CN104037757A - Brainstorming-based thermal power plant economic environment scheduling method - Google Patents

Abstract

The invention discloses a brainstorming-based thermal power plant economic environment scheduling method. The method includes the following steps of: determining mathematical models of thermal power plant environmental economic scheduling problems; step 2, obtaining various parameters in the models; step 3, obtaining an initial feasible scheduling set of the thermal power plant environmental economic scheduling problems; step 4, evaluating the feasible scheduling set; step 5, updating the feasible scheduling set; step 6, judging whether entire feasible scheduling is updated completely; step 7, updating a non-inferior scheduling solution set in an external archiving set; step 8, performing iterative output of a final feasible scheduling scheme, and the method terminates. According to the brainstorming-based thermal power plant economic environment scheduling method of the invention, a thermal power plant is adopted as a power supply system, and solutions to thermal power plant economic scheduling problems which are required by comprehensive environmental protection can be realized through collection and analysis of data of a power system and the thermal power plant and a brainstorming optimization algorithm based clustering and variation ideas.

Description

A kind of thermal power plant's economic environment dispatching method based on brainstorming

Technical field

The invention belongs to field of intelligent control technology, relate to a kind of thermal power plant's economic environment dispatching method based on brainstorming.

Background technology

Thermal power generation is the main force of current supply of electric power; due to pernicious gases such as a large amount of oxysulfide of power plant emission, nitrogen oxide and carbon dioxide; not only directly polluted atmospheric environment; and greenhouse effect have been caused; therefore; consider the research of the power system environment Economic Dispatch Problem of environmental protection and economic benefit, not only there is important theory significance, and be the most real selection of the energy and the electric power strategy of sustainable development.If consideration discharge amount of pollution, single-object problem originally will become multi-objective optimization question, has not only increased the complexity of problem, has brought difficulty and challenge also to the enforcement of scheduling.Because in power system environment Economic Dispatch Problem, between each target, be conflicting, therefore formulating rational operation plan is the emphasis of studying at present.

Brainstorming is that nineteen thirty-nine U.S.'s creation scholar A.F Ao Siben proposes first, claim again intelligence stimulation method, be the method that produces new concept or excite a kind of excitability thinking of innovative hypothesis, everybody produces new concept around specific interest worlds, and this sight is called brainstorming.Utilize the feature of free in brainstorm meeting, unfettered and positive optimism, within 2011, the beautiful time teacher of history is at (the The Second International Conference on Swarm Intelligence of colony intelligence international conference for the second time, be called for short ICSI11) middle a kind of new colony intelligence optimized algorithm---the brainstorming optimized algorithm (Brain Storm Optimization Algorithm is called for short BSO) that proposes.

Summary of the invention

The object of this invention is to provide a kind of thermal power plant's economic environment dispatching method based on brainstorming, solve existing dispatching method and between the economic benefit of electrical network and environmental benefit, be difficult to regulate, be difficult for obtaining the problem of optimum efficiency.

The technical solution adopted in the present invention is: a kind of thermal power plant's economic environment dispatching method based on brainstorming, according to following steps, implement:

Step 1, determine the Mathematical Modeling of thermal power plant's environmental economy scheduling problem

The Mathematical Modeling that environmental economy scheduling problem is set is:

$\min [\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}{F}_i\left(P_i\right),\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}{E}_i\left(P_i\right)],---\left(1\right)$

Wherein, N _gfor generator sum in system, i represents i platform generator, i=1, and 2 ..., N _g; P _ithe active power that represents i platform generator; F _i(P _i) representing that a certain moment thermal power plant's fuel used to generate electricity total burn-off is fuel total cost, calculation expression is:

$F_i\left(P_i\right)=a_i+b_i{P}_i+c_i{p}_i^2,---\left(2\right)$

A _i, b _i, c _ibe system parameters, represent respectively constant term, Monomial coefficient and the quadratic term coefficient of i platform generating set consumption characteristic, be known parameters in system;

E _i(P _i) representing the discharge amount of pollution of i platform generator, its calculation expression is:

E _i(P _i)＝α _i+β _iP _i+γ _iP _i ²+ξ _iexp(λ _iP _i)， (3)

α wherein _i, β _i, γ _i, ξ _i, λ _ibe system parameters, α _i, β _i, γ _ithe constant term, Monomial coefficient and the quadratic term coefficient that represent respectively i platform generating set disposal of pollutants flow characteristic, ξ _i, λ _ithe relevant parameter that represents exponential term, in environmental economy optimizing scheduling process, suffered constraints has:

Constraints 1: the inequality constraints of unit generation capacity:

P _i ^min＜P _i＜P _i ^max， (4)

Wherein, P _i ^min, P _i ^maxbe respectively minimum and the output of maximum active power of i platform thermoelectric generator;

Constraints 2: balance equality constraint, each generating set generated output sum of system should equal the total demand power of load and via net loss sum:

$P_D+P_{\mathrm{loss}}-\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}{P}_i=0,---\left(5\right)$

In formula (5), for a certain moment thermoelectricity gross power; P _dfor this moment system load demand; P _lossbe expressed as active power loss in t moment electrical network, obtaining of active power loss adopts B Y-factor method Y, and its formula is:

$P_{\mathrm{loss}}=\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}\underset{j=1}{\overset{N_G}{\mathrm{\Σ}}}{P}_i{B}_{\mathrm{ij}}{P}_j+\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}{B}_{0i}{P}_i+B_{00},---\left(6\right)$

B wherein _ij, B _0i, B ₀₀for B coefficient;

Step 2, obtain all kinds of parameters in model;

The initial feasible schedule set of step 3, acquisition environmental economy scheduling problem

In above-mentioned model, need to determine that decision variable is N _gthe active power of individual unit optimal value, produce feasible schedule set;

Step 4, to feasible schedule, set is evaluated

N the feasible schedule set that step 3 is produced be substitution target function type (2) and formula (3) respectively, respectively each feasible schedule is carried out the evaluation of environment and economy benefit, due to all corresponding two target functions of each feasible schedule, therefore need to carry out noninferior solution sequence to N feasible schedule, the feasible schedule sequence of not arranging is mutually kept in an outside set, is referred to as outside filing collection;

Step 5, feasible schedule is upgraded

Before upgrading, iterations initial value t=0 need to be set, and maximum iteration time T _max;

First random chooses m different feasible schedule as m Ge Leilei center, according to the active power of other all feasible schedule, to the Euclidean distance at each class center, N feasible schedule carried out to cluster, for simulating the forming process of brains storm processes thinking; By the class definition that contains noninferior solution, being elite's class, is general category without the class definition of noninferior solution, and on the basis of the fresh information obtaining in cluster, by selection, operates and mutation operation carries out iteration renewal to the active power of each feasible schedule generating set;

Step 6, judge whether whole feasible schedule has been upgraded

If i=N, shows that N feasible schedule upgraded, enter step 7; Otherwise return to step 5;

Step 7, the non-bad scheduling disaggregation that outside filing is concentrated are upgraded

The non-bad feasible schedule that Comparative economic benefit and discharge amount of pollution obtain is each time stored in to outside filing to be concentrated;

Step 8, the iteration of carrying out are exported final feasible schedule scheme

Judge whether iterations t reaches maximum of T _max, if not, iterations t=t+1 is set, forward step 4 to; If so, export the feasible schedule in current Noninferior Solution Set.

The invention has the beneficial effects as follows; using thermal power plant as electric power system; the collection analysis of utilization to electric power system and thermal power plant's data, in conjunction with the brainstorming optimized algorithm based on cluster and variation thought, realizes solving thermal power plant's Economic Dispatch Problem of integrated environment protection demand.By adjusting exerting oneself of each generating set of thermal power plant, meeting under the conditions such as line balancing and each unit output constraint, the energy output of each fired power generating unit of reasonable arrangement day part, for operation and the control of electric power system provides multiple feasible program, policymaker can therefrom determine final scheme according to different requirements, makes discharge amount of pollution and economic benefit in this control cycle reach comprehensive optimum.

The advantage of the inventive method specifically comprises:

The first, aspect real-time, there is good global convergence performance and convergence rate faster, realize simply, use it for the power system environment economic load distribution problem that solves, can effectively realize the scheduling real-time of electrical network.

Second, aspect practicality, under Electricity Market, power-management centre is according to indexs such as the coal consumption amount characteristic of each thermal power plant, dusty gas emission performance and quotations, consider each power plant combination of mutually coordinating, the unified overall planning energy output of considering under the energy, environmental protection, market benefits of different parties.Zuo Dui power plant generating Fang Eryan, can cost-savingly increase the benefit like this; To public users Fang Eryan, except obtaining satisfied electric weight, can also protection of the environment, energy savings, reach the effect that many factors considers.Than merely, according to each power plant, surf the Net at a competitive price and more meet the strategy of sustainable development, also than the Optimized Operation merely making overall arrangements with energy savings, more meet the requirement of the electric power system marketization.

Embodiment

Below in conjunction with embodiment, the present invention is described in detail.

Thermal power plant's economic environment dispatching method based on brainstorming of the present invention, according to following steps, implement:

Step 1, determine the Mathematical Modeling of thermal power plant's environmental economy scheduling problem

This thermal power plant's environmental economy scheduling problem is obvious two objective optimization problems, and the Mathematical Modeling that environmental economy scheduling problem is set is:

$\min [\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}{F}_i\left(P_i\right),\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}{E}_i\left(P_i\right)],---\left(1\right)$

Wherein, N _gfor generator sum in system, i represents i platform generator, i=1, and 2 ..., N _g; P _ithe active power that represents i platform generator; F _i(P _i) representing that a certain moment thermal power plant's fuel used to generate electricity total burn-off is fuel total cost, calculation expression is:

$F_i\left(P_i\right)=a_i+b_i{p}_i+c_i{p}_i^2,---\left(2\right)$

F _i(P _i) also embodied the fuel consumption characteristic of i platform generator simultaneously, the expense that has generating consume fuel to produce, and take market fuel used to generate electricity unit price as basis, a _i, b _i, c _ibe system parameters, represent respectively constant term, Monomial coefficient and the quadratic term coefficient of i platform generating set consumption characteristic, be known parameters in system;

E _i(P _i) representing the discharge amount of pollution of i platform generator, its calculation expression is:

E _i(P _i)＝α _i+β _iP _i+γ _iP _i ²+ξ _iexp(λ _iP _i)， (3)

α wherein _i, β _i, γ _i, ξ _i, λ _ibe system parameters, α _i, β _i, γ _ithe constant term, Monomial coefficient and the quadratic term coefficient that represent respectively i platform generating set disposal of pollutants flow characteristic, ξ _i, λ _ithe relevant parameter that represents exponential term is normal value for concrete system.

In environmental economy optimizing scheduling process, suffered constraints has:

Constraints 1: the inequality constraints of unit generation capacity:

P _i ^min＜P _i＜P _i ^max， (4)

Wherein, P _i ^min, P _i ^maxbe respectively minimum and the output of maximum active power of i platform thermoelectric generator;

Constraints 2: balance equality constraint, each generating set generated output sum of system should equal the total demand power of load and via net loss sum:

$P_D+P_{\mathrm{loss}}-\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}{P}_i=0,---\left(5\right)$

In formula (5), for a certain moment thermoelectricity gross power; P _dfor this moment system load demand; P _lossbe expressed as active power loss in t moment electrical network, obtaining of active power loss adopts B Y-factor method Y, and its formula is:

$P_{\mathrm{loss}}=\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}\underset{j=1}{\overset{N_G}{\mathrm{\Σ}}}{P}_i{B}_{\mathrm{ij}}{P}_j+\underset{i=1}{\overset{N_G}{\mathrm{\Σ}}}{B}_{0i}{P}_i+B_{00},---\left(6\right)$

B wherein _ij, B _0i, B ₀₀being B coefficient, obtaining respectively according to the character of unit, is known parameters.

Step 2, obtain all kinds of parameters in model

According to grid load curve, determine the load valley of electrical network and load peak period, determining the power demand P of day part sharing of load _i ^min, P _i ^max;

From grid dispatching center statistics, obtain the data of current time system, mainly comprise system total load P _d, total network loss P _lossparameter value B _ij, B _0i, B ₀₀;

The discharge capacity of the pernicious gases such as coal consumption amount, flue dust and carbon dioxide when the running performance parameters of unit mainly comprises operation, according to the supplemental characteristic α that soot emissions are fixed a price, reduction of discharging is fixed a price, CO2 emission exercise price obtains discharge amount of pollution _i, β _i, γ _i, ξ _i, λ _i;

Coal consumption price in producing according to electric power system assistant service price, unit of electrical energy, cost of electricity-generating while obtaining raw coal price, diesel-fuel price thermal power unit operation, obtain the parameter a of fuel total cost _i, b _i, c _i.

The initial feasible schedule set of step 3, acquisition environmental economy scheduling problem

In above-mentioned model, need to determine that decision variable is N _gthe active power of individual unit optimal value.

The present invention is based on existing colony intelligence optimized algorithm and brainstorming optimized algorithm, create a kind of new colony intelligence-brainstorming optimized algorithm, by cluster and variation two generic operations, realize the continuous renewal of feasible schedule, operation simply, easily realizes, in solving the multi-objective problems such as electric power system, effect is remarkable.

In this step, mainly produce feasible schedule set, concrete producing method is as follows:

Step1: to front N _g-1individual generating set, in meeting the scope of maximum generation active power and minimum generating active power, produces the front N of each unit at random _g-1the active-power P of individual generating set _i(i=1,2 ..., N _g-1);

Step2: according to equality constraint calculate N _gthe active power of individual unit;

Step3: computing network loss P _loss;

Step4: consider via net loss P _loss, according to formula calculate the active-power P of the last one dimension of each unit that meets equality constraint _d;

Step5: judge last N _gwhether individual unit meets the condition of capacity-constrained, if do not meet, re-executes step1 to step4 step, otherwise retains the feasible schedule producing;

According to same process, produce N feasible schedule set.

Step 4, to feasible schedule, set is evaluated

N the feasible schedule set that step 3 is produced be substitution target function type (2) and formula (3) respectively, respectively each feasible schedule is carried out the evaluation of environment and economy benefit, due to all corresponding two target functions of each feasible schedule, therefore need to carry out noninferior solution sequence to N feasible schedule, the feasible schedule sequence of not arranging is mutually kept in an outside set, is referred to as outside filing collection;

Step 5, feasible schedule is upgraded

From step 5, start to adopt the optimum ideals of brainstorming to upgrade feasible schedule.Before upgrading, iterations initial value t=0 need to be set, and maximum iteration time T _max;

First random chooses m different feasible schedule as m Ge Leilei center, according to the active power of other all feasible schedule, to the Euclidean distance at each class center, N feasible schedule carried out to cluster, for simulating the forming process of brains storm processes thinking; By the class definition that contains noninferior solution, it is elite's class, class definition without noninferior solution is general category, and on the basis of the fresh information obtaining in cluster, by selection, operated with mutation operation the active power of each feasible schedule generating set is carried out to iteration renewal, this update mode is the most important innovative point of the present invention.Specific implementation process is as follows:

5.1) select operation

In the t time iteration, for any one current feasible schedule, according to the distinctive selection mechanism of brainstorming algorithm below, select the feasible schedule that will upgrade.

Specifically, for i parent feasible schedule, produce the random number rand1 between a random value 0-1,

If rand1 is less than probability P 1, select feasible schedule in current class to upgrade; Specifically, produce a random number rand2, if rand2 is less than probability P 2, select feasible schedule in Yi Gelei center or class as upgating object,

When the random number producing is less than P3, select the feasible schedule in elite's class center or class; Otherwise select the feasible schedule in general category center or class;

If rand2 is more than or equal to probability P 2, select at random two classes to produce the active power of new feasible schedule; Specifically, produce a random value, if random number is less than probability P 4, the cluster centre in two classes that choose is carried out to linear combination; Otherwise the active power of random two feasible schedule selecting is carried out linear combination from two classes selecting;

Otherwise, with the probability of 1-P1, from filing, concentrate the active power of selecting the feasible schedule that will make a variation.

Above-mentioned P1, P2, P3, P4 are the probability parameter carrying in brainstorming algorithm, select really to fix a number between 0-1;

5.2) mutation operation

5.2.1) t iteration history before j feasible schedule selecting operation to obtain is dispatched in from 1 to N _g-1 random active power conduct of selecting i unit the active power of i unit the t+1 time in j feasible schedule iterative formula as follows:

$P_{{\mathrm{idx}\_\mathrm{popu}}_{\mathrm{ji}}}(t+1)=\left\{\begin{array}{cc}\mathrm{random}({P_i}^{\min },{P_i}^{\max })& \mathrm{rand}\left(\mathrm{0,1}\right)<0.05\\ P_{{\mathrm{idx}\_\mathrm{popu}}_{\mathrm{ji}}}\left(t\right)+\mathrm{rand}\left(\mathrm{0,1}\right)\&times;(P_{{\mathrm{best}1}_{\mathrm{ji}}}\left(t\right)-P_{{\mathrm{best}2}_{\mathrm{ji}}}\left(t\right))& \mathrm{otherwise}\end{array}\right.,---\left(7\right)$

In formula (7), with be the active power of j unit of i concentrated feasible schedule of the t time iteration filing, if file concentrated feasible schedule number, be less than 2, in the active power when two feasible schedule of the random selection of former generation; According to formula (7), produce the front N of new feasible schedule _g-1dimension, and check it whether to cross the border, be if so, bound; Otherwise the last one dimension by definite each the feasible schedule active power of power-balance constraint, guarantees that every one dimension of each feasible schedule active power is in its confining spectrum;

5.2.2) calculate the target function value of the new feasible schedule producing of i filial generation, filial generation and parent are carried out to non-ratio of less inequality, according to dominance relation, retain better feasible schedule;

By above iterative formula, the generating active power of the rate of change of each generating set and generating set will be constantly updated, and be convenient to like this find out N feasible schedule set that more meets the generating set generated output of target function under integrated environment factor.

The innovative point of this step is mainly: by all feasible schedule are carried out to cluster operation, according to having or not noninferior solution to be divided into elite's class and non-elite's class, introduce the corresponding generating set active power of more excellent target that filing is concentrated in the renewal process of feasible schedule with the introducing of this innovative point makes the flight of feasible schedule have more directivity, makes dispatching method have stronger exploring ability, has strengthened its optimization ability.

Step 6, judge whether whole feasible schedule has been upgraded

If i=N, shows that N feasible schedule upgraded, enter step 7; Otherwise return to step 5;

Step 7, the non-bad scheduling disaggregation that outside filing is concentrated are upgraded

Utilization to the extraction of relevant information in electric power system, analyze the method in conjunction with the comprehensive brainstorming algorithm of Clustering, realize the optimization of scheduling sequence; Because target is that economic benefit and disposal of pollutants are simultaneously optimum, therefore the non-bad feasible schedule that Comparative economic benefit and discharge amount of pollution obtain is each time stored in to outside filing and concentrates,

Outside the renewal of outside filing collection except the non-domination scheduling in population, also adopt crowding distance method to safeguard, concrete grammar is: the active power of the non-domination feasible schedule in population is put into outside filing one by one and concentrate, if the active power that the active power of this feasible schedule is filed concentrated feasible schedule by outside is arranged, the active power of this feasible schedule is deleted from filing to concentrate, otherwise the active power of this feasible schedule adds filing collection; If file the active power number of concentrated feasible schedule, be less than heap(ed) capacity, do not carry out deletion action, otherwise calculate the crowding distance that the active power of all feasible schedule is concentrated in current filing, the active power of deleting that feasible schedule of crowding distance minimum makes to file concentrated feasible schedule and remains on the number that is less than or equal to heap(ed) capacity.

The innovative point of this step is mainly: the method is different from all non-bad scheduling solution that NSGA-II etc. concentrates all non-domination feasible schedule producing in population and outside filing and sorts from big to small by crowding distance, select the feasible schedule that crowding distance is large to enter the next generation, the distribution that is conducive to feasible schedule solution is more even again.

Step 8, the iteration of carrying out are exported final feasible schedule scheme

Judge whether iterations t reaches maximum of T _max, if not, iterations t=t+1 is set, forward step 4 to; If so, export the feasible schedule in current Noninferior Solution Set.

Embodiment

Yi Mou thermal power plant 6 unit power system environment Economic Dispatch Problems are example, and the parameter setting relating in the implementation process of the inventive method is described.

Set the parameters in step 1 and step 2: the Mathematical Modeling of this environmental economy scheduling problem is suc as formula shown in (1)-Shi (7), the constant term of the i platform generating set fuel consumption characteristic in its Chinese style (2), once and quadratic term coefficient matrix be the fixed value of system, be respectively:

a _i＝[10 10 20 10 20 10]

b _i＝[200 150 180 100 180 150]

c _i＝[100 120 40 60 40 100]

The constant term of the i platform generating set disposal of pollutants flow characteristic in formula (3), once and the coefficient correlation of quadratic term coefficient matrix exponential term by the fixed value of system, be respectively:

α _i＝[4.091 2.543 4.258 5.326 4.258 6.131]

β _i＝[-5.554 -6.047 -5.094 -3.55 -5.094 -5.555]

γ _i＝[6.49 5.638 4.586 3.38 4.586 5.151]

ξ _i＝[0.0002 0.005 0.000001 0.002 0.000001 0.00001]

λ _i＝[2.857 3.333 8 2 8 6.667]

The total load P that generator in formula (5) is born _d=2.834MW,

The B coefficient that calculates network loss in formula (6) is existing fixed value, specific as follows:

$B_{\mathrm{ij}}=\left[\begin{array}{cccccc}0.0218& 0.0107& -0.00036& -0.0011& 0.00055& 0.0033\\ 0.0107& 0.01704& -0.0001& -0.00179& 0.00026& 0.0028\\ -0.0004& -0.0002& 0.02459& -0.01328& -0.0118& -0.0079\\ -0.0011& -0.00179& -0.01328& 0.0265& 0.0098& 0.0045\\ 0.00055& 0.00026& -0.0118& 0.0098& 0.0216& -0.0001\\ 0.0033& 0.0028& -0.00792& 0.0045& -0.00012& 0.02978\end{array}\right];$

B _0i＝[0.010731 1.7704 -4.0645 3.8453 1.3832 5.5503]×e-03；

B ₀₀＝0.0014；

In step 3, unit N _g=6 is system fixed value, and initial feasible schedule is counted N and rule of thumb selected 100, and maximum iteration time is rule of thumb selected T _max=1000;

Four probability parameters in step 5 according to experience respectively value be: P1=0.8, P2=0.4P3=0.2P4=0.2.

Through optimizing and the iterative process based on brainstorming thinking, the inventive method just can access the sharing of load scheme of each unit in the short period of time, this searching process without any restriction, does not rely on Mathematical Modeling to the target function of system and constraints, has good versatility.Compare with other algorithms, searching process of the present invention is realized simple, and easily operation has good global convergence performance and convergence rate faster.

In a word; dispatching method of the present invention; double factor for integrated environment protection and economic benefit; propose to utilize the cluster multiple target brainstorming optimized algorithm based on Knowledge Extraction; the constraint that has scheduling with peak and the low ebb of electricity consumption; the service requirement that can more conscientiously reflect system, realizes the environmental economy scheduling to thermal power plant.

Claims (5)

1. the thermal power plant's economic environment dispatching method based on brainstorming, its feature is, according to following steps, implements:

Step 1, determine the Mathematical Modeling of thermal power plant's environmental economy scheduling problem

The Mathematical Modeling that environmental economy scheduling problem is set is:

$\min [\underset{i=1}{\overset{N_G}{\mathrm{\&Sigma;}}}{F}_i\left(P_i\right),\underset{i=1}{\overset{N_G}{\mathrm{\&Sigma;}}}{E}_i\left(P_i\right)],---\left(1\right)$

Wherein, N _gfor generator sum in system, i represents i platform generator, i=1, and 2 ..., N _g; P _ithe active power that represents i platform generator; F _i(P _i) representing that a certain moment thermal power plant's fuel used to generate electricity total burn-off is fuel total cost, calculation expression is:

$F_i\left(P_i\right)=a_i+b_i{P}_i+c_i{p}_i^2,---\left(2\right)$

A _i, b _i, c _ibe system parameters, represent respectively constant term, Monomial coefficient and the quadratic term coefficient of i platform generating set consumption characteristic, be known parameters in system;

E _i(P _i) representing the discharge amount of pollution of i platform generator, its calculation expression is:

E _i(P _i)＝α _i+β _iP _i+γ _iP _i ²+ξ _iexp(λ _iP _i)， (3)

α wherein _i, β _i, γ _i, ξ _i, λ _ibe system parameters, α _i, β _i, γ _ithe constant term, Monomial coefficient and the quadratic term coefficient that represent respectively i platform generating set disposal of pollutants flow characteristic, ξ _i, λ _ithe relevant parameter that represents exponential term, in environmental economy optimizing scheduling process, suffered constraints has:

Constraints 1: the inequality constraints of unit generation capacity:

P _i ^min＜P _i＜P _i ^max， (4)

Wherein, P _i ^min, P _i ^maxbe respectively minimum and the output of maximum active power of i platform thermoelectric generator;

Constraints 2: balance equality constraint, each generating set generated output sum of system should equal the total demand power of load and via net loss sum:

$P_D+P_{\mathrm{loss}}-\underset{i=1}{\overset{N_G}{\mathrm{\&Sigma;}}}{P}_i=0,---\left(5\right)$

In formula (5), for a certain moment thermoelectricity gross power; P _dfor this moment system load demand; P _lossbe expressed as active power loss in t moment electrical network, obtaining of active power loss adopts B Y-factor method Y, and its formula is:

$P_{\mathrm{loss}}=\underset{i=1}{\overset{N_G}{\mathrm{\&Sigma;}}}\underset{j=1}{\overset{N_G}{\mathrm{\&Sigma;}}}{P}_i{B}_{\mathrm{ij}}{P}_j+\underset{i=1}{\overset{N_G}{\mathrm{\&Sigma;}}}{B}_{0i}{P}_i+B_{00},---\left(6\right)$

B wherein _ij, B _0i, B ₀₀for B coefficient;

Step 2, obtain all kinds of parameters in model;

The initial feasible schedule set of step 3, acquisition environmental economy scheduling problem

In above-mentioned model, need to determine that decision variable is N _gthe active power of individual unit optimal value, produce feasible schedule set;

Step 4, to feasible schedule, set is evaluated

N the feasible schedule set that step 3 is produced be substitution target function type (2) and formula (3) respectively, respectively each feasible schedule is carried out the evaluation of environment and economy benefit, due to all corresponding two target functions of each feasible schedule, therefore need to carry out noninferior solution sequence to N feasible schedule, the feasible schedule sequence of not arranging is mutually kept in an outside set, is referred to as outside filing collection;

Step 5, feasible schedule is upgraded

Before upgrading, iterations initial value t=0 need to be set, and maximum iteration time T _max;

First random chooses m different feasible schedule as m Ge Leilei center, according to the active power of other all feasible schedule, to the Euclidean distance at each class center, N feasible schedule carried out to cluster, for simulating the forming process of brains storm processes thinking; By the class definition that contains noninferior solution, being elite's class, is general category without the class definition of noninferior solution, and on the basis of the fresh information obtaining in cluster, by selection, operates and mutation operation carries out iteration renewal to the active power of each feasible schedule generating set;

Step 6, judge whether whole feasible schedule has been upgraded

If i=N, shows that N feasible schedule upgraded, enter step 7; Otherwise return to step 5;

Step 7, the non-bad scheduling disaggregation that outside filing is concentrated are upgraded

The non-bad feasible schedule that Comparative economic benefit and discharge amount of pollution obtain is each time stored in to outside filing to be concentrated;

Step 8, the iteration of carrying out are exported final feasible schedule scheme

Judge whether iterations t reaches maximum of T _max, if not, iterations t=t+1 is set, forward step 4 to; If so, export the feasible schedule in current Noninferior Solution Set.

2. the thermal power plant's economic environment dispatching method based on brainstorming according to claim 1, its feature is: in described step 2,

According to grid load curve, determine the load valley of electrical network and load peak period, determining the power demand P of day part sharing of load _i ^min, P _i ^max;

From grid dispatching center statistics, obtain the data of current time system, comprise system total load P _d, total network loss P _lossparameter value B _ij, B _0i, B ₀₀;

According to soot emissions price, reduction of discharging price, CO2 emission exercise price, obtain the supplemental characteristic α of discharge amount of pollution _i, β _i, γ _i, ξ _i, λ _i;

Coal consumption price in producing according to electric power system assistant service price, unit of electrical energy, cost of electricity-generating while obtaining raw coal price, diesel-fuel price thermal power unit operation, obtain the parameter of fuel total cost.

3. the thermal power plant's economic environment dispatching method based on brainstorming according to claim 1, its feature is: in described step 3, the concrete producing method of feasible schedule set is as follows:

Step1: to front N _g-1individual generating set, in meeting the scope of maximum generation active power and minimum generating active power, produces the front N of each unit at random _g-1the active-power P i of individual generating set, i=1,2 ...., N _g-1;

Step2: according to equality constraint calculate N _gthe active power of individual unit;

Step3: computing network loss P _loss;

Step4: consider via net loss P _loss, according to formula calculate the active-power P of the last one dimension of each unit that meets equality constraint _d;

Step5: judge last N _gwhether individual unit meets the condition of capacity-constrained, if do not meet, re-executes step1 to step4 step, otherwise retains the feasible schedule producing;

According to same process, produce N feasible schedule set.

4. the thermal power plant's economic environment dispatching method based on brainstorming according to claim 1, its feature is: in described step 5, specific implementation process is as follows:

5.1) select operation

In the t time iteration, for any one current feasible schedule, according to brainstorming algorithm below, select the feasible schedule that will upgrade,

For i parent feasible schedule, produce the random number rand1 between a random value 0-1,

If rand1 is less than probability P 1; Select feasible schedule in current class to upgrade; Produce a random number rand2, if rand2 is less than probability P 2, select feasible schedule in Yi Gelei center or class as upgating object,

When the random number producing is less than P3, select the feasible schedule in elite's class center or class; Otherwise select the feasible schedule in general category center or class;

If rand2 is more than or equal to probability P 2, select at random two classes to produce the active power of new feasible schedule; Produce a random value, if random number is less than probability P 4, the cluster centre in two classes that choose is carried out to linear combination; Otherwise the active power of random two feasible schedule selecting is carried out linear combination from two classes selecting;

Otherwise, with the probability of 1-P1, from filing, concentrate the active power of selecting the feasible schedule that will make a variation,

Above-mentioned P1, P2, P3, P4 are the probability parameter carrying in brainstorming algorithm, select really to fix a number between 0-1;

5.2) mutation operation

5.2.1) in t iteration history before j feasible schedule selecting operation to obtain is dispatched the active power of i unit as the active power of i unit the t+1 time in j feasible schedule iterative formula as follows:

$P_{{\mathrm{idx}\_\mathrm{popu}}_{\mathrm{ji}}}(t+1)=\left\{\begin{array}{cc}\mathrm{random}({P_i}^{\min },{P_i}^{\max })& \mathrm{rand}\left(\mathrm{0,1}\right)<0.05\\ P_{{\mathrm{idx}\_\mathrm{popu}}_{\mathrm{ji}}}\left(t\right)+\mathrm{rand}\left(\mathrm{0,1}\right)\&times;(P_{{\mathrm{best}1}_{\mathrm{ji}}}\left(t\right)-P_{{\mathrm{best}2}_{\mathrm{ji}}}\left(t\right))& \mathrm{otherwise}\end{array}\right.,---\left(7\right)$

In formula (7), with be the active power of j unit of i concentrated feasible schedule of the t time iteration filing, if file concentrated feasible schedule number, be less than 2, in the active power when two feasible schedule of the random selection of former generation; According to formula (7), produce the front N of new feasible schedule _g-1 dimension, and check it whether to cross the border, be if so, bound; Otherwise the last one dimension by definite each the feasible schedule active power of power-balance constraint, guarantees that every one dimension of each feasible schedule active power is in its confining spectrum;

5.2.2) calculate the target function value of the new feasible schedule producing of i filial generation, filial generation and parent are carried out to non-ratio of less inequality, according to dominance relation, retain better feasible schedule.

5. the thermal power plant's economic environment dispatching method based on brainstorming according to claim 1, its feature is: in described step 7, outside the renewal of outside filing collection except the non-domination scheduling in population, also adopt crowding distance method to safeguard, concrete grammar is: the active power of the non-domination feasible schedule in population is put into outside filing one by one and concentrate, if the active power that the active power of this feasible schedule is filed concentrated feasible schedule by outside is arranged, the active power of this feasible schedule is deleted from filing to concentrate, otherwise the active power of this feasible schedule adds filing collection, if file the active power number of concentrated feasible schedule, be less than heap(ed) capacity, do not carry out deletion action, otherwise calculate the crowding distance that the active power of all feasible schedule is concentrated in current filing, the active power of deleting that feasible schedule of crowding distance minimum makes to file concentrated feasible schedule and remains on the number that is less than or equal to heap(ed) capacity.