CN103762589B - A kind of new forms of energy capacity ratio hierarchy optimization method in electrical network - Google Patents

Abstract

The invention discloses a kind of new forms of energy capacity ratio hierarchy optimization method in electrical network.The present invention is divided into and inside and outside two-layerly carries out iterative computation, computation model is as follows: internal layer is exerted oneself to it and carried out time series modeling on the basis considering this area's new forms of energy characteristic, best for target with electrical network energy-saving and emission-reduction benefit, consider the factors such as part throttle characteristics, peak load regulation characteristic, variety classes thermal power plant unit coupled thermomechanics characteristic, establish the annual sequential production simulation simulation model taking into account generation of electricity by new energy; Skin is capacity ratio Optimized model, within the energy-saving and emission-reduction benefit of layer model be that fitness function upgrades individual search direction, determine generation of electricity by new energy proportioning capacity, decrease the blindness of stochastic generation new forms of energy installed capacity, improve optimization efficiency and precision.The present invention can be used for the optimization of provincial power network new forms of energy capacity, and to economizing the new forms of energy installed capacity planning of (district) electrical network, low-carbon electric power requires lower net source planning and practical power systems to dispatch all to have important directive significance.

Description

A kind of new forms of energy capacity ratio hierarchy optimization method in electrical network

Technical field

The present invention relates to new forms of energy capacity ratio optimization method in a kind of electrical network, particularly relate to a kind of new forms of energy capacity ratio hierarchy optimization method in electrical network, belong to energy-conserving and emission-cutting technology field.

Background technology

The low carbonization of power industry is reply global warming, realizes the key of Chinese society sustainable economic development.In this context, China is proposed a series of energy development policy, encourages large-scale development and the utilization of regenerative resource, particularly wind energy and solar energy resources.After Wind Power Generation Industry fast development, just starting solar energy industry exploitation upsurge in recent years, the large province of many generations of electricity by new energy has planned jumbo wind energy and solar power generation simultaneously, but extensive planning is only separately carried out for foundation to wind-powered electricity generation, photovoltaic development with wind energy and solar energy resources in various places at present, does not consider wind-powered electricity generation, its optimum proportioning capacity of photovoltaic generation operation characteristic coordination optimization.Because the design and construction cycle of wind energy and solar power generation is short, plan with regional normal power supplies in development process, Electric Power Network Planning disconnects, cause " abandoning wind " in actual motion, " abandoning light " phenomenon is serious.In order to better promote the effect of electrical network in development low-carbon economy, give full play to solar energy and the wind energy in time and geographically natural complementary advantage had, improve Large Copacity wind energy and solar power generation receiving ability to greatest extent, program results is made more to press close to electric power system practical operation situation, must based on wind energy and solar power generation year receiving ability, the installed capacity of unified coordinated planning wind-powered electricity generation and photovoltaic, plays its greatest benefit.

Existing document is studied the honourable proportioning economizing (district) electrical network at present.Document one " Size optimization fora hybrid photovoltaic-wind energy system " (Electrical Power and Energy Systems the 42nd volume the 448th page) establishes the capacity Optimized model of honourable association system, has carried out capacity configuration calculating based on this model for various boundary conditions.But the method is wind, light receiving situation under load peak-valley difference maximum case, wind-powered electricity generation every day, photovoltaic power producing characteristics and the whole network operational mode can not be embodied, if be used for instructing local wind electricity, photovoltaic installed capacity, result of calculation will be too conservative, is unfavorable for the energy-saving and emission-reduction benefit promoting electrical network.Document two " Multicriteria OptimalSizing of Photovoltaic-Wind Turbine Grid Connected Systems " (IEEE Trans on EnergyConversion the 28th volume the 2nd phase the 370th page) adopts the improve PSO algorithm based on time stimulatiom to solve the optimum capacity configuration of somewhere scene, and sensitivity analysis has been carried out to wind speed size, intensity of illumination, obtain the honourable installed capacity configuration under different natural conditions.Because model adopts stochastic simulation to obtain wind speed size and intensity of illumination, can not accurately reflect this area's wind, light exerts oneself temporal characteristics and electric power sequential balance, thus effective technical support can not be provided for the planning of this area's actual electric network and scene construction.Document three " A New Methodology for Optimizing the Size ofHybrid PV/wind System " (IEEE International Conference on Sustainable EnergyTechnologies the 922nd page) adopts improvement of differential evolution algorithm to solve the best scene proportioning of certain regional power grid, proves that wind-light combined power generation system has better economy and reliability than independent wind generator system.But, in the process of founding mathematical models, do not consider the start and stop characteristic of conventional power unit and the coupled thermomechanics characteristic of thermal power plant unit, cause the operation result of conventional power unit and practical power systems deviation comparatively large, affect the credibility of honourable proportioning result of calculation.

In sum, the many employings of existing method are based on the balance of electric power and ener method of typical case's day, what provide is the generation of electricity by new energy balance under most serious conditions, can not embody every day wind-powered electricity generation and photovoltaic power producing characteristics and the whole network how to optimize Unit Commitment and service arrangement, and when carrying out modeling to electrical network production simulation, consideration is comprehensive not.If be used for the installed capacity of guiding plan wind-powered electricity generation and photovoltaic, result of calculation will be too conservative, is unfavorable for the energy-saving and emission-reduction benefit improving electrical network.

Summary of the invention

Technical problem to be solved by this invention is to overcome prior art deficiency, a kind of new forms of energy capacity ratio hierarchy optimization method in electrical network is provided, consider the actual condition change of electric power system, further increase the confidence level of optimum results, there is strong adaptability, advantage that reliability is high.

The present invention solves the problems of the technologies described above by the following technical solutions:

A kind of new forms of energy capacity ratio hierarchy optimization method in electrical network, the energy source in described electrical network comprises thermoelectricity and at least one new forms of energy, and described optimization method comprises the following steps:

The installed capacity of all kinds of new forms of energy in step 1, initialization electrical network, as the initial value of outer Optimized model;

Step 2, the installed capacity input internal layer Optimized model of current all kinds of new forms of energy that outer Optimized model is obtained; Internal layer Optimized model, by solving with drag, obtains under the installed capacity of current all kinds of new forms of energy, the whole year operation state of the fired power generating unit making CO2 emissions F minimum:

$\min F=\underset{j=1}{\overset{N_j}{\Σ}}\underset{t=1}{\overset{T}{\Σ}}\{{\mathrm{\α}}_j{Y}_j^t+{\mathrm{\β}}_j{Z}_j^t+a_j{P}_j^t+b_j\}\·\mathrm{\γ}$

Be tied in:

$P_j^{t+1}-P_j^t\≤{\mathrm{\ΔP}}_j^{up}$

$P_j^t-P_j^{t+1}\≤{\mathrm{\ΔP}}_j^{down}$

$X_j^t\·P_j^{min}\≤P_j^t\≤X_j^t\·P_j^{max}$

$0\≤Y_j^t+Z_j^t\≤1$

$Y_j^t+\underset{i=1}{\overset{k}{\Σ}}{Z}_j^{t\mathrm{+i}}\≤1$

$Z_j^t+\underset{i=1}{\overset{k}{\Σ}}{Y}_j^{t\mathrm{+i}}\≤1$

$P_{BY}^t=C_j^b\·H_j^t$

$H_j^t\·C_j^b\≤P_{CQ}^t\≤P_j^{max}-H_j^t\·C_j^v$

$\underset{j=1}{\overset{N_j}{\Σ}}{P}_j^t+P_0^t=P_1^t$

$-\underset{j=1}{\overset{N_j}{\Σ}}{P}_j^{max}-C_N^t\≤-P_1^t-S_p$

$\underset{j=1}{\overset{N_j}{\Σ}}{P}_j^{min}+C_N^t\≤P_1^t-S_N$

$0\≤P_i^t\≤N_i\·P_i^{t*},i=1,2,...,CL$

Wherein, for binary variable, represent the starting state of jth platform fired power generating unit t, 1 represents that unit starts, and 0 represents that unit is not at starting state; also be binary variable, represent jth platform fired power generating unit t stopped status, 1 represents that unit is shut down, and 0 represents that unit is not in stopped status; for the jth platform fired power generating unit of t is exerted oneself; and be independent variable; N _jrepresent the total number of units participating in optimizing fired power generating unit; T represents internal layer Optimized model simulation time length; α _jfor jth platform fired power generating unit opens machine-made egg-shaped or honey-comb coal briquets consumption; β _jfor jth platform fired power generating unit shuts down coal consumption; a _jfor the coal consumption of separate unit fired power generating unit is with changed power slope; b _jfor separate unit fired power generating unit coal consumption constant; γ is CO2 emission coefficient; be respectively swash ratio of slope and the lower climbing rate of jth platform fired power generating unit; be respectively minimum load value and the maximum output value of jth platform fired power generating unit; for binary variable, represent jth platform fired power generating unit t running status, 1 represents that unit runs, and 0 represents that unit does not run; K is the default minimum time step opening machine or shutdown; be respectively in the heat supply phase, in fired power generating unit, back pressure type thermal power plant unit and bleeder thermal power plant unit t exerts oneself; for t load of heat; for thermal power plant unit coupled thermomechanics coefficient; for total power load of t electrical network; for the electric power sum of all kinds of new forms of energy generations that t electrical network is received; S _p, S _nbe respectively the positive/negative spinning reserve of electrical network; for generation of electricity by new energy is at the credible capacity of day part; for the electric power of the i-th class new forms of energy generation that t electrical network is received; N _ifor the installed capacity of the i-th class new forms of energy in the electrical network that outer Optimized model inputs; for the i-th class new forms of energy long time scale in electrical network is exerted oneself seasonal effect in time series normalized value; CL is the classification sum of new forms of energy in electrical network;

Step 3, export the minimum CO2 emissions obtained to outer Optimized model;

Step 4, outer Optimized model judge whether to meet the stopping criterion for iteration preset, in this way, then by the installed capacity N of each new forms of energy minimum for CO2 emissions _iand the starting state of fired power generating unit stopped status exert oneself with fired power generating unit export as final optimum results, optimize and terminate; As no, then using the minimum CO2 emissions of internal layer Optimized model output as fitness function value, the installed capacity of all kinds of new forms of energy upgraded, then goes to step 2.

Preferably, described internal layer Optimized model uses branch and bound method (Branch and Bound is called for short BAB) to solve described model.

Preferably, described outer Optimized model uses particle cluster algorithm (Particle Swarm optimization, be called for short PSO) or based on the improvement bacterial foraging algorithm (Bacteria foraging algorithm-Particle Swarmoptimization, be called for short BFAPSO) of particle cluster algorithm.

Preferably, the variable solution space constraint of described outer Optimized model is as follows:

$N_i^{\min }\≤{\mathrm{\θ}}_i^m\≤N_i^{\max },=1,2,...,CL$

In formula, θ represents that the variable that needs are optimized, m represent m the individuality of variable θ, represent the planning installed capacity maximum of the i-th class new forms of energy and existing installed capacity value respectively.

Compared to existing technology, technical solution of the present invention and optimal technical scheme thereof have following beneficial effect:

1, the present invention utilizes historical data to generate long time scale time series, and it can be used as the constraints of optimized algorithm, substantially increases the reliability of optimum results, and comparatively accurately can reflect all kinds of new forms of energy in this area, load exerts oneself temporal characteristics.

2, hierarchy optimization algorithm can reduce amount of calculation to greatest extent, effectively solves electrical network new forms of energy installation planning problem, meets planning personnel's computation requirement.

3, outer algorithm adopts BFAPSO algorithm, can effectively improve computational accuracy and computational efficiency.

4, internal layer adopts BAB Algorithm for Solving based on the production simulation problem of time stimulatiom method, all kinds of new forms of energy year characteristic can be taken into full account, maximize and improve all kinds of new forms of energy actual online quantity of electricity, the operation of more realistic electric power system and the low carbonization requirement of electric power, to practical power systems planning and scheduling, Correspondence policy formulation have important directive significance.

5, the production simulation that have employed based on sequential due to internal layer emulates, the more realistic electric power system of this model, on this basis to new forms of energy installed capacity planning, can increase reasonability and the credibility of new forms of energy capacity ratio result.

6, the production simulation that have employed based on sequential due to internal layer of the present invention emulates, analysis and assessment can be carried out to generation of electricity by new energy ruuning situation under planning scene in planning process, namely can carry out coordination optimization to generation of electricity by new energy and the generating of conventional thermoelectricity, unit in net is dispatched by Various Seasonal; Reference frame can be provided for new forms of energy annual running mode, industrial development planning.With the multiple electricity of new forms of energy for principle, increase economy and the energy-saving and emission-reduction benefit of system cloud gray model; Consider the factor of rationing the power supply of operation of power networks, science and the reasonability of generation of electricity by new energy plan can be ensured.

Accompanying drawing explanation

Fig. 1 is the basic flow sheet of the inventive method embodiment;

Northeast province of Tu2Shi China forcasted years year normalization wind-powered electricity generation sequence;

Northeast province of Tu3Shi China forcasted years year normalization photovoltaic sequence;

Northeast province of Tu4Shi China forcasted years load is exerted oneself sequence;

Northeast province of Tu5Shi China weekly honourable population mean rations the power supply rate distribution map.

Embodiment

Below in conjunction with accompanying drawing, technical scheme of the present invention is described in detail:

New forms of energy capacity ratio hierarchy optimization method in the electrical network that the present invention proposes, be divided into and inside and outside two-layerly carry out iterative computation, computation model is as follows respectively: internal layer is exerted oneself to it and carried out time series modeling on the basis considering all kinds of new forms of energy characteristic in this area, best for target with electrical network energy-saving and emission-reduction benefit, consider part throttle characteristics, peak load regulation characteristic, the factors such as variety classes thermal power plant unit coupled thermomechanics characteristic, establish the annual sequential production simulation simulation model taking into account generation of electricity by new energy, the more realistic electric power system of this model, on this basis new forms of energy installed capacity is planned, reasonability and the credibility of new forms of energy proportioning result can be increased.Skin is new forms of energy capacity ratio Optimized model, within the energy-saving and emission-reduction benefit of layer model be that fitness function upgrades individual search direction, determine all kinds of generation of electricity by new energy proportioning capacity, decrease the blindness of stochastic generation new forms of energy installed capacity, improve optimization efficiency and precision.

Above-mentioned internal layer Optimized model is a typical mixed integer programming problem, and the present invention preferably adopts efficient branch and bound method to solve.And for large complicated electric power system, model is complicated, relate to variable more.For improving optimization efficiency further, the present invention preferably adopts the improvement bacterial foraging algorithm (BFAPSO) based on particle cluster algorithm, expands search volume, avoid precocious generation, strengthen local search ability, improve optimizing ability to reach, greatly improve the object of efficiency of algorithm.Above-mentioned algorithm is stable ripe computing, possesses very high engineering practicability after tested, meets the demand of technical solution of the present invention.

Have more realistic meaning for the ease of public understanding, below only to consider the provincial power network of wind-powered electricity generation and these two kinds of new forms of energy of photovoltaic generation, a preferred embodiment of the present invention to be described.

The rudimentary algorithm thinking of the preferred embodiment is as follows: the installed capacity of outer employing BFAPSO algorithm initialization scene; Internal layer, after outer wind-powered electricity generation, photovoltaic installed capacity are imported into, adopts BAB algorithm to carry out sequential production simulation, optimizes Unit Commitment machine plan and unit output, on the basis that the system of guarantee CO2 emissions is minimum, as much as possible receives wind-powered electricity generation more, photovoltaic exerts oneself.Now CO2 emission value is turned back to outer layer model (as fitness function value), adopt BFAPSO algorithm to wind-powered electricity generation, photovoltaic installed capacity optimizing, until meet the condition of termination.

The basic procedure of this preferred embodiment as shown in Figure 1, specifically comprises the following steps:

(1) time series modeling:

To long time scale wind power output time series modeling, and be normalized:

$P_w^{t*}=P_{w,o}^t/C_w$

In formula: for normalization wind power output time series, for history wind power output time series, C _wfor the total installation of generating capacity of this year region wind-powered electricity generation.

Long time scale photovoltaic is exerted oneself time series modeling, and is normalized:

$P_v^{t*}=P_{v,o}^t/C_v$

In formula: for normalization photovoltaic is exerted oneself time series, for history photovoltaic is exerted oneself time series, C _vfor the total installation of generating capacity of this year region photovoltaic.

Obtaining load time series of exerting oneself according to demand history data is P _l ^t.

(2) initialization wind-powered electricity generation and photovoltaic installed capacity:

S _W,min<S _W,0<S _W,max

S _S,min<S _S,0<S _S,max

In formula: S _{w, 0}for initial installed capacity of wind-driven power, S _{s, 0}for initial photovoltaic installed capacity, S _{w, min}for existing installed capacity of wind-driven power, S _{s, min}for existing photovoltaic installed capacity, S _{w, max}for allowing the maximum installed capacity of wind-powered electricity generation, S _{s, max}for allowing the maximum installed capacity of photovoltaic.

(3) inner layer B AB algorithm optimization calculates:

Set up target function as follows:

$F=\underset{j=1}{\overset{N_j}{\&Sigma;}}\underset{t=1}{\overset{T}{\&Sigma;}}\{{\mathrm{\&alpha;}}_j{Y}_j^t+{\mathrm{\&beta;}}_j{Z}_j^t+a_j{P}_j^t+b_j\}\&CenterDot;\mathrm{\&gamma;}$

In formula, F is CO2 emissions; for binary variable, represent the starting state of jth platform fired power generating unit t, 1 represents that unit starts, and 0 represents that unit is not at starting state; also be binary variable, represent jth platform fired power generating unit t stopped status, 1 represents that unit is shut down, and 0 represents that unit is not in stopped status; for the jth platform fired power generating unit of t is exerted oneself; and be independent variable; N _jrepresent the total number of units participating in optimizing fired power generating unit; T represents internal layer simulation time length; α _jfor jth platform fired power generating unit opens machine-made egg-shaped or honey-comb coal briquets consumption; β _jfor jth platform fired power generating unit shuts down coal consumption; a _jfor the coal consumption of separate unit fired power generating unit is with changed power slope; b _jfor separate unit fired power generating unit coal consumption constant; γ is CO2 emission coefficient.

Fired power generating unit constraints is set:

1) fired power generating unit climbing rate constraint

$P_j^{t+1}-P_j^t\&le;{\mathrm{\&Delta;P}}_j^{up}$

$P_j^t-P_j^{t+1}\&le;{\mathrm{\&Delta;P}}_j^{down}$

In formula: be respectively swash ratio of slope and the lower climbing rate of jth platform unit.

2) fired power generating unit units limits

$X_j^t\&CenterDot;P_j^{min}\&le;P_j^t\&le;X_j^t\&CenterDot;P_j^{max}$

In formula: be respectively minimum load value and the maximum output value of unit; for binary variable, represent jth platform unit t running status, 1 represents that unit runs, and 0 represents that unit does not run.

3) fired power generating unit start and stop state constraint

$0\&le;Y_j^t+Z_j^t\&le;1$

$Y_j^t+\underset{i=1}{\overset{k}{\&Sigma;}}{Z}_j^{t+i}\&le;1$

$Z_j^t+\underset{i=1}{\overset{k}{\&Sigma;}}{Y}_j^{t+i}\&le;1$

In formula: k opens machine or minimum parameter decision downtime by unit is minimum, that reflects the minimum time step opening machine or shutdown.The consideration of this constraint, physical characteristic and Unit Commitment machine-made egg-shaped or honey-comb coal briquets mainly owing to being subject to unit consume the restriction of cost, and unit can not start and stop frequently.

4) thermal power plant unit heat supply phase power producing characteristics constraint

The present invention considers 2 type thermal power plant unit: back pressure type thermal power plant unit and bleeder thermal power plant unit.

Back pressure unit units limits:

$P_{BY}^t=C_j^b\&CenterDot;H_j^t$

To bleed unit units limits:

$H_j^t\&CenterDot;C_j^b\&le;P_{CQ}^t\&le;P_j^{max}-H_j^t\&CenterDot;C_j^v$

In formula: for t load of heat; for thermal power plant unit coupled thermomechanics coefficient.

System constraints is set:

1) region account load balancing constraints

$\underset{j=1}{\overset{N_j}{\&Sigma;}}{P}_j^t+P_w^t+P_v^t=P_1^t$

In formula: for the total power load of system (electrical network); for the wind-powered electricity generation electric power that t is received; for the photovoltaic electric power that t is received.

2) the positive/negative spinning reserve capacity constraint of system

$-\underset{j=1}{\overset{N_j}{\&Sigma;}}{P}_j^{\max }-C_N^t\&le;-P_1^t-S_p$

$\underset{j=1}{\overset{N_j}{\&Sigma;}}{P}_j^{min}+C_N^t\&le;P_1^t-S_N$

In formula: S _p, S _nthe positive/negative spinning reserve of the system that is respectively; for generation of electricity by new energy is at the credible capacity of day part, day part generation of electricity by new energy can be got according to actual conditions and to exert oneself the certain proportion of sum, for the wind in the present embodiment, light generating, its value be taken as day part scene exert oneself sum 60%-80% between; Credible capacity of being exerted oneself by day part scene includes conventional power unit start calculation of capacity category in, reduces its start capacity, can better receive wind energy and solar power generation.

3) system wind-powered electricity generation, photovoltaic units limits

$0\&le;P_w^t\&le;N_w\&CenterDot;P_w^{t*}$

$0\&le;P_v^t\&le;N_v\&CenterDot;P_v^{t*}$

In formula: N _w, N _vrepresent the installed capacity of wind-powered electricity generation and photovoltaic respectively, its value is imported into after being optimized by skin; be respectively the normalized value that wind power output and photovoltaic are exerted oneself, obtained by step (1).

According to above-mentioned constraints and target function, mature and stable BAB algorithm is utilized to be optimized calculating, under obtaining the restriction of current peak modulation capacity, the carbon dioxide minimum emissions of system.After remembering the n-th suboptimization iterative computation, the carbon dioxide minimum emissions obtaining system is F _n.

(4) outer BFAPSO algorithm:

The constraint of variable solution space is set:

$N_s^{\min }\&le;{\mathrm{\&theta;}}_s^m\&le;N_s^{\max }$

In formula: θ represents that the variable that needs are optimized, m represent m the individuality of variable θ, and s represents individual dimension, represents installed capacity of wind-driven power during s=1, represents photovoltaic installed capacity during s=2; represent the installed capacity value of existing scene, the honourable installed capacity maximum of representative planning.

Fitness function is set:

$J=\min F\left({\mathrm{\&theta;}}_s^m\right)$

In formula: J is the fitness function of outer layer model, and its value is internal layer target function value, is system CO2 emissions, outer algorithm upgrades individual search direction according to its value size.

For the n-th suboptimization iteration, by carbon dioxide minimum emissions F _n-1substitute in mature and stable BFAPSO algorithm, utilize above-mentioned constraints and fitness function, solve the wind-powered electricity generation after being optimized and photovoltaic installed capacity is respectively S _w,nand S _s,n

(5) judge whether to meet end condition:

The setting iterations upper limit is n _maxif, Optimized Iterative frequency n=n _maxtime, terminate Optimized Iterative, go to step (6).If n<n _max, then return step (3) and be optimized iterative computation.

(6) optimal result is exported:

Export the result of optimum iteration, namely time, wind-powered electricity generation and the photovoltaic installed capacity of system optimal are respectively S _w,pand S _s,p, and under this installed capacity, the machine that the opens state of fired power generating unit, stopped status exert oneself with fired power generating unit

In order to test the validity of the inventive method, the method in application specific embodiment has carried out simulating, verifying to the honourable proportioning planning of northeast province of China.

According to this province's Master Plan requirement, planning level year wind-powered electricity generation, photovoltaic total installed capacity is no more than 8000MW.This province has blower fan 2646.4MW at present, and photovoltaic installation 530.83MW, wind-powered electricity generation and photovoltaic summation account for 14.01% of system total installed capacity.Forcasted years year wind-powered electricity generation sequence after normalization, annual photovoltaic sequence, load exert oneself sequence as shown in figs 2-4, and simulation time step-length is 1 hour.Honourable ratio optimization calculating is carried out to above-mentioned data, obtains result as follows:

The result that employing algorithms of different carries out honourable ratio optimization is as shown in table 1.Comparative analysis result of calculation is known: interior layer model uses the operational mode proposed based on typical case's day, is the receiving situation of scene when considering the most serious.Adopt the emulation of sequential production simulation than interior layer model, system then more gives off 0.054 hundred million ton of carbon dioxide, and new forms of energy receive capacity calculation relatively conservative, and program results confidence level is not high.Under the constraint that electrical network low-carbon (LC) requires, wind-powered electricity generation installation rises to 3668MW from 2646.4MW, increases by 38.60%; Photovoltaic installation rises to 3185MW from 530.83MW, increases by 500%.In this area, honourable existing proportioning is 4.985:1, and the honourable proportioning after planning is 1.152:1, and solar energy industry needs the support of Correspondence policy badly.

Table 1 algorithms of different optimum results

It is as shown in table 2 that hierarchy optimization algorithm and the method for exhaustion optimize the time.The time that this project internal layer sequential production simulation simulation optimization is 1 time is 21 minutes, therefore adopts hierarchy optimization method than adopting the method for exhaustion and reduces 186291 minutes computing time, effectively meet and economize requirement computing time of (district) Electric Power Network Planning.

Table 2 algorithms of different solves the time and compares

The production simulation that have employed based on sequential due to internal layer emulates, and can carry out analysis and assessment, now research and analyse further honourable optimum proportioning case in planning process to generation of electricity by new energy ruuning situation under planning scene.Fig. 5 to ration the power supply rate distribution map for honourable population mean weekly, and can find that system is better than the heat supply phase in the wind-powered electricity generation of non-heat supply phase, photovoltaic ability of receiving, rate of rationing the power supply is less.This is the enhancing due to non-heat supply phase thermal power plant unit peak modulation capacity.Therefore the time stimulatiom method of model internal layer, can carry out management and running to unit in net by Various Seasonal, increase the receiving ability of wind energy and solar energy.

The optimization that the present invention is particularly useful for provincial power network generation of electricity by new energy installed capacity calculates, and economizes the new forms of energy installed capacity planning of (district) electrical network, low-carbon electric power requires lower net source planning and practical power systems to dispatch all to have important directive significance to China.

Claims (8)

1. a new forms of energy capacity ratio hierarchy optimization method in electrical network, the energy source in described electrical network comprises thermoelectricity and at least one new forms of energy, it is characterized in that, described optimization method comprises the following steps:

The installed capacity of all kinds of new forms of energy in step 1, initialization electrical network, as the initial value of outer Optimized model;

Step 2, the installed capacity input internal layer Optimized model of current all kinds of new forms of energy that outer Optimized model is obtained; Internal layer Optimized model, by solving with drag, obtains under the installed capacity of current all kinds of new forms of energy, the whole year operation state of the fired power generating unit making CO2 emissions F minimum:

$\min F=\underset{j=1}{\overset{N_j}{\&Sigma;}}\underset{t=1}{\overset{T}{\&Sigma;}}\{{\mathrm{\&alpha;}}_j{Y}_j^t+{\mathrm{\&beta;}}_j{Z}_j^t+a_j{P}_j^t+b_j\}\&CenterDot;\mathrm{\&gamma;}$

Be tied in:

$P_j^{t+1}-P_j^t\&le;{\mathrm{\&Delta;P}}_j^{up}$

$P_j^t-P_j^{t+1}\&le;{\mathrm{\&Delta;P}}_j^{down}$

$X_j^t\&CenterDot;P_j^{min}\&le;P_j^t\&le;X_j^t\&CenterDot;P_j^{max}$

$0\&le;Y_j^t+Z_j^t\&le;1$

$Y_j^t+\underset{i=1}{\overset{k}{\&Sigma;}}{Z}_j^{t+i}\&le;1$

$Z_j^t+\underset{i=1}{\overset{k}{\&Sigma;}}{Y}_j^{t+i}\&le;1$

$P_{BY}^t=C_j^b\&CenterDot;H_j^t$

$H_j^t\&CenterDot;C_j^b\&le;P_{CQ}^t\&le;P_j^{max}-H_j^t\&CenterDot;C_j^v$

$\underset{j=1}{\overset{N_j}{\&Sigma;}}{P}_j^t+P_0^t=P_1^t$

$-\underset{j=1}{\overset{N_j}{\&Sigma;}}{p}_j^{max}-C_N^t\&le;-P_1^t-S_p$

$\underset{j=1}{\overset{N_j}{\&Sigma;}}{P}_j^{min}+C_N^t\&le;P_1^t-S_N$

$0\&le;P_i^t\&le;N_i\&CenterDot;P_i^{t*},i=1,2,...,CL$

Wherein, for binary variable, represent the starting state of jth platform fired power generating unit t, 1 represents that unit starts, and 0 represents that unit is not at starting state; also be binary variable, represent jth platform fired power generating unit t stopped status, 1 represents that unit is shut down, and 0 represents that unit is not in stopped status; for the jth platform fired power generating unit of t is exerted oneself; and be independent variable; N _jrepresent the total number of units participating in optimizing fired power generating unit; T represents internal layer Optimized model simulation time length; α _jfor jth platform fired power generating unit opens machine-made egg-shaped or honey-comb coal briquets consumption; β _jfor jth platform fired power generating unit shuts down coal consumption; a _jfor the coal consumption of separate unit fired power generating unit is with changed power slope; b _jfor separate unit fired power generating unit coal consumption constant; γ is CO2 emission coefficient; be respectively swash ratio of slope and the lower climbing rate of jth platform fired power generating unit; be respectively minimum load value and the maximum output value of jth platform fired power generating unit; for binary variable, represent jth platform fired power generating unit t running status, 1 represents that unit runs, and 0 represents that unit does not run; K is the default minimum time step opening machine or shutdown; be respectively in the heat supply phase, in fired power generating unit, back pressure type thermal power plant unit and bleeder thermal power plant unit t exerts oneself; for t load of heat; for thermal power plant unit coupled thermomechanics coefficient; for total power load of t electrical network; for the electric power sum of all kinds of new forms of energy generations that t electrical network is received; S _p, S _nbe respectively the positive/negative spinning reserve of electrical network; for generation of electricity by new energy is at the credible capacity of day part; for the electric power of the i-th class new forms of energy generation that t electrical network is received; N _ifor the installed capacity of the i-th class new forms of energy in the electrical network that outer Optimized model inputs; for the i-th class new forms of energy long time scale in electrical network is exerted oneself seasonal effect in time series normalized value; CL is the classification sum of new forms of energy in electrical network;

Step 3, export the minimum CO2 emissions obtained to outer Optimized model;

Step 4, outer Optimized model judge whether to meet the stopping criterion for iteration preset, in this way, then by the installed capacity N of each new forms of energy minimum for CO2 emissions _iand the starting state of fired power generating unit stopped status exert oneself with fired power generating unit export as final optimum results, optimize and terminate; As no, then using the minimum CO2 emissions of internal layer Optimized model output as fitness function value, the installed capacity of all kinds of new forms of energy upgraded, then goes to step 2.

2. new forms of energy capacity ratio hierarchy optimization method in electrical network as claimed in claim 1, is characterized in that, use branch and bound method to solve described internal layer Optimized model.

3. new forms of energy capacity ratio hierarchy optimization method in electrical network as claimed in claim 1, it is characterized in that, described outer Optimized model uses particle cluster algorithm or the improvement bacterial foraging algorithm based on particle cluster algorithm.

4. new forms of energy capacity ratio hierarchy optimization method in electrical network as claimed in claim 1, it is characterized in that, the variable solution space constraint of described outer Optimized model is as follows:

$N_i^{min}\&le;{\mathrm{\&theta;}}_i^m\&le;N_i^{\max },i=1,2,...,CL$

In formula, θ represents that the variable that needs are optimized, m represent m the individuality of variable θ, represent the planning installed capacity maximum of the i-th class new forms of energy and existing installed capacity value respectively.

5. new forms of energy capacity ratio hierarchy optimization method in electrical network as claimed in claim 1, it is characterized in that, described default stopping criterion for iteration is the maximum iteration time of default outer Optimized model.

6. new forms of energy capacity ratio hierarchy optimization method in electrical network as claimed in claim 1, it is characterized in that, described new forms of energy are wind power generation and photovoltaic generation.

7. new forms of energy capacity ratio hierarchy optimization method in electrical network as claimed in claim 6, it is characterized in that, described generation of electricity by new energy is at the credible capacity of day part value be taken as day part new forms of energy exert oneself sum 60%-80% between.

8. new forms of energy capacity ratio hierarchy optimization method in electrical network as claimed in claim 1, it is characterized in that, described electrical network is provincial power network.