CN106295883A - Based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes - Google Patents

Abstract

The invention discloses a kind of based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes, including: S1 sets up microgrid multi-objective optimization scheduling model；S2 initial beggar population at individual, repairs the individuality of sub-population according to the constraint of micro-source climbing rate；S3 carries out genetic manipulation respectively and obtains the sub-population of filial generation each sub-population；The individuality of the sub-population of filial generation is repaired according to the constraint of micro-source climbing rate；S4 merges all of sub-population and the sub-population of filial generation generates mixed population, and mixed population is divided into new sub-population；The scale reparation of each new sub-population is the sub-population scale set by S5S；S6 retains all non-bad individualities in new sub-population；S7 judges whether to meet the condition of convergence, if meeting, retaining and exporting all feasible individuals.The present invention devises the process strategy of climbing rate constraint, keeps multiformity and the adaptivity solved by the way of " partition protecting ", meets the robustness demand under all kinds of Parameters variation during microgrid runs.

Description

Based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes

Technical field

The invention belongs to microgrid Optimized Operation field, particularly relate to a kind of based on improving the micro-of subregion multi-target evolution optimization Net Optimization Scheduling.

Background technology

The scheduling of microgrid multiple target running optimizatin is the Important Problems that industrial quarters is paid close attention in recent years.In state of the art Restriction under, by the scheme of exerting oneself of each micro battery in research microgrid, give full play to microgrid flexible, economical, the advantage of environmental protection, and Coordinate bulk power grid safe and reliable operation, will be to promoting that power system energy-saving and emission-reduction have important practical significance.To conventional electric power For system, microgrid structure is complicated, both uses regenerative resource and uncontrollable micro-electricity just like wind-driven generator, photovoltaic cell etc. Source, uses non-renewable energy resources just like miniature gas turbine, fuel cell etc. again and exports controlled micro battery, the most also including The energy storage devices such as accumulator.Therefore, microgrid multi-objective optimization scheduling presents that variable composition is various, object function is non-linear By force, constraints is complicated and the Parameters variation feature such as frequently.

For microgrid Multi-Objective Scheduling, owing to the Optimized model under different operational modes and constraints make The complexity obtaining optimization problem has bigger difference, and the optimized algorithm therefore used also is not quite similar.The most the more commonly used Mathematical Planning and intelligent algorithm etc., although achieving certain entering in nonlinear optimal problem, constrained optimization problems solving Step, but inefficient for solving microgrid multi-objective optimization scheduling, exist and priori is required that height, poor robustness etc. lack Point.On the other hand, existing to the process of constraints in microgrid multi-objective optimization scheduling, use based on dominance relation more Individual ordering strategy, and the climbing rate such as that have ignored violates this kind of more difficult constraint directly eliminated by penalty of item.Cause This, the ability of the microgrid multi-objective optimization question that existing optimisation technique processes belt restraining is poor, and obtainable feasible individual is relatively Few, it is impossible to meet the demand of practical engineering application.

Summary of the invention

It is an object of the invention to provide a kind of based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes, should Method devises the process strategy of climbing rate constraint, keeps multiformity and the adaptivity solved by the way of " partition protecting ", Meet the robustness demand under all kinds of Parameters variation during microgrid runs.

For reaching above-mentioned purpose, it is a kind of based on improving the microgrid optimization tune that subregion multi-target evolution optimizes that the present invention provides Degree method, including step:

S1 sets up the microgrid multi-objective optimization scheduling model being made up of object function and constraints；Described constraint Condition includes that micro-source climbing rate retrains；

S2 initial beggar population at individual, and repair sub-population at individual according to the constraint of micro-source climbing rate；Initialize evolutionary generation Gen=0；

Described repairs sub-population at individual according to the constraint of micro-source climbing rate, particularly as follows:

1. the sub-variable vector P=[P that each micro-source in current sub-population that obtains is corresponding₁,P₂,...,P_M], make m=2；

If 2. | P_m-P_m-1|≤P_r, do not repair；If P_m-P_m-1＞ P_r, by P_mRepair as P_m=min (P_max,P_m+P_r)；If P_m-1-P_m＞ P_r, by P_mRepair as P_m=max (P_min,P_m-P_r)；

3. judge that m, whether equal to M, if being equal to, terminates；Otherwise, make m=m+1, repeat sub-step 2.；

Wherein, P_mAnd P_m-1Represent that micro-source of m, m-1 time point is actual respectively to exert oneself, when micro-source is energy storage device, Actual the exerting oneself in micro-source uses filling or discharge capacity of micro-source；M=1,2 ... M, M express time point sum；P_r、P_max、P_minTable respectively Show the maximum climbing rate in micro-source, EIAJ, minimum load；

S3 makes evolutionary generation gen=gen+1, performs step S4；

S4 carries out genetic manipulation respectively and obtains the sub-population of filial generation each sub-population, and makes the sub-population scale of filial generation reach to set Sub-population scale S；

S5 merges all of sub-population and the sub-population of filial generation generates mixed population, according to individual in its target in mixed population The distribution of subspace, space, is divided into new sub-population by mixed population；

The scale reparation of each new sub-population is the sub-population scale S set by S6, and this step farther includes:

In 6.1 pairs of mixed populations, individuality carries out prioritization；

6.2 pairs of scales are more than the new sub-population of S, reject q the individuality that priority is minimum from new sub-population；

6.3 pairs of scales are less than the new sub-population of S, from subspace, new sub-population place, choose and the subregion of new sub-population Q the feasible individual new sub-population of addition that vector distance is the shortest；If the feasible individual lazy weight q outside subspace, then by preferential Level selects infeasible individual to add new sub-population from subspace, and the individual sum adding new sub-population is q；

6.4 pairs of scales are equal to the new sub-population of S, do not process；

Described q represents the difference of the sub-population scale of new sub-population scale and setting；

S7 retains all non-bad individualities in new sub-population；

S8 judges whether to meet the condition of convergence, if meeting, retaining and exporting all feasible individuals；Otherwise, step S3 is returned.

Further, object function includes the minimum object function of integrated cost and discharge amount of pollution minimum target function.

Further, constraints also include micro-source exert oneself bound constraint, energy-storage system fill or discharge capacity constraint, energy storage The constraint of system state-of-charge and system power Constraints of Equilibrium.

In step S4, antithetical phrase population carries out genetic manipulation, particularly as follows:

4.1 use binary tournament method to select two individualities from sub-population, use simulation binary system to intersect and multinomial The method of variation obtains two offspring individuals；

4.2 duplicon steps 4.1, until the scale of the sub-population of filial generation reaches the sub-population scale S set.

Sub-step 6.1 particularly as follows:

The priority of feasible individual is higher than infeasible individual；

Desired value according to feasible individual carries out the sequence of non-of inferior quality level to feasible individual；

Violating degree value according to comprehensive constraint and infeasible individual carries out prioritization, comprehensive constraint violates degree value more Little individuality, its first level is the highest；

Described comprehensive constraint violates degree valueWherein:

V_kRepresent that the comprehensive constraint of kth infeasible individual violates degree value；

w_RRepresent the constraint of climbing rate always violates weight, and its degree of concern retrained climbing rate according to policymaker takes Value；

w_r,nRepresent the climbing rate constraint violation weight in micro-source of the n-th tool climbing rate constraint, its according to policymaker to n-th The climbing rate in micro-source of individual tool climbing rate constraint is violated the degree of concern of situation and is carried out value, micro-source of all tool climbing rates constraint Climbing rate constraint violation weight should meet

v_k,r,nRepresent the climbing rate constraint violation degree in micro-source of the n-th tool climbing rate constraint under kth infeasible individual Value, its value is the micro-source of the n-th tool climbing rate constraint climbing rate constraint violation degree value sum under each time point；

w_lFor the infeasible individual violation weight to l constraints in addition to retraining except climbing rate, it is by policymaker's root Each constraints is violated the degree of concern decision of situation according to infeasible individual, and infeasible individual is to the institute in addition to retraining except climbing rate The violation weight of Prescribed Properties should meet

v_k,lRepresent the kth infeasible individual violation degree value to l constraints in addition to retraining except climbing rate, its Value is the kth infeasible individual violation degree value sum to l constraints under each time point；

N represents the quantity in micro-source of tool climbing rate constraint；

L represents the quantity of other constraintss in addition to climbing rate retrains.

As preferably, always violate weight w_RSpan be 0.3～0.7.

Compared with prior art, the present invention has the advantages that:

(1) improve sub-population scale reparation to measure, improve the multiformity of speed of searching optimization and feasible individual.

(2) use the mode that individual self-correcting combines with dominance relation sequence to process the constraint of micro-source climbing rate, solve Traditional constraints processes the problem that strategy cannot reduce climbing rate constraint violation degree.

(3) priori is required less, the workload of field adjustable personnel can be reduced, meet microgrid multiple-objection optimization and adjust Degree problem is to algorithm robustness and the demand of adaptivity.

Accompanying drawing explanation

Fig. 1 is the particular flow sheet of the inventive method.

Detailed description of the invention

The mode that the present invention uses individual self-correcting to combine with dominance relation sequence retrains to process micro-source climbing rate, and The subregion multi-objective Evolutionary Algorithm improved is used to solve the microgrid multi-objective optimization scheduling with multiple constraint.

Below in conjunction with accompanying drawing, the specific embodiment of the invention is elaborated.

Specifically comprising the following steps that of the inventive method

S1 sets up microgrid multi-objective optimization scheduling model, and described microgrid multi-objective optimization scheduling model includes Object function and constraints.

In this detailed description of the invention, object function includes the minimum object function of integrated cost and discharge amount of pollution minimum target Function.Wherein, integrated cost includes fuel cost, maintenance cost, depreciable cost etc.；Discharge amount of pollution includes carbon emission amount, nitrogen Compound discharge capacity and sulfide emission amount etc..Constraints includes exerting oneself micro-source, and bound retrains, micro-source climbing rate retrains, energy storage System is filled or discharge capacity constraint, the constraint of energy-storage system state-of-charge and system power Constraints of Equilibrium.Object function mentioned above And the expression formula of constraints is techniques well known, refers to existing document and obtain, do not repeat.

Before running multi-objective Evolutionary Algorithm, set population scale R, subproblem number K, K partition vector F₁、F₂、…… F_K, sub-population scale S, end condition and genetic operation operator.In this detailed description of the invention, R=100, K=10, K subregion to Object space is divided into K sub spaces, kth subspace Ω by amount_kIt is expressed as:

${\mathrm{\&Omega;}}_k=\{u\&Element;R_{+}^m|<u,F_k>\&le;<u,F_j>,j=1,...,K\}---\left(1\right)$

In formula (1),Represent that m ties up object space；<u,F_k> represent vector u and partition vector F_kBetween acute angle；<u, F_jRepresent vector u and partition vector F_jBetween acute angle.

When certain individuality is distributed in subspace Ω at the point that object space is corresponding_kIn, this individuality then position belongs to subspace Ω_kRight The sub-population answered.

S2 initializes the individuality of each sub-population according to the boundary constraint in constraints, and repaiies according to the constraint of micro-source climbing rate The individuality of multiple sub-population；Initialize evolutionary generation gen=0 simultaneously.

Here, the micro-source during boundary constraint refers to constraints exert oneself bound constraint and energy-storage system fill or discharge capacity about Bundle.

The described individuality repairing sub-population according to the constraint of micro-source climbing rate, farther includes:

2.1 obtain the sub-variable vector P=[P that in current sub-population, each micro-source is corresponding₁,P₂,…,P_M], wherein, P_mFor son Variable, represents that micro-source of m-th time point is actual exerts oneself, and when micro-source is energy storage device, actual the exerting oneself in micro-source uses micro-source Fill or discharge capacity；M=1,2 ... M, M express time point sum, in this detailed description of the invention, M takes 24.

2.2 make m=2, perform sub-step 2.3；

If 2.3 | P_m-P_m-1|≤P_r, do not repair；If P_m-P_m-1＞ P_r, by P_mRepair as P_m=min (P_max,P_m+P_r)；If P_m-1-P_m＞ P_r, by P_mRepair as P_m=max (P_min,P_m-P_r)；Wherein, P_mAnd P_m-1Represent the micro-of m, m-1 time point respectively Source is actual exerts oneself；P_r、P_max、P_minRepresent the maximum climbing rate in micro-source, EIAJ, minimum load respectively；

2.4 judge that m, whether equal to M, if being equal to, terminates repairing；Otherwise, make m=m+1, repeat sub-step 2.3.

In the present invention, sub-population is made up of individuality, and individuality is " solution ", and individuality includes that in sub-population, all micro-sources are when each Between point sub-variable.In this step, repair by individuality being decomposed into sub-variable vector.

S3 makes evolutionary generation gen=gen+1, then performs step S4.

S4 carries out genetic manipulation respectively to each sub-population, it is thus achieved that the sub-population of filial generation.

In this detailed description of the invention, antithetical phrase population carry out genetic manipulation particularly as follows:

4.1 use binary tournament method to select two individualities from sub-population, use simulation binary system to intersect and multinomial The method of variation obtains two offspring individuals；

4.2 duplicon steps 4.1, until the scale of the sub-population of filial generation reaches the sub-population scale S set.

S5 merges all of sub-population and the sub-population of filial generation generates mixed population, according to individual in its target in mixed population The distribution of subspace, space, is divided into new sub-population by mixed population.

The scale reparation of each new sub-population is the sub-population scale S set by S6.

Below in conjunction with embodiment, this step is described in detail.

In the present embodiment, tool constraints micro-source (the most controlled micro-source) have 5, including 2 miniature gas burners (MT), 2 Individual fuel cell (FC) and 1 accumulator (BT).Different controlled micro-sources have different constraints, the constraints of MT and FC Including micro-source climbing rate constraint and micro-source exert oneself bound constraint, the constraints of BT include energy-storage system fill or discharge capacity constraint Retraining with energy-storage system state-of-charge, system constraints is system power Constraints of Equilibrium.Therefore, the present embodiment has climbing Micro-source of rate constraint includes 2 MT and 2 FC, i.e. N=4；Constraints has 11, removes the climbing rate of 2 MT and 2 FC Constraints after constraint also remains 7, i.e. L=7.

This step farther includes:

In 6.1 pairs of mixed populations, individuality carries out prioritization.

Feasible individual priority is higher than infeasible individual.To feasible individual, according to the desired value of feasible individual to feasible Body carries out the sequence of non-of inferior quality level, and desired value calculates according to object function and obtains；The priority of infeasible individual is inferior to feasible individual, Comprehensive constraint according to infeasible individual is violated degree value and infeasible individual is carried out prioritization, and comprehensive constraint violates degree Being worth the least, individual priority is the highest.

Comprehensive constraint violates degree value V_kComputing formula as follows:

$V_k=w_R\&CenterDot;\underset{n=1}{\overset{N}{\&Sigma;}}(w_{r,n}\&CenterDot;v_{k,r,n})+\underset{l=1}{\overset{L}{\&Sigma;}}(w_l\&CenterDot;v_{k,l})---\left(2\right)$

In formula (2):

V_kRepresent that the comprehensive constraint of kth infeasible individual violates degree value；

w_RRepresent the constraint of climbing rate always violates weight；

w_r,nRepresent the climbing rate constraint violation weight in micro-source of the n-th tool climbing rate constraint；

v_k,r,nRepresent the climbing rate constraint violation degree in micro-source of the n-th tool climbing rate constraint under kth infeasible individual Value；

w_lRepresent the infeasible individual violation weight to l constraints in addition to retraining except climbing rate；

v_k,lRepresent the kth infeasible individual violation degree value to l constraints in addition to retraining except climbing rate；

N represents the quantity in micro-source of tool climbing rate constraint；

L represents the quantity of other constraintss in addition to climbing rate retrains.

Above-mentioned, w_RValue is carried out, if thinking preferably to solve climbing rate about according to the degree of concern that climbing rate is retrained by policymaker Bundle, w_RThen take higher value；w_RGeneral in 0.3～0.7 scope value, in the present embodiment, w_RTake 0.3.

Above-mentioned, w_r,nAccording to policymaker, the climbing rate in micro-source of the n-th tool climbing rate constraint is violated the degree of concern of situation Carrying out value, degree of concern is high, w_r,nThen take higher value, the climbing rate constraint violation weight in micro-source of all tool climbing rates constraint Should meet

Above-mentioned, v_k,r,nIt is used for reflecting the climbing rate constraint in micro-source of the n-th tool climbing rate constraint under kth infeasible individual Violation degree.V is provided below_k,r,nA kind of circular, micro-source climbing rate constraints is designated as | P_n,m-P_n,m-1| ≤P_n,rIf, | P_n,m-P_n,m-1| ＞ P_n,r, then climbing rate constraint violation degree value v of m-th time point_k,r,n,m=| P_n,m-P_n,m-1 |；If meeting the constraint of micro-source climbing rate, then climbing rate constraint violation degree value v of m-th time point_k,r,n_,mIt is 0.So, n-th The climbing rate constraint violation degree value in micro-source of individual tool climbing rate constraintWherein, P_n,mAnd P_n,m-1Table respectively Show m, m-1 time point n-th have climbing rate constraint the actual of micro-source exert oneself, P_n,rIt is the micro-of the n-th tool climbing rate constraint The maximum climbing rate in source.

Above-mentioned, w_lAccording to infeasible individual, each constraints in addition to retraining except climbing rate is violated the pass of situation by policymaker Note degree determines, if more paying close attention to the violation situation of certain constraints, corresponding violation weight then takes higher value；Vice versa. The violation weight of the institute's Prescribed Properties in addition to retraining except climbing rate should be met by infeasible individual

Above-mentioned, v_k,lIt is used for the violation reflecting kth infeasible individual to l constraints in addition to retraining except climbing rate Degree.In the present embodiment, the constraints in addition to climbing rate retrains includes exerting oneself bound constraints, 1 energy storage in 4 micro-sources System is filled or discharge capacity constraints, 1 energy-storage system state-of-charge constraints and 1 system power equilibrium constraint, I.e. L=7.

V is provided below_k,lThe circular of value.

Kth infeasible individual bound constraints of exerting oneself micro-source violates being calculated as follows of degree value:

Micro-source bound constraints of exerting oneself is designated as P_n1,min≤P_n1,m≤P_n1,maxIf, P_n1,m＞ P_n1,max, then the m-th time Under Dian, exert oneself violation degree value v of bound constraints in micro-source_k,l,m=| P_n1,m-P_n1,max|；If P_n1,m＜ P_n1,min, then m Under individual time point, exert oneself violation degree value v of bound constraints in micro-source_k,l,m=| P_n1,m-P_n1,min|；If meeting micro-source to exert oneself Bound constraints, then under m-th time point, exert oneself violation degree value v of bound constraints in micro-source_k,l,mIt is 0.So, Exerted oneself the violation degree value of bound constraints in micro-source by kth infeasible individualWherein, P_n1,maxWith P_n1,minRepresent that the n-th 1 have micro-source and exert oneself the exert oneself upper limit and lower limit of exerting oneself, the P in micro-source of bound constraint respectively_n1,mRepresent the N1 have micro-source exert oneself bound constraint the actual of micro-source exert oneself.In the present embodiment, have micro-source exert oneself bound constraint micro- Source quantity is 4, then micro-source bound constraints quantity of exerting oneself also is 4.

Energy-storage system is filled by kth infeasible individual or discharge capacity constraints violates being calculated as follows of degree value:

Energy-storage system fills or discharge capacity constraints is designated as P_-,n2≤P_+/-,n2,m≤P_+,n2If, P_+/-,n2,m＞ P_+,n2, then m Under individual time point, energy-storage system fills or violation degree value v of discharge capacity constraints_k,l,m=| P_+/-,n2,m-P_+,n2|；If P_+/-,n2,m ＜ P_-,n2, then under m-th time point, energy-storage system fills or violation degree value v of discharge capacity constraints_k,l,m=| P_+/-,n2,m- P_-,n2|；If meeting energy-storage system to fill or discharge capacity constraints, then under m-th time point, energy-storage system fills or discharge capacity constraint Violation degree value v in micro-source of condition_k,l,mIt is 0.So, energy-storage system is filled or discharge capacity constraint bar by kth infeasible individual The violation degree value of partWherein, P_-,n2And P_+,n2Represent respectively the n-th 2 tool energy-storage systems fill or discharge capacity about The maximum pd quantity in micro-source of bundle and maximum charge amount, P_+/-,n2,mRepresent that the n-th 2 tool energy-storage systems of m-th time point fill or put The actual discharge amount in micro-source of Constraint or actual charge volume.In the present embodiment, tool energy-storage system fill or discharge capacity constraint Micro-source quantity is 1, then energy-storage system fills or discharge capacity constraints quantity is also 1.

Energy-storage system state-of-charge constraints is violated being calculated as follows of degree value by kth infeasible individual:

Energy-storage system state-of-charge is constrained to SOC_n3,min≤SOC_n3,m≤SOC_n3,max,If SOC_n3,m＞ SOC_n3,max, then energy-storage system state-of-charge constraint under m-th time point Violation degree value v of condition_k,l,m=| SOC_n3,m-SOC_n3,max|；If SOC_n3,m＜ SOC_n3,min, then energy storage under m-th time point Violation degree value v of system state-of-charge constraints_k,l,m=| SOC_n3,m-SOC_n3,min|；If meeting energy-storage system state-of-charge Constraint, then violation degree value v of energy-storage system state-of-charge constraints under m-th time point_k,l,mIt is 0.So, kth is not The feasible individual violation degree value to energy-storage system state-of-charge constraintsWherein, SOC_n3,m、SOC_int、 SOC_n3,maxAnd SOC_n3,minRepresent micro-source reality at m-th time point of the n-th 3 tool energy-storage system state-of-charge constraints respectively State-of-charge, initial state-of-charge, maximum state-of-charge and minimum state-of-charge；P_+/-,n3,jRepresent jth time point the n-th 3 The actual discharge amount in micro-source of tool energy-storage system state-of-charge constraint or actual charge volume.In the present embodiment, has energy-storage system lotus Micro-source quantity of electricity condition constraint is 1, then energy-storage system state-of-charge constraints quantity is also 1.

System power equilibrium constraint is violated being calculated as follows of degree value by kth infeasible individual:

System power equilibrium constraint is designated asRepresent all controlled micro-sources The actual sum of exerting oneself of m time point, N' represents all controlled micro-sources number, P_n',mRepresent n-th ' individual controlled micro-source m-th time The actual of point is exerted oneself.IfThen under m-th time point system power equilibrium constraint disobey Return angle valueIf meeting system power equilibrium constraint, then under m-th time point Violation degree value v of system power equilibrium constraint_k,l,mIt is 0.So, system power is balanced about by kth infeasible individual The violation degree value of bundle conditionWherein, N_s、P_L,mRepresent energy-storage system quantity and m-th time point system respectively The workload demand of system, P_+/-,n4,mRepresent the actual discharge amount of the n-th 4 energy-storage systems of m-th time point or actual charge volume.This reality Execute in example, N_s=1, l take 7.

6.2 pairs of scales are more than the new sub-population of the sub-population scale S set, by new sub-population scale and the sub-population of setting The difference of scale is designated as q, rejects q the individuality that priority is minimum from new sub-population.

6.3 pairs of scales are less than the new sub-population of the sub-population scale S set, by new sub-population scale and the sub-population of setting The difference of scale is designated as q, and from subspace, new sub-population place, choosing q the shortest with the partition vector of new sub-population distance can Row individuality adds new sub-population；If the feasible individual lazy weight q outside subspace, then the infeasible individual from subspace is pressed Priority selects.In detailed description of the invention, distance uses Euclidean distance in object space.

6.4 pairs of scales are equal to the new sub-population of the sub-population scale S set, and do not process.

S7 retains all non-bad individualities in new sub-population.

S8 judges whether to meet the condition of convergence, if meeting, performs step S9；Otherwise, execution step S3 is returned.

The known technology being set to iterative algorithm of the condition of convergence, typically can be set to evolutionary generation and reach maximum evolutionary generation Or the difference of adjacent iteration gained solution value is less than preset value.In this detailed description of the invention, the condition of convergence is set to evolutionary generation and reaches To maximum evolutionary generation.

S9 retains and exports all feasible individuals.

Claims (6)

1., based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes, it is characterized in that, including step:

S1 sets up the microgrid multi-objective optimization scheduling model being made up of object function and constraints；Described constraints Retrain including micro-source climbing rate；

S2 initial beggar population at individual, and repair sub-population at individual according to the constraint of micro-source climbing rate；Initialize evolutionary generation gen= 0；

Described repairs sub-population at individual according to the constraint of micro-source climbing rate, particularly as follows:

1. the sub-variable vector P=[P that each micro-source in current sub-population that obtains is corresponding₁,P₂,...,P_M], make m=2；

If 2. | P_m-P_m-1|≤P_r, do not repair；If P_m-P_m-1＞ P_r, by P_mRepair as P_m=min (P_max,P_m+P_r)；If P_m-1-P_m ＞ P_r, by P_mRepair as P_m=max (P_min,P_m-P_r)；

3. judge that m, whether equal to M, if being equal to, terminates；Otherwise, make m=m+1, repeat sub-step 2.；

Wherein, P_mAnd P_m-1Represent that micro-source of m, m-1 time point is actual respectively to exert oneself, when micro-source is energy storage device, micro-source Actual exerting oneself uses filling or discharge capacity of micro-source；M=1,2 ... M, M express time point sum；P_r、P_max、P_minRepresent micro-respectively The maximum climbing rate in source, EIAJ, minimum load；

S3 makes evolutionary generation gen=gen+1, performs step S4；

S4 carries out genetic manipulation respectively and obtains the sub-population of filial generation each sub-population, and makes the sub-population scale of filial generation reach the son set Population scale S；

S5 merges all of sub-population and the sub-population of filial generation generates mixed population, according to individual at its object space in mixed population The distribution of subspace, is divided into new sub-population by mixed population；

The scale reparation of each new sub-population is the sub-population scale S set by S6, and this step farther includes:

In 6.1 pairs of mixed populations, individuality carries out prioritization；

6.2 pairs of scales are more than the new sub-population of S, reject q the individuality that priority is minimum from new sub-population；

6.3 pairs of scales are less than the new sub-population of S, from subspace, new sub-population place, choose and the partition vector of new sub-population Q the feasible individual new sub-population of addition that distance is the shortest；If the feasible individual lazy weight q outside subspace, the most according to priority from Selecting infeasible individual to add new sub-population outside subspace, the individual sum adding new sub-population is q；

6.4 pairs of scales are equal to the new sub-population of S, do not process；

Described q represents the difference of the sub-population scale of new sub-population scale and setting；

S7 retains all non-bad individualities in new sub-population；

S8 judges whether to meet the condition of convergence, if meeting, retaining and exporting all feasible individuals；Otherwise, step S3 is returned.

2., as claimed in claim 1 based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes, it is characterized in that:

Described object function includes the minimum object function of integrated cost and discharge amount of pollution minimum target function.

3., as claimed in claim 1 based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes, it is characterized in that:

Described constraints also includes exerting oneself micro-source, and bound retrains, energy-storage system fills or discharge capacity constraint, energy-storage system lotus Electricity condition constraint and system power Constraints of Equilibrium.

4., as claimed in claim 1 based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes, it is characterized in that:

In step S4, antithetical phrase population carries out genetic manipulation, particularly as follows:

4.1 use binary tournament method to select two individualities from sub-population, use simulation binary system to intersect and multinomial makes a variation Method obtain two offspring individuals；

4.2 duplicon steps 4.1, until the scale of the sub-population of filial generation reaches the sub-population scale S set.

5., as claimed in claim 1 based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes, it is characterized in that:

Sub-step 6.1 particularly as follows:

The priority of feasible individual is higher than infeasible individual；

Desired value according to feasible individual carries out the sequence of non-of inferior quality level to feasible individual；

Violating degree value according to comprehensive constraint and infeasible individual carries out prioritization, it is the least that comprehensive constraint violates degree value Individuality, its first level is the highest；

Described comprehensive constraint violates degree valueWherein:

V_kRepresent that the comprehensive constraint of kth infeasible individual violates degree value；

w_RRepresent the constraint of climbing rate always violates weight, and its degree of concern retrained climbing rate according to policymaker carries out value；

w_r,nRepresenting the climbing rate constraint violation weight in micro-source of the n-th tool climbing rate constraint, it has n-th according to policymaker The climbing rate in micro-source of climbing rate constraint is violated the degree of concern of situation and is carried out value, climbing of micro-source of all tool climbing rates constraint Ratio of slope constraint violation weight should meet

v_k,r,nRepresent the climbing rate constraint violation degree value in micro-source of the n-th tool climbing rate constraint under kth infeasible individual, its Value is the micro-source of the n-th tool climbing rate constraint climbing rate constraint violation degree value sum under each time point；

w_lFor infeasible individual to the violation weight of l constraints in addition to retraining except climbing rate, its by policymaker according to can not The individual degree of concern that each constraints is violated situation of row determines, infeasible individual is to the institute's Constrained in addition to retraining except climbing rate The violation weight of condition should meet

v_k,lRepresenting the kth infeasible individual violation degree value to l constraints in addition to retraining except climbing rate, its value is Kth infeasible individual violation degree value sum to l constraints under each time point；

N represents the quantity in micro-source of tool climbing rate constraint；

L represents the quantity of other constraintss in addition to climbing rate retrains.

6., as claimed in claim 5 based on improving the microgrid Optimization Scheduling that subregion multi-target evolution optimizes, it is characterized in that:

Described total violation weight w_RSpan be 0.3～0.7.