CN106127336A - Small hydropower station optimal scheduling method based on multi-target moth algorithm - Google Patents

Abstract

The invention relates to a small hydropower station optimal scheduling method based on a multi-objective moth algorithm. And then, taking the established model as a target function to be brought into a multi-target moth algorithm for optimization calculation, finally returning a set containing an optimized scheduling scheme after the optimization calculation is carried out through the algorithm, and finally making the scheduling scheme by a decision maker by referring to the given optimized scheduling scheme set. The method provided by the invention emphasizes on improving the accuracy and high efficiency of the optimized dispatching of the small hydropower station, solves the problems existing in the prior art on models and methods, and has important significance for promoting the development of the optimized dispatching of the small hydropower station and improving the economic benefit.

Description

A kind of small hydropower station Optimization Scheduling based on multiple target moth algorithm

Technical field

The present invention relates to small water conservancy hydroelectric project run and Optimized Operation field, particularly relate to one and fly based on multiple target The small hydropower station Optimization Scheduling of moth algorithm.

Background technology

Along with the development of China's economy, people's living standard gradually steps up, the most increasing to the demand of electric power, by This has also driven developing rapidly of each power industry, and the even tidal power generation of water power, thermoelectricity, nuclear power, photovoltaic industry, wind-power electricity generation is all There is significant progress, achieve remarkable achievement.By the end of the year in 2010, whole nation hydropower installed capacity reached 2.1 hundred million kW, Ranking first in the world, annual electricity generating capacity reaches 563,300,000,000 kW h, accounts for whole nation electric power installed capacity and 21.6% He of annual electricity generating capacity respectively 16.4%.Relative to thermal power plant, power station not only has steadily in the long term, operating cost is low and the significant advantage of peak-frequency regulation, and And along with country's raising to environment protection control requirements, the environmental protection advantage of water power is more noticeable.China's water resource is enriched, Water energy technical exploitation amount is about 5.42 hundred million kW, and that has the most developed is only the 36% of total amount.

Small power station is also under the jurisdiction of water power category, is in the end in all power stations for calm measuring angle, and it generally refers to The power station of capacity 50,000 below kW.China's small hydropower resources is concentrated mainly on vast rural area and remote mountain areas, therefore small power station Stand and be commonly referred to rural small hydropower.

Rural hydropower based on small power station is clean reproducible energy, is the important foundation of rural economy social development Facility, is also the important means of environmental conservation and ecological construction.China's small hydropower resources is the abundantest, and technical exploitation amount reaches 1.28 hundred million kW.At the end of 2010, built small hydropower station 4.5 ten thousand, whole nation rural hydropower installation amount is more than 59,000,000 kW, Nian Fa Electricity 200,000,000,000 kW h；Small power station has spread all over the country the region of 1/2, the counties and cities of 1/3.

In recent years, although to optimizing scheduling of reservoir research and practice, oneself becomes through achieving much gratifying research both at home and abroad Really, goal in research develops into step and across basin multi-reservoir from single reservoir, and stream flow description describes from deterministic type and develops at random Type describes, and various optimum theories, algorithm and Optimal Operation Model are also evolving, and some aspect is also answered in the middle of reality With but for the Optimized Operation of rural small hydropower, owing to rural small hydropower has with great river large watershed is different, no matter Being from basin area coverage or hydrology and water conservancy characteristic, rural small hydropower all has the aspect of oneself uniqueness, therefore, optimization routine Some of scheduling are theoretical and computational methods are not applied for rural small hydropower, need to explore a set of suitable theory and process；Separately On the one hand, due to features such as the complexity of water reservoir system, multiformity, the mathematical model of various intelligent optimization methods and scheduling is often There is dimension calamity and cause being difficult to solve, and the inconsiderate congruence shortcoming that considers a problem, Optimal Scheduling of Multi-reservoir System problem is far from reaching To satisfactory solution, remain bigger gap between optimum theory and actual application, await our further effort Theory is made preferably to serve practice.

Summary of the invention

The present invention is to overcome above-mentioned weak point, it is therefore intended that provide a kind of miniature water based on multiple target moth algorithm Power optimization dispatching method, the inventive method is for the shortcoming of existing reservoir operation technology, it is provided that one flies based on multiple target The small hydropower station Optimization Scheduling of moth algorithm.The present invention improves emphatically accuracy and the high efficiency of small hydropower station Optimized Operation, Solve prior art in problem present on model and method, the development that promotion small hydropower station optimization is adjusted, raising economy effect Yidu has great significance.

The present invention is to reach above-mentioned purpose by the following technical programs: a kind of small-sized water power based on multiple target moth algorithm Stand Optimization Scheduling, comprise the steps:

(1) the essential information data in collection scheduling target power station, and combine constraints based on essential information data and build Vertical target is the mathematical model that Energy Maximization is minimum with ecological holographic amount；

(2) model of foundation is brought in multiple target moth algorithm as object function it is optimized calculating, returned Set containing Optimized Operation scheme；

(3) policymaker is with reference to calculated Optimized Operation scheme set, uses Multiobjective Decision Making Method to formulate dispatching party Case.

As preferably, described constraints for combine storage capacity constraint, water quantity restraint, power generation dispatching constraint, constraint of supplying water, The constraints of boundary constraint.

As preferably, the step that described step (2) obtains returning the set containing Optimized Operation scheme is as follows:

1) determine the bound of day part regulation goal Hydropower Plant Reservoir water level value, set initial parameter；

2) carry out regulation goal Hydropower Plant Reservoir initializing population and initializing adaptive mesh mechanism；

3) initialization population is carried out preliminary optimizing outside storing initial non-dominant disaggregation；

4) it is updated optimizing according to algorithmic formula, and judges whether optimizing local iteration reaches end condition, wherein, eventually Only condition is default；

5) if reaching end condition, updating all parameters and finding optimum non-dominant disaggregation, updating step 3) outside storage Non-dominant disaggregation；If not up to end condition, redirect execution step 4)；

6) judge whether outside storage reaches maximum storage value, the non-dominant disaggregation if reaching maximum storage value, to storage Carry out screening and perform step 7 after rejecting operation)；If not up to, being directly entered step 7)；

7) judge non-domination solution concentrates whether there is solution beyond border, if beyond, update grid mechanism and cover beyond border Solution, enter step 8)；If no, being directly entered step 8)；

8) judging whether to reach overall situation end condition, wherein overall situation end condition is default；If reaching the overall situation to terminate bar Part, the output set containing Optimized Operation scheme；If not up to, redirecting execution step 4).

As preferably, described setting initial parameter includes determining target dimension, initial population number, iterations, outside Store size, the grid coefficient of expansion, select operation ginseng

Number, deletion action parameter.

As preferably, described step 4) to be updated the formula of optimizing according to algorithmic formula as follows:

M_i=S (M_i,F_j)

Wherein, M_iRepresent i-th moth, F_jRepresenting jth group light source, S is Spirallike Functions, and wherein the starting point of spiral is moth Position, terminal is the position of light source.

As preferably, described Spirallike Functions S formula is as follows:

S(M_i, F_j)=D_i·e^ht·cos(2πt)+F_j

Wherein, D_iRepresent i-th moth and jth group light source between distance, h is the constant of screw, t be interval [- 1,1] random number on；Solve D_iFormula be defined as follows:

D_i=| F_j-M_i|

Wherein M_iRepresent i-th moth, F_jRepresent jth group light source.

As preferably, described step 6) carry out screening and use the form of roulette dish to screen, its screening probability is by such as Lower formula defines:

$P_i=\frac{\mathrm{\α}}{N_i}$

Wherein, α is the constant that a ratio 1 is big, N_iRepresent the number having obtained non-domination solution in i-th section.

The beneficial effects of the present invention is: (1) meets the requirement of reservoir Multiobjective Optimal Operation；(2) initial population is utilized The ergodic of high initial population, randomness and multiformity, can avoid random initial population of poor quality and concentrate on some Regional area is so that the algorithm problem that is absorbed in local optimum；(3) the outside storage of Dynamic Updating Mechanism is used, it is ensured that non-domination solution Individuality is evenly distributed, and has good multiformity, accelerates global convergence.

Accompanying drawing explanation

Fig. 1 is the main-process stream schematic diagram of Optimization Scheduling of the present invention；

Fig. 2 is the signal of embodiment of the present invention small hydropower station based on multiple target moth algorithm Optimization Scheduling flow process Figure.

Detailed description of the invention

Below in conjunction with specific embodiment, the present invention is described further, but protection scope of the present invention is not limited in This:

Embodiment: as shown in Figure 1 and Figure 2, the present invention is directed to traditional method, easily to sink into locally optimal solution, convergence rate slower Etc. defect, and for avoiding the occurrence of initial population distribution property difference, searching process is had an impact, it is provided that a kind of based on multiple target moth The small hydropower station Optimization Scheduling of algorithm, the method uses the outside storage of Dynamic Updating Mechanism and introduces moth orientation Spiral search process, it is achieved the ability of algorithm global optimizing, and on the basis of non-dominant disaggregation, use Multi-Objective Decision Theory From main separation optimal scheduling scheme, it is achieved the Multiobjective Optimal Operation of small hydropower station, concrete steps include:

Step (1), sets up small hydropower station power benefit model:

S=∑ c_iN_iΔt (1)

N=kQH (2) power benefit S depends on efficiency k, generating flow Q, productive head H and electricity price factor c.For machine The small hydropower station that type is fixing, k value is relevant with Q, H, wherein k=f (Q, H).Q is generating flow, is typically based on the water yield and regulation Depending on degree.

Step (2), in conjunction with practical situation and the market demand, sets up the scheduling model with Energy Maximization as target:

$F=max\underset{i=1}{\overset{T}{\Σ}}{\mathrm{KQ}}_t\stackrel{\‾}{H_t}\mathrm{\Δ}t---\left(3\right)$

Wherein F is the annual electricity generating capacity in power station, and K is the comprehensive power factor in power station, reflects the potential energy of current For the energy loss in power process.Q_tFor power station at the generating flow of t period,Send out the average of t period for power station Electricity net water head, T is calculating total period in the scheduling year of power station, and Δ t is the length of t period.

Step (3), in conjunction with electricity needs, from the angle of market economy, in conjunction with the feature of water power to transporting at electric power Battalion obtains the economic benefit of maximum.Optimized model with maximizing generation profit as target is as follows:

$F=max\underset{t=1}{\overset{T}{\Σ}}{\mathrm{Kp}}_t{Q}_t\stackrel{\‾}{H_t}\mathrm{\Δ}t---\left(4\right)$

Wherein, F is the year gene-ration revenue in power station, and K is the comprehensive power factor in power station, and the potential energy reflecting current turns Turn to the energy loss in power process.p_tFor the electricity price factor, Q_tFor power station at the generating flow of t period,For power station In the net water head that averagely generates electricity of t period, T is calculating total period in the scheduling year of power station, and Δ t is the length of t period.

Step (4), for the water demand for natural service in river course, The more the better from the point of view of ecological view, but be built upon ensureing necessarily On the premise of generated energy.Here, the water demand for natural service in river course is converted into river channel ecology hydropenia, hereafter carries out multiple target to facilitate Optimization processes, and sets up water demand for natural service model:

$F=\min W_Q=\{\begin{array}{cc}(Q_x-Q_s)\×\mathrm{\Δ}t& Q_x\≤Q_s\\ 0& Q_x>Q_s\end{array}---\left(5\right)$

Wherein, W_QRepresent river channel ecology hydropenia, Q_xAnd Q_sRepresent letdown flow and river channel ecology flow respectively.

Step (5), sets up constraints, sends out scheduling about including reservoir water quantity restraint, reservoir capacity constraint, reservoir The constraint of bundle, output, boundary condition constraint.Its particular content is as follows:

Reservoir water balance side becomes to retrain:

V_t+1=V_t+(q_t-Q_t)K_t (6)

Wherein V_t+1Represent the reservoir storage of small hydropower station t period, v_tRepresent the reservoir storage at the beginning of the small hydropower station t period, q_t Represent the average reservoir inflow of small hydropower station t period, Q_tRepresent the generated energy of small hydropower station t period, K_tRepresent time conversion Coefficient.

Reservoir capacity retrains:

max[VL_t,V_et]≤V_t≤VH_t (7)

VL_tRepresent the minimum capacity of a reservoir of reservoir, V_etRepresent the ecological characteristic storages of reservoir of reservoir, VH_tRepresent the utilizable capacity of reservoir.

Hydropower station schedule constraints:

Qq_t≥Qd_min (8)

Wherein Qd_minRepresent t period minimum generating flow.

Supply water and retrain:

Uq_t≥Qg_min (9)

Wherein Qg_minRepresent t period minimum output.

Boundary condition retrains:

V_(k,1)=V_(k,n) (10)

Wherein V_(k,1)Represent the initial storage after iteration K time, V_(k,n)Represent the last storage capacity after iteration K time.

Step (6), for the model set up, uses multiple target moth Algorithm for Solving Multiobjective Optimal Operation mathematical model Pareto optimal solution.Specific as follows:

1) bound Ub of day part reservoir water place value, Lb are determined；

2) setting initial parameter, including determining target dimension, initial population number, iterations, outside storage is big Little, the grid coefficient of expansion, selection manipulation parameter, deletion action parameter.

3) carry out reservoir initializing kind of a group operation.

4) adaptive mesh mechanism is initialized.

5) initialization population is carried out preliminary optimizing and stores initial non-domination solution.

6) being updated optimizing according to algorithmic formula, its formula is as follows:

M_i=S (M_i,F_j) (11)

Wherein, M_iRepresent i-th moth, F_jRepresenting jth group light source, S is Spirallike Functions, and wherein the starting point of spiral is moth Position, terminal is the position of light source.In algorithm, Spirallike Functions S formula is as follows:

S(M_i, F_j)=D_i·e^ht·cos(2πt)+F_j

(12)

Wherein D_iRepresent i-th moth and jth group light source between distance, h is the constant of screw, t be interval [- 1,1] random number on.Solve D_iFormula be defined as follows:

D_i=| F_j-M_i| (13)

Wherein M_iRepresent i-th moth, F_jRepresent jth group light source.

7) judge whether local iteration reaches end condition, such as not up to return 6；

8) all parameters are updated.

9) optimum non-domination solution need to be looked for, and update stored data in outside storage.

10) judge outside storage condition, if reaching maximum storage value, carrying out screening and deleting and more new data, at algorithm In, use the form of roulette dish to screen, its screening probability is defined by equation below:

$P_i=\frac{\mathrm{\α}}{N_i}---\left(14\right)$

Wherein, α is the constant that a ratio 1 is big, N_iRepresent the number having obtained non-domination solution in i-th section.

11) determine whether to solve beyond border, except covering new explanation beyond then renewal grid mechanism

12) judge whether to reach overall situation end condition, as not up to, return 6).

13) optimal solution set is returned.

Step (7), the optimal solution collection that policymaker is given with algorithm is combined into reference, considers formulation dispatching party Case.

It is the specific embodiment of the present invention and the know-why used described in Yi Shang, if conception under this invention institute Make change, function produced by it still without departing from description and accompanying drawing contained spiritual time, must belong to the present invention's Protection domain.

Claims (7)

1. a small hydropower station Optimization Scheduling based on multiple target moth algorithm, it is characterised in that comprise the steps:

(1) the essential information data in collection scheduling target power station, and combine constraints based on essential information data and set up mesh It is designated as the mathematical model that Energy Maximization is minimum with ecological holographic amount；

(2) model of foundation is brought in multiple target moth algorithm as object function it is optimized calculating, obtain return and contain The set of Optimized Operation scheme；

(3) policymaker is with reference to calculated Optimized Operation scheme set, uses Multiobjective Decision Making Method to formulate scheduling scheme.

A kind of small hydropower station Optimization Scheduling based on multiple target moth algorithm the most according to claim 1, it is special Levy and be: described constraints is to combine storage capacity constraint, water quantity restraint, power generation dispatching constraint, supply water constraint, boundary constraint Constraints.

A kind of small hydropower station Optimization Scheduling based on multiple target moth algorithm the most according to claim 1, it is special Levy and be: the step that described step (2) obtains returning the set containing Optimized Operation scheme is as follows:

1) determine the bound of day part regulation goal Hydropower Plant Reservoir water level value, set initial parameter；

2) carry out regulation goal Hydropower Plant Reservoir initializing population and initializing adaptive mesh mechanism；

3) initialization population is carried out preliminary optimizing outside storing initial non-dominant disaggregation；

4) it is updated optimizing according to algorithmic formula, and judges whether optimizing local iteration reaches end condition, wherein, terminate bar Part is default；

5) if reaching end condition, updating all parameters and finding optimum non-dominant disaggregation, updating step 3) outside non-stored Join disaggregation；If not up to end condition, redirect execution step 4)；

6) judging whether outside storage reaches maximum storage value, if reaching maximum storage value, the non-dominant disaggregation of storage being carried out Screening performs step 7 after rejecting operation)；If not up to, being directly entered step 7)；

7) judge non-domination solution concentrates whether there is solution beyond border, if beyond, update grid mechanism and cover beyond border Solve, enter step 8)；If no, being directly entered step 8)；

8) judging whether to reach overall situation end condition, wherein overall situation end condition is default；If reaching overall situation end condition, defeated Go out the set containing Optimized Operation scheme；If not up to, redirecting execution step 4).

A kind of small hydropower station Optimization Scheduling based on multiple target moth algorithm the most according to claim 3, it is special Levy and be: described setting initial parameter includes determining target dimension, initial population number, iterations, external storage size, net The lattice coefficient of expansion, selection manipulation parameter, deletion action parameter.

A kind of small hydropower station Optimization Scheduling based on multiple target moth algorithm the most according to claim 3, it is special Levy and be: described step 4) to be updated the formula of optimizing according to algorithmic formula as follows:

M_i=S (M_i,F_j)

Wherein, M_iRepresent i-th moth, F_jRepresenting jth group light source, S is Spirallike Functions, and wherein the starting point of spiral is the position of moth Putting, terminal is the position of light source.

A kind of small hydropower station Optimization Scheduling based on multiple target moth algorithm the most according to claim 5, it is special Levy and be: described Spirallike Functions S formula is as follows:

S(M_i, F_j)=D_i·e^ht·cos(2πt)+F_j

Wherein, D_iRepresenting the distance between i-th moth and jth group light source, h is the constant of screw, and t is interval [-1,1] On random number；Solve D_iFormula be defined as follows:

D_i=| F_j-M_i|

Wherein M_iRepresent i-th moth, F_jRepresent jth group light source.

A kind of small hydropower station Optimization Scheduling based on multiple target moth algorithm the most according to claim 3, it is special Levy and be: described step 6) carry out screening use roulette dish form screen, its screening probability defined by equation below:

$P_i=\frac{\mathrm{\α}}{N_i}$

Wherein, α is the constant that a ratio 1 is big, N_iRepresent the number having obtained non-domination solution in i-th section.