CN105281372B - The multiple target multiagent distributed game optimization method of the Based on Distributed energy - Google Patents

Abstract

The invention discloses the multiple target multiagent distributed game optimization method of the Based on Distributed energy, belong to the technical field of Automation of Electric Systems.The present invention is directed to the multiple target that multiple distributed energy resource system presents, multiple constraint, non-linear and multi-agent Game characteristic proposes a kind of multiple target multiagent distributed game optimization method, economical according to multiple distributed energy combined optimization, the target requirement such as the feature of environmental protection and high efficiency, in conjunction with exerting oneself of various distributed energies self, multi-energy system multiple target combined optimization model is set up in climbing rate constraint etc., based on distributed coordination optimum theory, block mold is decomposed into several subsystem multiple target combined optimization models, the multi-objective optimization algorithm using improvement again solves the Pareto disaggregation obtaining each subsystem to it, ultimately form the optimum Pareto disaggregation of whole system, reliable decision support is provided for policymaker.

Description

The multiple target multiagent distributed game optimization method of the Based on Distributed energy

Technical field

The invention discloses the multiple target multiagent distributed game optimization method of the Based on Distributed energy, belong to electricity The technical field of Force system automation.

Background technology

Along with increasing wind-powered electricity generation, the photovoltaic distributed energy access network system, the associating of multi-energy system Optimization problem presents the complex characteristics such as multiple target, multiple constraint, meanwhile, multiple distributed energy Interest Main Body it Between there is mutual gaming characteristics, traditional distributed optimization method normally only consider its multiple target characteristic or Multi-agent Game characteristic, have ignored multiple target, the coordination problem of multiagent during distributed energy optimizes.Due to many Planting the energy all has self different independent characteristic and various distributed energy typically to adhere to different Interest Main Bodies separately, only Consider that the prioritization scheme of the multiple target characteristic of optimization problem have ignored comprehensive consideration of individual character.Additionally, due to Distributed energy is distributed rationally has diversified target requirement, only considers the simple target of its multiagent characteristic Demand, it is impossible to meet its actual requirement of engineering.

Summary of the invention

The technical problem to be solved is the deficiency for above-mentioned background technology, it is provided that Based on Distributed The multiple target multiagent distributed game optimization method of the energy, it is achieved that the optimum of multiple distributed energy resource is joined Put, solve the technical problem that multiple target multiagent distributed coordination optimizes.

The present invention adopts the following technical scheme that for achieving the above object

The multiple target multiagent distributed game optimization method of the Based on Distributed energy, comprises the steps:

A, set up the multiple target global optimization model of multi-energy system；

B, it is with each energy group according to distributed coordination optimum theory by described multiple target global optimization model decomposition Subsystem Model for Multi-Objective Optimization for main body；

C, solve each subsystem Model for Multi-Objective Optimization and obtain each energy group game strategies under multiple target demand Set, obtains multiple target global optimization model scheme collection under each target requirement in conjunction with distributed coordination theory；

D, to be solved multi-energy system by multiple target global optimization model scheme collection under each target requirement overall Pareto optimal solution set.

Multiple target multiagent distributed game optimization method the most excellent as the described Based on Distributed energy Change scheme, step A is for comprising wind power group, photovoltaic group, the multi-energy system of thermoelectricity group, with economic benefit Maximum, environmental pollution minimum, the minimum target of line loss, it is considered to account load balancing constraints, each energy group's unit Exert oneself restriction and the multiple target global optimization model of multi-energy system is set up in the constraint of climbing rate:

Multiple target: $\{\begin{array}{c}{\mathrm{maxf}}_1=\max \left(\underset{t=1}{\overset{T}{\Σ}}\left(\underset{i=1}{\overset{I}{\Σ}}{P}_{wit}{Q}_{wt}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}{Q}_{pt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}{Q}_{ct}\right)\right)\\ {\mathrm{minf}}_2=\min \left(\underset{t=1}{\overset{T}{\Σ}}\underset{k=1}{\overset{K}{\Σ}}\left(a_k{P_{ckt}}^2+b_k{P}_{ckt}+c_k\right)\right)\\ {\mathrm{minf}}_3=\min \underset{h=1}{\overset{H}{\Σ}}\underset{o=1,o\≠h}{\overset{H}{\Σ}}\left(V_h^2+V_o^2-2V_h{V}_o*\cos \left({\mathrm{\θ}}_h-{\mathrm{\θ}}_o\right)\right)g_{ho}\end{array},$

Account load balancing constraints: $\underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}=L_t+P_{loss,t},$

The restriction of exerting oneself of each energy group's unit: $\left\{\begin{array}{c}{P}_{wi\min }\≤P_{wit}\≤P_{wi\max },i=1,2,\mathrm{...},I\\ P_{pj\min }\≤P_{pjt}\≤P_{pj\max },j=1,2,\mathrm{...},J\\ P_{ck\min }\≤P_{ckt}\≤P_{ck\max },k=1,\mathrm{2...}K\end{array}\right.,$

The climbing rate constraint of each energy group's unit: $\left\{\begin{array}{c}{Z}_{wi\min }\≤P_{wi,t+1}-P_{wit}\≤Z_{wi\max },i=1,2,\mathrm{...},I\\ Z_{pj\min }\≤P_{pj,t+1}-P_{pjt}\≤Z_{pj\max },j=1,2,\mathrm{...},J\\ Z_{ck\min }\≤P_{ck,t+1}-P_{ckt}\≤Z_{ck\max },k=1,2,\mathrm{...},K\end{array}\right.,$

Wherein, P_wit、P_pjt、P_cktIt is respectively i-th Wind turbines, jth photovoltaic unit, kth thermal motor Group exerting oneself in t, P_wi,t+1、P_pj,t+1、P_ck,t+1Be respectively i-th Wind turbines, jth photovoltaic unit, Kth fired power generating unit is exerted oneself the t+1 moment, Q_wt、Q_pt、Q_ctRepresent wind-powered electricity generation, photovoltaic, thermoelectricity respectively Electricity price price, I, J, K are respectively the unit quantity of wind power group, photovoltaic group, thermoelectricity group, L_tBear for t Lotus aggregate demand, P_loss,tFor t line loss, V_h、V_oIt is respectively arbitrary node h, the voltage of node o, θ_h、 θ_oIt is respectively arbitrary node h, the phase angle of node o, g_hoFor the transconductance between node h, node o, H is Interstitial content, P_{wi min}、P_{pj min}、P_{ck min}It is respectively i-th Wind turbines, jth photovoltaic unit, kth The minimum load of fired power generating unit limits, P_{wi max}、P_{pj max}、P_{ck max}It is respectively i-th Wind turbines, jth light Volt unit, the EIAJ of kth fired power generating unit limit, Z_{wi min}、Z_{pj min}、Z_{ck min}It is respectively i-th wind-powered electricity generation Unit, jth photovoltaic unit, the climbing rate lower limit of kth fired power generating unit, Z_wimax、Z_{pj max}、Z_{ck max}Point Not Wei i-th Wind turbines, jth photovoltaic unit and the climbing rate upper limit of kth fired power generating unit, T is the time Yardstick, a_k、b_k、c_kEnvironmental pollution parameter for kth fired power generating unit.

Further, in the multiple target multiagent distributed game optimization method of the described Based on Distributed energy, step Each energy group described in rapid B is that the subsystem Model for Multi-Objective Optimization of main body includes: wind power group subsystem model, Photovoltaic group's subsystem model, thermoelectricity group's subsystem model,

Wind power group subsystem model: $\left\{\begin{array}{c}{F}_w=({\mathrm{maxf}}_1,{\mathrm{minf}}_3)\\ \underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}=L_t+P_{loss,t}\\ P_{wi\min }\≤P_{wit}\≤P_{wimax}\\ Z_{wi\min }\≤P_{wi,t+1}-P_{wit}\≤Z_{wimax}\end{array}\right.,$

Photovoltaic group's subsystem model: $\left\{\begin{array}{c}{F}_p=({\mathrm{maxf}}_1,{\mathrm{minf}}_3)\\ \underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}=L_t+P_{loss,t}\\ P_{pjmin}\≤P_{pjt}\≤P_{pj\max }\\ Z_{pjmin}\≤P_{pj,t+1}-P_{pjt}\≤Z_{pjmax}\end{array}\right.,$

Thermoelectricity group's subsystem model: $\left\{\begin{array}{c}{F}_c=({\mathrm{maxf}}_1,{\mathrm{minf}}_2,{\mathrm{minf}}_3)\\ \underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}=L_t+P_{loss,t}\\ P_{ckmin}\≤P_{ckt}\≤P_{ck\max }\\ Z_{ckmin}\≤P_{ck,t+1}-P_{ckt}\≤Z_{ckmax}\end{array}\right.,$

Wherein, F_w、F_p、F_cIt is respectively wind power group, photovoltaic group, the multiple objective function of thermoelectricity group's subsystem.

Further, in the multiple target multiagent distributed game optimization method of the described Based on Distributed energy, Step C combines distributed coordination theory and obtains multiple target global optimization model scheme collection under each target requirement Including:

Maximization of economic benefit scheme collection: $\left\{\begin{array}{c}{P}_{wit}^{n+1}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_1}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}\\ P_{pjt}^{n+1}=P_{pjt}^{\left(n\right)}+{\mathrm{\η}}_{pjt}*\frac{\∂f_1}{\∂P_{pjt}}+{\mathrm{\λ}}_{pjt}*\frac{\∂Y}{\∂P_{pjt}}\\ P_{ckt}^{n+1}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_1}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}\end{array}\right.,$

Environmental pollution minimizes scheme collection: $P_{ckt}^{(n+1)}=P_{ckt}^{\left(n\right)}+{\mathrm{\η}}_{ckt}*\frac{\∂f_2}{\∂P_{ckt}}+{\mathrm{\λ}}_{ckt}*\frac{\∂Y}{\∂P_{ckt}},$

Line loss minimizes scheme collection: $\left\{\begin{array}{c}{P}_{wit}^{n+1}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_3}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}\\ P_{pjt}^{n+1}=P_{pjt}^{\left(n\right)}+{\mathrm{\η}}_{pjt}*\frac{\∂f_3}{\∂P_{pjt}}+{\mathrm{\λ}}_{pjt}*\frac{\∂Y}{\∂P_{pjt}}\\ P_{ckt}^{n+1}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_3}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}\end{array}\right.,$

Wherein,It is respectively i-th Wind turbines, jth photovoltaic unit, kth thermoelectricity The nth iteration value that unit t is exerted oneself,It is respectively i-th Wind turbines, jth (n+1)th iterative value that individual photovoltaic unit, kth fired power generating unit t are exerted oneself, $Y=\underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}-L_t-P_{loss,t},{\mathrm{\η}}_{wit},{\mathrm{\η}}_{pjt},{\mathrm{\η}}_{ckt}$ For iterative parameter, λ_wit、λ_pjt、λ_ckt For Lagrangian, n is positive integer.

Further, in the multiple target multiagent distributed game optimization method of the described Based on Distributed energy, It is overall that step D is solved multi-energy system by multiple target global optimization model scheme collection under each target requirement The method of Pareto optimal solution set is: the multiple target global optimization model obtaining step C is under each target requirement Scheme collection carry out limited number of time iteration choosing and be positioned at limited of iteration result prostatitis and solve composition multi-energy system Overall Pareto optimal solution set.

As the described Based on Distributed energy multiple target multiagent distributed game optimization method further Prioritization scheme, solves each subsystem Model for Multi-Objective Optimization and obtains each energy group in multiple target demand in step C The method of lower game strategies set particularly as follows:

C1, each energy group predict that other energy group in the generating information of future time instance and estimates that other energy group's is individual Body strategy；

C2, with each self history of energy group generating information as participant and with the individual strategy of other energy group for competing The person of striving, carries out game according to each subsystem Model for Multi-Objective Optimization and obtains each energy group winning under multiple target demand Play chess strategy set.

Further, more the entering of the multiple target multiagent distributed game optimization method of the described Based on Distributed energy One-step optimization scheme, uses intelligent optimization algorithm to try to achieve each main body optimum game strategies set after step C2, Multiple target global optimization model scheme collection under each target requirement is obtained in conjunction with distributed coordination theory.

The present invention uses technique scheme, has the advantages that the Based on Distributed energy of proposition is many Target multiagent distributed game optimization method, had not only considered multiple target characteristic but also account for distributed energy and adhered to separately This individual character of different interests main body, carries for coordinating multiple target, multiagent problem in distributed energy optimization Supply a kind of feasible program, be primarily based on distributed coordination optimum theory by many for multiagent distributed energy resource system mesh Mark Optimized model is decomposed into several subsystem Model for Multi-Objective Optimization, then complication system is converted into several letters Single subsystem, and use the intelligent optimization method of advanced person that each subsystem Optimized model is solved and then inquired into Go out the optimal solution set of each subsystem, ultimately form the optimal Pareto scheme collection of multiple distributed energy total system, Combined optimization for multiple-energy-source resource configures the reliable decision support of offer, it is achieved that multiple distributed energy resource Allocation optimum.

Aspect and advantage that the present invention adds will part be given in the following description, and these will be from explained below In become obvious, or recognized by the practice of the present invention.

Accompanying drawing explanation

In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, required in embodiment being described below The accompanying drawing used is briefly described, it should be apparent that, the accompanying drawing in describing below is only some of the present invention Embodiment, for those of ordinary skill in the art, on the premise of not paying creative work, also may be used To obtain other accompanying drawing according to these accompanying drawings.

Fig. 1 is the block diagram that the Based on Distributed energy distributed game of multiple target multiagent of the present invention optimizes.

Detailed description of the invention

Embodiments of the present invention are described below in detail, and the embodiment described below with reference to accompanying drawing is example Property, it is only used for explaining the present invention, and is not construed as limiting the claims.

It will be understood to those skilled in the art that unless otherwise defined, all terms used herein (include skill Art term and scientific terminology) have the ordinary technical staff in the technical field of the invention be commonly understood by identical Meaning.Should also be understood that those terms defined in such as general dictionary should be understood that have with The meaning that meaning in the context of prior art is consistent, and unless defined as here, will not be by ideal Change or the most formal implication is explained.

For ease of the understanding to the embodiment of the present invention, below with as shown in Figure 1 comprise wind power group, photovoltaic group, The multiple target multiagent distributed game optimization method of the present invention is illustrated as a example by the multiple-energy-source total system of thermoelectricity group. This embodiment does not constitute the restriction to the embodiment of the present invention.

(1) combine the target requirements such as the economy of multiple distributed energy, the feature of environmental protection and high efficiency, comprehensively examine Consider the constraintss such as wind power group, photovoltaic group and restrictions of exerting oneself of thermoelectricity group's unit, climbing rate restriction, foundation with The various different distributions formula energy are the multiple target multiagent combined optimization model of Interest Main Body

(1) target:

Economic benefit: ${\mathrm{maxf}}_1=max\left(\underset{t=1}{\overset{T}{\Σ}}\right(\underset{i=1}{\overset{I}{\Σ}}{P}_{wit}{Q}_{wt}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}{Q}_{pt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}{Q}_{ct}))---\left(1\right),$

Environmental pollution: ${\mathrm{minf}}_2=min\left(\underset{t=1}{\overset{T}{\Σ}}\underset{k=1}{\overset{K}{\Σ}}\right(a_k{P_{ckt}}^2+b_k{P}_{ckt}+c_k))---\left(2\right),$

Line loss: ${\mathrm{minf}}_3=min\underset{h=1}{\overset{H}{\Σ}}\underset{o=1,o\≠h}{\overset{H}{\Σ}}(V_h^2+V_o^2-2V_h{V}_o*cos({\mathrm{\θ}}_h-{\mathrm{\θ}}_o))g_{ho}---\left(3\right),$

Wherein, P_wit、P_pjt、P_cktIt is respectively i-th Wind turbines, jth photovoltaic unit, kth thermal motor Group exerting oneself in t, P_wi,t+1、P_pj,t+1、P_ck,t+1Be respectively i-th Wind turbines, jth photovoltaic unit, Kth fired power generating unit is exerted oneself the t+1 moment, Q_wt、Q_pt、Q_ctRepresent wind-powered electricity generation, photovoltaic, thermoelectricity respectively Electricity price price, I, J, K are respectively the unit quantity of wind power group, photovoltaic group, thermoelectricity group, V_h、V_oRespectively For arbitrary node h, the voltage of node o, θ_h、θ_oIt is respectively arbitrary node h, the phase angle of node o, g_hoFor Transconductance between node h, node o, H is interstitial content, and T is time scale, a_k、b_k、c_kFor kth The environmental pollution parameter of individual fired power generating unit.

(2) constraints:

Account load balancing constraints: $\underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}=L_t+P_{loss,t}---\left(4\right),$

Wind power output retrains: P_{wi min}≤P_wit≤P_{wi max}, i=1,2 ..., I (5),

Photovoltaic units limits: P_{pj min}≤P_pjt≤P_{pj max}, j=1,2 ..., J (6),

Thermoelectricity units limits: P_{ck min}≤P_ckt≤P_{ck max}, k=1,2 ..., K (7),

Wind-powered electricity generation climbing rate retrains: Z_{wi min}≤P_wi,t+1-P_wit≤Z_{wi max}, i=1,2 ..., I (8),

Photovoltaic climbing rate retrains: Z_{pj min}≤P_pj,t+1-P_pjt≤Z_{pj max}, j=1,2 ..., J (9),

Thermoelectricity climbing rate retrains: Z_{ck min}≤P_ck,t+1-P_ckt≤Z_{ck max}, k=1,2 ..., K (10),

Wherein, L_tFor t load aggregate demand, P_loss,tFor t line loss, P_{wi min}、P_{pj min}、P_{ck min}Point Not Wei i-th Wind turbines, jth photovoltaic unit, kth fired power generating unit minimum load limit, P_{wi max}、P_{pj max}、P_{ck max}It is respectively i-th Wind turbines, jth photovoltaic unit, kth fired power generating unit EIAJ limits, Z_{wi min}、Z_{pj min}、Z_{ck min}Be respectively i-th Wind turbines, jth photovoltaic unit, the The climbing rate lower limit of k fired power generating unit, Z_wimax、Z_{pj max}、Z_{ck max}It is respectively i-th Wind turbines, jth The climbing rate upper limit of photovoltaic unit and kth fired power generating unit.

(2) theoretical according to large system decomposing coordination, can be by the multiple-objection optimization mould of above-mentioned multiple-energy-source total system The subsystem Model for Multi-Objective Optimization that it is different interests main body with wind power group, photovoltaic group and thermoelectricity group that type is decomposed into, Specifically can be to be decomposed into following form:

Wind power group subsystem model: $\left\{\begin{array}{c}{F}_w=({\mathrm{maxf}}_1,{\mathrm{minf}}_3)\\ \underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}=L_t+P_{loss,t}\\ P_{wi\min }\≤P_{wit}\≤P_{wimax}\\ Z_{wi\min }\≤P_{wi,t+1}-P_{wit}\≤Z_{wimax}\end{array}\right.---\left(11\right),$

Photovoltaic group's subsystem model: $\left\{\begin{array}{c}{F}_p=({\mathrm{maxf}}_1,{\mathrm{minf}}_3)\\ \underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}=L_t+P_{loss,t}\\ P_{pjmin}\≤P_{pjt}\≤P_{pj\max }\\ Z_{pj\min }\≤P_{pj,t+1}-P_{pjt}\≤Z_{pjmax}\end{array}\right.---\left(12\right),$

Thermoelectricity group's subsystem model: $\left\{\begin{array}{c}{F}_c=({\mathrm{maxf}}_1,{\mathrm{minf}}_2,{\mathrm{minf}}_3)\\ \underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}=L_t+P_{loss,t}\\ P_{ckmin}\≤P_{ckt}\≤P_{ck\max }\\ Z_{ckmin}\≤P_{ck,t+1}-P_{ckt}\≤Z_{ckmax}\end{array}\right.---\left(13\right),$

(3) according to subsystem model obtained above, each energy group predicts that other energy group is at future time instance Generating information and estimate the individual strategy of other energy group, with each self history of energy group generating information for ginseng Individuality with person and with other energy group is tactful as competitor, carries out according to each subsystem Model for Multi-Objective Optimization Game obtains each energy group game strategies set under multiple target demand, can obtain wind power group in each target Game strategies collection under demand is combined into u_w1,u_w3, photovoltaic group strategy set under each target requirement is u_p1,u_p3, Thermoelectricity group strategy set under each target requirement is u_c1,u_c2,u_c3

Order $Y=\underset{i=1}{\overset{I}{\Σ}}{P}_{wit}+\underset{j=1}{\overset{J}{\Σ}}{P}_{pjt}+\underset{k=1}{\overset{K}{\Σ}}{P}_{ckt}-L_t-P_{loss,t},$ Then have (all in the range of feasible zone):

u_w1Set of strategies is: $P_{wit}^{(n+1)}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_1}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}---\left(14\right),$

u_w3Set of strategies is: $P_{wit}^{(n+1)}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_3}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}---\left(15\right),$

u_p1Set of strategies is: $P_{pjt}^{(n+1)}=P_{pjt}^{\left(n\right)}+{\mathrm{\η}}_{pjt}*\frac{\∂f_1}{\∂P_{pjt}}+{\mathrm{\λ}}_{pjt}*\frac{\∂Y}{\∂P_{pjt}}---\left(16\right),$

u_p3Set of strategies is: $P_{pjt}^{(n+1)}=P_{pjt}^{\left(n\right)}+{\mathrm{\η}}_{pjt}*\frac{\∂f_3}{\∂P_{pjt}}+{\mathrm{\λ}}_{pit}*\frac{\∂Y}{\∂P_{pjt}}---\left(17\right),$

u_c1Set of strategies is: $P_{ckt}^{(n+1)}=P_{ckt}^{\left(n\right)}+{\mathrm{\η}}_{ckt}*\frac{\∂f_1}{\∂P_{ckt}}+{\mathrm{\λ}}_{ckt}*\frac{\∂Y}{\∂P_{ckt}}---\left(18\right),$

u_c2Set of strategies is: $P_{ckt}^{(n+1)}=P_{ckt}^{\left(n\right)}+{\mathrm{\η}}_{ckt}*\frac{\∂f_2}{\∂P_{ckt}}+{\mathrm{\λ}}_{ckt}*\frac{\∂Y}{\∂P_{ckt}}---\left(19\right),$

u_c3Set of strategies is: $P_{ckt}^{(n+1)}=P_{ckt}^{\left(n\right)}+{\mathrm{\η}}_{ckt}*\frac{\∂f_3}{\∂P_{ckt}}+{\mathrm{\λ}}_{ckt}*\frac{\∂Y}{\∂P_{ckt}}---\left(20\right),$

Formula (14) in formula (20),It is respectively i-th Wind turbines, jth photovoltaic The nth iteration value that unit, kth fired power generating unit t are exerted oneself,It is respectively i-th (n+1)th iterative value that individual Wind turbines, jth photovoltaic unit, kth fired power generating unit t are exerted oneself, η_wit、η_pjt、η_cktFor iterative parameter, λ_wit、λ_pjt、λ_cktFor Lagrangian, n is positive integer.

Intelligent optimization algorithm can also be used after step C2 to try to achieve each main body game strategies optimal solution, with rich Play chess strategy optimal solution as the game strategies set under each target requirement of each main body.

System global optimization model game strategies collection under each target requirement is combined into (u_w1,u_p1,u_c1), (u_c2), (u_w3,u_p3,u_c3).For above-mentioned game strategies set generation in the following ways:

f₁Under target: $\left\{\begin{array}{c}{P}_{wit}^{n+1}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_1}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}\\ P_{pjt}^{n+1}=P_{pjt}^{\left(n\right)}+{\mathrm{\η}}_{pjt}*\frac{\∂f_1}{\∂P_{pjt}}+{\mathrm{\λ}}_{pjt}*\frac{\∂Y}{\∂P_{pjt}}\\ P_{ckt}^{n+1}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_1}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}\end{array}\right.---\left(21\right),$

f₂Under target: $P_{ckt}^{(n+1)}=P_{ckt}^{\left(n\right)}+{\mathrm{\η}}_{ckt}*\frac{\∂f_2}{\∂P_{ckt}}+{\mathrm{\λ}}_{ckt}*\frac{\∂Y}{\∂P_{ckt}}---\left(22\right),$

f₃Under target: $\{\begin{array}{c}{P}_{wit}^{n+1}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_3}{\∂P_{wit}}+{\mathrm{\λ}}_{wit}*\frac{\∂Y}{\∂P_{wit}}\\ P_{pjt}^{n+1}=P_{pjt}^{\left(n\right)}+{\mathrm{\η}}_{pjt}*\frac{\∂f_3}{\∂P_{pjt}}+{\mathrm{\λ}}_{pjt}*\frac{\∂Y}{\∂P_{pjt}}\\ P_{ckt}^{n+1}=P_{wit}^{\left(n\right)}+{\mathrm{\η}}_{wit}*\frac{\∂f_3}{\∂P_{ckt}}+{\mathrm{\λ}}_{ckt}*\frac{\∂Y}{\∂P_{ckt}}\end{array}---\left(23\right),$

(4) on the basis of the initial value set, each in conjunction with (21), (22) and (23) formula iterative Exerting oneself of unit.Owing to the result of multiple-objection optimization is that Pareto optimizes disaggregation, so patent of the present invention uses In the following manner generates optimum Pareto disaggregation

(1) first basis (21) formula iteration 500 times, and preserve every generation result, at 500 iteration result In choose front 10 optimal result；

(2) similarly, according to 500 (P of formula (22) iteration_wit,P_pjtAll at feasible zone), preserve Iteration result each time, and choose front 10 optimal result；

(3) identical with step (1), choose 10 results of optimum.

Finally, 10 disaggregation of this every part are merged into a disaggregation having 30 optimal solutions, then this disaggregation Optimum Pareto disaggregation for whole system.

In sum, the multiple target multiagent distributed game optimization side of the Based on Distributed energy that the present invention proposes Method, had not only considered multiple target characteristic but also account for distributed energy and adhered to this individual character of different interests main body separately, For coordinating multiple target in distributed energy optimization, multiagent problem provide a kind of feasible program, be primarily based on point Multiagent distributed energy resource system Model for Multi-Objective Optimization is decomposed into several subsystems by cloth coordination optimization theory System Model for Multi-Objective Optimization, then complication system is converted into several simple subsystems, and use the intelligence of advanced person Each subsystem Optimized model can solve and then inquire into the optimal solution set each subsystem, finally by optimization method Forming the optimal Pareto scheme collection of multiple distributed energy total system, the combined optimization for multiple-energy-source resource is joined Put the reliable decision support of offer, it is achieved that the allocation optimum of multiple distributed energy resource.

One of ordinary skill in the art will appreciate that: accompanying drawing is the schematic diagram of an embodiment, the mould in accompanying drawing Block or flow process are not necessarily implemented necessary to the present invention.

As seen through the above description of the embodiments, those skilled in the art is it can be understood that arrive this Bright can add the mode of required general hardware platform by software and realize.Based on such understanding, the present invention's The part that prior art is contributed by technical scheme the most in other words can embody with the form of software product Out, this computer software product can be stored in storage medium, such as ROM/RAM, magnetic disc, CD etc., Including some instructions with so that a computer equipment (can be personal computer, server, or network Equipment etc.) perform embodiments of the invention or the method described in some part of embodiment.

Claims (6)

1. the multiple target multiagent distributed game optimization method of the Based on Distributed energy, it is characterised in that comprise the steps:

A, set up the multiple target global optimization model of multi-energy system,

For comprising wind power group, photovoltaic group, the multi-energy system of thermoelectricity group, maximum with economic benefit, environmental pollution minimum, the minimum target of line loss, it is considered to account load balancing constraints, each energy group's unit exert oneself restriction and the multiple target global optimization model of multi-energy system is set up in the constraint of climbing rate:

Multiple target:

Account load balancing constraints:

The restriction of exerting oneself of each energy group's unit:

The climbing rate constraint of each energy group's unit:

Wherein, P_wit、P_pjt、P_cktIt is respectively i-th Wind turbines, jth photovoltaic unit, kth fired power generating unit exerting oneself in t, P_wi,t+1、P_pj,t+1、P_ck,t+1Respectively i-th Wind turbines, jth photovoltaic unit, kth fired power generating unit is exerted oneself the t+1 moment, Q_wt、Q_pt、Q_ctRepresenting the electricity price price of wind-powered electricity generation, photovoltaic, thermoelectricity respectively, I, J, K are respectively the unit quantity of wind power group, photovoltaic group, thermoelectricity group, L_tFor t load aggregate demand, P_loss,tFor t line loss, V_h、V_oIt is respectively arbitrary node h, the voltage of node o, θ_h、θ_oIt is respectively arbitrary node h, the phase angle of node o, g_hoFor the transconductance between node h, node o, H is interstitial content, P_{wi min}、P_{pj min}、P_{ck min}It is respectively the minimum load restriction of i-th Wind turbines, jth photovoltaic unit, kth fired power generating unit, P_{wi max}、P_{pj max}、P_{ck max}It is respectively the EIAJ restriction of i-th Wind turbines, jth photovoltaic unit, kth fired power generating unit, Z_{wi min}、Z_{pj min}、Z_{ck min}It is respectively i-th Wind turbines, jth photovoltaic unit, the climbing rate lower limit of kth fired power generating unit, Z_wimax、Z_{pj max}、Z_{ck max}Being respectively i-th Wind turbines, jth photovoltaic unit and the climbing rate upper limit of kth fired power generating unit, T is time scale, a_k、b_k、c_kEnvironmental pollution parameter for kth fired power generating unit；

B, it is based on the subsystem Model for Multi-Objective Optimization of each energy group according to distributed coordination optimum theory by described multiple target global optimization model decomposition；

C, solve each subsystem Model for Multi-Objective Optimization and obtain each energy group game strategies set under multiple target demand, obtain multiple target global optimization model scheme collection under each target requirement in conjunction with distributed coordination theory；

D, solved multi-energy system entirety Pareto optimal solution set by multiple target global optimization model scheme collection under each target requirement.

The multiple target multiagent distributed game optimization method of the Based on Distributed energy the most according to claim 1, it is characterized in that, each energy group described in step B is that the subsystem Model for Multi-Objective Optimization of main body includes: wind power group subsystem model, photovoltaic group's subsystem model, thermoelectricity group's subsystem model

Wind power group subsystem model:

Photovoltaic group's subsystem model:

Thermoelectricity group's subsystem model:

Wherein, F_w、F_p、F_cIt is respectively wind power group, photovoltaic group, the multiple objective function of thermoelectricity group's subsystem.

The multiple target multiagent distributed game optimization method of the Based on Distributed energy the most according to claim 2, it is characterised in that step C combines distributed coordination theory and obtains multiple target global optimization model scheme collection under each target requirement and include:

Maximization of economic benefit scheme collection:

Environmental pollution minimizes scheme collection:

Line loss minimizes scheme collection:

Wherein,The nth iteration value that respectively i-th Wind turbines, jth photovoltaic unit, kth fired power generating unit t are exerted oneself,(n+1)th iterative value that respectively i-th Wind turbines, jth photovoltaic unit, kth fired power generating unit t are exerted oneself,η_wit、η_pjt、η_cktFor iterative parameter, λ_wit、λ_pjt、λ_cktFor Lagrangian, n is positive integer.

The multiple target multiagent distributed game optimization method of the Based on Distributed energy the most according to claim 3, it is characterized in that, the method that step D is solved multi-energy system entirety Pareto optimal solution set by multiple target global optimization model scheme collection under each target requirement is: the multiple target global optimization model obtaining step C scheme collection under each target requirement carries out limited number of time iteration choosing and is positioned at limited of iteration result prostatitis and solves and form multi-energy system entirety Pareto optimal solution set.

5. according to the multiple target multiagent distributed game optimization method of the Based on Distributed energy described in any one claim in Claims 1-4, it is characterized in that, step C solves each subsystem Model for Multi-Objective Optimization obtain each energy group game strategies set under multiple target demand method particularly as follows:

C1, each energy group predict that other energy group in the generating information of future time instance and estimates that other energy group's is individual tactful；

C2, with each self history of energy group generating information as participant and with the individual strategy of other energy group as competitor, carry out game according to each subsystem Model for Multi-Objective Optimization and obtain each energy group game strategies set under multiple target demand.

The multiple target multiagent distributed game optimization method of the Based on Distributed energy the most according to claim 5, it is characterized in that, use intelligent optimization algorithm to try to achieve each main body optimum game strategies set after step C2, obtain multiple target global optimization model scheme collection under each target requirement in conjunction with distributed coordination theory.