water supply type reservoir dispatching diagram considering human and ecological needs |
| CN102004835A (en) * | 2010-11-26 | 2011-04-06 | 武汉大学 | Power generation risk-oriented hydropower station optimal operation chart drawing method |
| CN103544533A (en) * | 2012-07-12 | 2014-01-29 | 上海宝信软件股份有限公司 | Diversion project water resource optimal allocation dynamic model establishing system |
| CN103544533B (en) * | 2012-07-12 | 2016-09-21 | 上海宝信软件股份有限公司 | A kind of water diversion project water resource optimal allocation dynamic model constructing system |
| CN102780235A (en) * | 2012-08-02 | 2012-11-14 | 南通大学 | Pumped storage power station dispatching method on basis of genetic algorithm |
| CN102780235B (en) * | 2012-08-02 | 2014-10-08 | 南通大学 | Pumped storage power station dispatching method on basis of genetic algorithm |
| CN102867226B (en) * | 2012-10-16 | 2016-01-20 | 国网电力科学研究院 | A kind of method of Optimized Scheduling of Hydroelectric Power |
| CN102867226A (en) * | 2012-10-16 | 2013-01-09 | 国网电力科学研究院 | Optimizing and dispatching method of hydropower station |
| CN104049532B (en) * | 2013-03-15 | 2017-04-12 | 洛克威尔自动控制技术股份有限公司 | Sequential deterministic optimization based control system and method |
| CN104049532A (en) * | 2013-03-15 | 2014-09-17 | 洛克威尔自动控制技术股份有限公司 | Sequential deterministic optimization based control system and method |
| US10317857B2 (en) | 2013-03-15 | 2019-06-11 | Rockwell Automation Technologies, Inc. | Sequential deterministic optimization based control system and method |
| US10871752B2 (en) | 2013-03-15 | 2020-12-22 | Rockwell Automation Technologies, Inc. | Sequential deterministic optimization based control system and method |
| CN104063808B (en) * | 2014-06-27 | 2015-06-24 | 大连理工大学 | Trans-provincial power transmission cascade hydropower station group peak-shaving dispatching two-phase search method |
| CN104063808A (en) * | 2014-06-27 | 2014-09-24 | 大连理工大学 | Trans-provincial power transmission cascade hydropower station group peak-shaving dispatching two-phase search method |
| CN105354630A (en) * | 2015-10-20 | 2016-02-24 | 华中科技大学 | Variable-stage gradual optimization based joint optimization scheduling method for cascaded hydropower stations |
| CN105354630B (en) * | 2015-10-20 | 2019-11-12 | 华中科技大学 | A kind of step hydropower station joint optimal operation method based on change stage successive optimization |
| CN106845712A (en) * | 2017-01-20 | 2017-06-13 | 中国水利水电科学研究院 | The extracting method of scheduling rule is decomposed in polymerization towards big system multi-reservoir |
| CN106845712B (en) * | 2017-01-20 | 2018-05-11 | 中国水利水电科学研究院 | The extracting method of scheduling rule is decomposed in polymerization towards big system multi-reservoir |
| CN112633578A (en) * | 2020-12-24 | 2021-04-09 | 国电电力发展股份有限公司和禹水电开发公司 | Optimized dispatching method for lower-grade reservoir group under influence of diversion project |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1971601A (en) | Feasible searching method of optimizing scheduling of reservoir | |
| CN109713737B (en) | New energy delivery capacity evaluation method and system for accessing flexible direct current power grid | |
| CN102097866B (en) | Mid-long-term unit commitment optimizing method | |
| CN105046395B (en) | Method for compiling day-by-day rolling plan of power system containing multiple types of new energy | |
| CN102855591A (en) | Method and system for optimizing scheduling for short-term combined generation of cascade reservoir group | |
| WO2017071230A1 (en) | Method for short-term optimal scheduling of multi-agent hydropower station group | |
| CN105787589B (en) | Carry-over storage drought is restricted water supply the method for optimally controlling and dedicated control system of position | |
| CN106200577A (en) | A kind of multi-storied garage dispatch control method | |
| CN102682409A (en) | Optimal scheduling method of nonlinear-programming cascade reservoir group based on GAMS (general algebraic modeling system) | |
| Archibald et al. | Nested Benders decomposition and dynamic programming for reservoir optimisation | |
| CN112862194B (en) | Power distribution network power supply planning method, device, equipment and readable storage medium | |
| CN102298731A (en) | Cascade reservoir short-term electricity generation optimal dispatching method considering comprehensive requirements of tide stemming water supply | |
| CN109978404A (en) | The passway for transmitting electricity Expansion Planning method that meter and uncertain factor and scene are cut down | |
| CN104636830A (en) | Water power and thermal power generation real-time load adjusting method of provincial grid under inflow change | |
| CN108649605A (en) | A kind of grid-connected allowed capacity planing methods of DER based on the double-deck scene interval trend | |
| CN102034159A (en) | Intelligent scheduling system of large-scale spanning watershed hydroelectric power stations | |
| CN112072710B (en) | Source network load integrated economic dispatching method and system considering demand response | |
| CN108695859A (en) | A kind of capacitor group addressing constant volume method of access active distribution network | |
| CN109472463B (en) | Multi-power-grid peak regulation scheduling method and system for hydropower station group in extra large basin | |
| CN102593855B (en) | Method for stabilizing fluctuation of output power of renewable energy power supply in power system | |
| CN111082466A (en) | New energy access and grid frame extension optimization method considering wind power uncertainty | |
| Simonovic | The implicit stochastic model for reservoir yield optimization | |
| CN103276704B (en) | Determination method of hydropower station water storing and releasing dispatch and operation scheme based on energy storage analysis | |
| CN102904248A (en) | Electric power system dispatching method based on wind electricity output uncertainty aggregation | |
| CN113991640A (en) | Multi-energy complementary energy base energy configuration planning method based on thermal power |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20070530 |