CN106532710A - Microgrid power flow optimization method considering voltage stability constraint - Google Patents

Abstract

The invention discloses a microgrid power flow optimization method considering voltage stability constraint. The method comprises a step of establishing a voltage stability index, a step of establishing the optimization mathematical model of a micro grid considering the influences of a distributed power supply and an energy storage unit, a step of constructing a historical fitting prediction-correction interior point method considering voltage stability constraint based on a prediction-correction interior point method, a step of repeatedly calling a three-phase power flow calculation module in a solution process to carry out three-phase power flow calculation on a microgrid optimization scheme, feeding the obtained voltage stability index to the historical fitting prediction-correction interior point method to carry out optimization processing. The method has the advantages that the historical fitting prediction-correction interior point method is used to process a discrete control variable, the contradiction between the calculation accuracy and iteration efficiency is solved, the method has a better global optimization characteristic, and a prediction-correction interior point method advantage is maintained.

Description

The micro-capacitance sensor tide optimization method of meter and Voltage Stability Constraints

Technical field

The present invention relates to a kind of tide optimization method of micro-capacitance sensor, more particularly to a kind of micro- electricity for considering Voltage Stability Constraints Net tide optimization method.

Background technology

Optimal load flow (Optimal Power Flow, OPF) is one of important means of Operation of Electric Systems and control, It is the important component part of electric energy management system.So-called optimal load flow, is exactly the structural parameters and load feelings when power system When condition gives, by the preferred of control variables, find and can meet all constraintss specified, and make system one or more Trend distribution when performance indications are optimal.With the continuous propulsion that intelligent grid is built, the regenerative resource of clean and effective Generation technology is of great interest, in the last few years, by various distributed power sources (Distributed Generation, DG), the micro-grid system that distributed energy storage unit (Storage Unit, SU), load and monitoring and protection device are assembled Permeability in power system is improved constantly, and its role is increasingly projected.Photovoltaic, wind-powered electricity generation distributed in micro-capacitance sensor Power supply has very strong intermittence, has had a strong impact on the safe operation of system.Correlative study shows that voltage stability is micro-capacitance sensor The important safety factor considered required for being incorporated into the power networks.For the running optimizatin problem of micro-capacitance sensor, net can be generally used Economy, the index such as security after network optimization weighing the quality of prioritization scheme, and the DG in these indexs and micro-capacitance sensor and The running status of SU is closely related.Therefore, in DG and the SU in a large number micro-capacitance sensor of infiltration, need badly and redesign the suitable of micro-capacitance sensor Operating index, optimized mathematical model and corresponding derivation algorithm.

Existing tide optimization method is difficult to the needs for meeting micro-capacitance sensor development well and building both at home and abroad at present, mainly Show：

1. most tide optimization models fail the impact of fully meter and distributed power source and energy-storage units.Although occurring in that The tide optimization method that some meters and DG affect, but the modeling to DG still shows coarse, and DG is considered as common power or " negative usually Load " processed, and it is equivalent not carry out classification, lacks the Science modeling to SU with fashion.

2. when micro-grid connection is run, voltage stability is affected larger by intermittent micro- source, although occurred in that some meters And the tide optimization method of voltage stability constraint, but traditional voltage stability index fails to take into full account the grid-connected of micro-capacitance sensor The effect of operation characteristic and DG, SU, so as to preferably assessment can not be done to voltage stability.

3. micro-capacitance sensor has obvious three-phase imbalance feature, in the index such as network loss and voltage of calculation optimization scheme, Existing many optimization methods fail the disequilibrium for taking into full account micro-capacitance sensor triphase parameter and load, so as to affect to calculate essence Degree.

4. the characteristics of there is micro-capacitance sensor Optimized model continuous variable and discrete variable to coexist, traditional mathematicses law of planning process from Scattered variable is more difficult, and global optimizing ability is not strong；Existing intelligent algorithm calculating speed is slow, and is easily absorbed in local optimum.

The content of the invention

The purpose of the present invention is exactly to solve the above problems, it is proposed that the micro-capacitance sensor tide of a kind of meter and Voltage Stability Constraints Flow-optimized method, the method has that meter and voltage stability, Optimized model be more accurate, calculating speed faster, global optimizing efficiency Higher the advantages of.

For achieving the above object, concrete scheme of the invention is as follows：

The micro-capacitance sensor tide optimization method of a kind of meter and Voltage Stability Constraints, comprises the following steps：

Step one：With reference to the characteristic that is incorporated into the power networks of micro-capacitance sensor, voltage stability index is set up；

Step 2；Optimization aim is up to load balancing degree so that the active loss of micro-capacitance sensor network is minimum respectively, sets up micro- The Mathematical Modeling of network optimization；

Step 3：Determine the constraints of micro-capacitance sensor optimized mathematical model, including：Power-balance constraint, node voltage are about Beam, reactive-load compensation capacitor compensation constraint, transformer capacity constraint, the injecting power constraint of DG, the power B of SU_iWith energy S_iAbout Beam and Voltage Stability Constraints；

Step 4：Optimize the discrete control variables in case to micro-capacitance sensor using the process of history matching method, obtain continuous Optimal objective variable；

Step 5：Based on predictor-corrector interior point method, by introducing optimal objective variable, node is carried out using AMD algorithms excellent Change numbering, construct the history matching predictor-corrector interior point method of meter and voltage stability constraint, micro-capacitance sensor optimization case is carried out Optimal load flow is calculated, and obtains voltage stability index, and calculated voltage stability index is fed back to history matching again Predictor-corrector interior point method is optimized process, obtains the optimal solution of micro-capacitance sensor optimized mathematical model.

Further, to the micro-grid system containing distributed power source and energy-storage units high permeability, pushed away using three phase fronts Weakly loops are converted into Radial network by back substitution power flow algorithm, by the method in breakpoint both sides injecting compensating amount come equivalent micro- electricity The impact of the contained loop of net；Simultaneously according to the interface mode of all kinds of distributed power sources and Power System Interconnection and their operation and Control mode, sets up its Equivalent Model in Load flow calculation respectively, changes distributed power source place node in each iteration step For PQ nodes, PI nodes or PV node.

Further, to the micro-grid system containing distributed power source and energy-storage units high permeability, the tool of Load flow calculation Body method is：

(1) initial data of micro-capacitance sensor is initialized, including microgrid topology information, component parameters, load data；

(2) form loop-impedance matrice Z；

(3) calculate the Injection Current of each load bus；

(4) determine the equivalent Injection Current of distributed power source and energy-storage units；

(5) whether there is looped network loop in judging network, to the node superposition Injection Current that unlinks if it there is looped network loop；

(6) the three-phase Injection Current of energy-storage units and distributed power source is superimposed, three phase fronts is carried out and is pushed back Load flow calculation；

(7) forward calculation is proceeded by from feeder terminal node, branch current is sued for peace, so as to obtain each bar branch road The three-phase current of top node；

(8) start to push back the voltage for calculating each branch road endpoint node from feeder line headend node, while the three-phase to each node Voltage is modified；

(9) reactive power of PV node injection is corrected, then judges whether the reactive power of PV node crosses the border, if sending out Life is crossed the border, and PV node is changed into PQ nodes proceeds to step (3) and re-start calculating, otherwise proceeds to (10)；

(10) whether default precision is met as the condition of convergence with the adjacent voltage amplitude of the iteration twice value difference of node, if meeting condition Then iteration convergence, proceeds to (11), otherwise corrects again the Injection Current of place node, then proceeds to step (3) and recalculate；

(11) calculate and terminate, export calculation of tidal current.

Further, in the step one, voltage stability index L_mgSpecially：

Wherein,To be connected with the voltage phasor of DG nodes；To be connected with the voltage phasor of SU nodes,For load bus electricity Pressure phasor；F_ijFor load participation factors；S_GTo be connected with the set of DG nodes；S_STo be connected with the set of SU nodes；S_LFor whole loads The set of node.

Further, in the step 2, the Mathematical Modeling of the micro-capacitance sensor optimization of foundation is specially：

Wherein, f₁For network active loss；f₂For branch road peak load rate；M={ A, B, C } is phase ordered sets；B is micro- electricity Net all set of fingers；P_ij.m、Q_ij.mAnd I_ij.mActive power, reactive power and the electric current of branch road ij are flow through respectively；R_ij.mFor The resistance of branch road ij；V_ij.mFor the terminal voltage of branch road ij；I_N.ij.mFor the rated current of branch road ij.

Further, in the step 3,

(1) power-balance constraint is specially：

To any node i in micro-capacitance sensor, following condition need to be met：

Wherein, P_iAnd Q_iThe respectively injection active power and reactive power of node i；P_loss、Q_lossRespectively micro-capacitance sensor Active loss and reactive loss；V_iAnd V_jThe respectively voltage magnitude of node i and node j；G_ijAnd B_ijRespectively branch admittance Y_ij Imaginary part and real part；θ_ijFor node i and the phase angle difference of node j；N is micro-capacitance sensor node set；

(2) node voltage constraint is specially：

V_i.min≤V_i≤V_i.maxi∈N；

Wherein, V_i.minAnd V_i.maxThe respectively lower voltage limit and the upper limit of node i；V_iFor the voltage of node；

(3) reactive-load compensation capacitor compensation constraint is specially：

Q_cri.min≤Q_cri≤Q_cri.max i∈N_C；

Wherein, Q_criFor the reactive-load compensation amount of reactive-load compensation capacitor；Q_cri.max、Q_cri.minFor compensation rate Q_criBound； N_CFor candidate compensation buses set.

Further, in the step 3,

(1) the injecting power constraint of DG is specially：

P_Gi.min≤P_Gi≤P_Gi.max i∈G_P；

Q_Gi.min≤Q_Gi≤Q_Gi.max i∈G_Q；

Wherein, P_GiAnd Q_GiThe active and idle output of respectively i-th DG；P_Gi.maxAnd P_Gi.minIt is respectively which corresponding Active power upper and lower limit；Q_Gi.maxAnd Q_Gi.minIt is respectively its corresponding reactive power upper and lower limit；G_PAnd G_QIt is respectively active adjustable With idle adjustable controllable DG set；

(2) the power B of SU_iWith energy S_iConstraint is specially：

P_charge≤B_i≤P_dischargei∈N；

S_low≤S_i≤S_highi∈N；

Wherein, P_dischargeFor SU maximum discharge powers；P_chargeFor SU maximum charge power；S_lowFor the least residue of SU Energy, S_highFor the maximum residual energy of SU；

(3) Voltage Stability Constraints condition is specially：

L_mgi.min≤L_mgi≤L_mgi.maxi∈N；

Wherein, L_mgFor voltage stability index, L_mgi.maxAnd L_mgi.minRespectively L_mgiBound.

Further, the concrete grammar of the step 4 is：

(4-1) discrete control variables to be adjusted is selected, determines position of the discrete control variables in micro-capacitance sensor；

(4-2) setup parameter method of adjustment, carries out automatic history matching；

(4-3) data analysis, if data results are unreasonable, returns the starting point of circulation, selects other adjustment areas The sample range in domain, the parameter for changing adjustment or modification parameter；If result is rationally, and parameter converges to special value, then enter Enter next step；

(4-4) optimal result fitting is got up, obtains a continuous optimal objective variable.

Further, the concrete grammar of the step 5 is：

(5-1) data initialization, reads in the underlying parameter and service data of micro-capacitance sensor, while given original variable and antithesis The initial value of variable, and ensure slack variable u, l >=0, Lagrange multiplier y ≠ 0, z >=0, w >=0, obstruction factor μ >=0 is arranged Primary iteration number of times k=0, maximum iteration time K_max, convergence difference ε₁、ε₂；

(5-2) needed for judging, whether adjustment control variables is discrete variable；If discrete variable, then history plan is carried out to which Continuous optimal objective variable is combined into, (5-3) is then proceeded to, (5-3) if continuous variable, is then directly proceeded to；

(5-3) using AMD algorithms according to the higher limit of node degree selecting to want numbered node principle, repair to improving interior point method The coefficient matrix of positive equation carries out node optimizing code；

(5-4) calculate maximum norm F of complementary clearance G ap and K-T conditions_max=max | | L_x| |, | | L_y| |, | | L_z| |, | |L_w||}；

(5-5) Gap ＜ ε are judged₁And F_max＜ ε₂Whether set up, directly proceed to (5-6) of being false, if setting up, output is optimum Solution terminates to calculate；

(5-6) calculate the Jacobin matrix of each class functionWith Hessian matrix

(5-7) solve the affine step-length of affine equation original variable and dual variableAnd under radiation direction Complementary clearance G ap^Aff；

(5-8) according to prediction Center Parameter σ=(Gap^Aff/Gap)³, calculate obstruction factor μ；

(5-9) iterative equation is solved, obtains iteration direction Δ x, Δ y, Δ u, Δ w, Δ l, Δ z；

(5-10) solve the iteration step length step of original variable and dual variable_d、step_d, update original variable with to mutation Amount；

(5-11) judge current iteration number of times k whether less than maximum iteration time k_max, k=k+1 is put if meeting and proceed to (5-4), it is unsatisfactory for, " output is calculated and do not restrained " simultaneously terminates calculating.

Beneficial effects of the present invention：

1. the voltage stability index set up can more effectively assess the voltage stability of each node of micro-grid system.

2. the multiple target multiconstraint optimization model constructed by can count simultaneously and distributed power source and energy-storage units impact.

3. many power supply three-phase Forward and backward substitution methods for being adopted can consider the three-phase imbalance feature of micro-capacitance sensor well, And calculating speed is fast；

4. the history matching predictor-corrector interior point method for being adopted can be preferably processed to discrete control variables, solved Contradiction between computational accuracy and iteration efficiency, with more preferable global optimizing feature, and maintains predictor-corrector interior point method Advantage.

Description of the drawings

Overall design drawings of the Fig. 1 for micro-capacitance sensor tide optimization；

Fig. 2 is feeder line model；

Fig. 3 is micro-capacitance sensor three-phase power flow flow chart；

Fig. 4 is discrete control variables history matching flow chart.

Fig. 5 is history matching predictor-corrector interior point method flow chart.

Specific embodiment：

The present invention is described in detail below in conjunction with the accompanying drawings：

The micro-capacitance sensor tide optimization method of meter and Voltage Stability Constraints, it comprises the concrete steps that：As shown in figure 1, tying first The characteristic that is incorporated into the power networks of micro-capacitance sensor is closed, voltage stability index L is set up_mg；Then the shadow of meter and distributed power source and energy-storage units Sound sets up multiple target, the multiconstraint optimization Mathematical Modeling of micro-capacitance sensor；Obtained by introducing history matching based on predictor-corrector interior point method To continuous optimal objective variable and carry out node optimizing code using AMD algorithms, construct meter and voltage stability constraint History matching predictor-corrector interior point method；Three-phase power flow module is called repeatedly in solution process to micro-capacitance sensor prioritization scheme Three-phase power flow is carried out, voltage stability index L obtained from entering_mg, then feed back to history matching predictor-corrector interior point method and enter Row optimization processing.

Fig. 2 is feeder line model, wherein, I_iFor Injection Current from branch road ij to node i, Y_ijFor the admittance of branch road ij, I_LiFor The load side Injection Current of node i, I_GiAnd I_SiRespectively charging and discharging currents of the Injection Current and SU of the connect DG of node i, I_i.1 And I_i.kRespectively the 1st of node i goes out to prop up and kth goes out the output current propped up.

Below the inventive method is described in detail：

1. the voltage stability index L for being adapted to micro-grid connection operation characteristic is set up_mg：

In formula：To be connected with the voltage phasor of DG nodes；To be connected with the voltage phasor of SU nodes,For load bus electricity Pressure phasor；F_ijFor load participation factors；S_GTo be connected with the set of DG nodes；S_STo be connected with the set of SU nodes；S_LFor whole loads The set of node.

2. the Mathematical Modeling of micro-capacitance sensor optimization is set up, and the object function in model there are two, and respectively micro-capacitance sensor network has Work(is lost minimum and load balancing degree highest, and expression formula is：

In formula：f₁For network active loss；f₂For branch road peak load rate；M={ A, B, C } is phase ordered sets；B is micro- electricity Net all set of fingers；P_ij.m、Q_ij.mAnd I_ij.mActive power, reactive power and the electric current of branch road ij are flow through respectively；R_ij.mFor The resistance of branch road ij；V_ij.mFor the terminal voltage of branch road ij；I_N.ij.mFor the rated current of branch road ij.

The constraints of micro-capacitance sensor optimized mathematical model includes：

1) power-balance constraint, to any node i in micro-capacitance sensor, need to meet following condition：

In formula：P_iAnd Q_iThe respectively injection active power and reactive power of node i；P_loss、Q_lossRespectively micro-capacitance sensor Active loss and reactive loss；V_iAnd V_jThe respectively voltage magnitude of node i and node j；G_ijAnd B_ijRespectively branch admittance Y_ij Imaginary part and real part；θ_ijFor node i and the phase angle difference of node j；N is micro-capacitance sensor node set.

2) node voltage constraint, i.e., the voltage V of each node_iShould meet：

V_i.min≤V_i≤V_i.max i∈N (6)

In formula：V_i.minAnd V_i.maxThe respectively lower voltage limit and the upper limit of node i.

3) reactive-load compensation capacitor compensation constraint：

Q_cri.min≤Q_cri≤Q_cri.max i∈N_C (7)

In formula：Q_criFor the reactive-load compensation amount of reactive-load compensation capacitor；Q_cri.max、Q_cri.minFor compensation rate Q_criBound； N_CFor candidate compensation buses number.

4) transformer capacity constraint：

|S_t|≤S_N.t t∈T (8)

In formula：S_tAnd S_N.tThe respectively actual power and rated capacity of transformer t；T is the set of all transformers.

5) injecting power of DG is constrained to：

P_Gi.min≤P_Gi≤P_Gi.max i∈G_P (9)

Q_Gi.min≤Q_Gi≤Q_Gi.max i∈G_Q (10)

In formula：P_GiAnd Q_GiThe active and idle output of respectively i-th DG；P_Gi.maxAnd P_Gi.minIt is respectively which corresponding Active power upper and lower limit；Q_Gi.maxAnd Q_Gi.minIt is respectively its corresponding reactive power upper and lower limit；G_PAnd G_QIt is respectively active adjustable With idle adjustable controllable DG set.

6) the power B of SU_iWith energy S_iConstraint, i.e. Optimized model need to meet the physical limits such as the power and energy of SU operations：

P_charge≤B_i≤P_discharge i∈N-1 (11)

S_low≤S_i≤S_high i∈N (12)

In formula：P_dischargeFor SU maximums discharge power (positive number)；P_chargeFor SU maximum charge power (negative)；S_lowFor The least residue energy of SU, S_highFor the maximum residual energy of SU.

7) Voltage Stability Constraints condition, based on the voltage stability index L set up in formula (1)_mg, micro-grid connection is run The voltage stability constraints that should be met is expressed as：

L_mgi.min≤L_mgi≤L_mgi.max i∈N (13)

In formula：L_mgi.maxAnd L_mgi.minRespectively L_mgiBound；L_mgiThe span of index is 0～1, when the index When being close to 0, represent voltage stability preferably, and when index is close to 1, system operating point will be close to collapse of voltage point.Therefore The condition of global voltage stabilization is that the index that all load points are calculated is respectively less than 1.

3., for the micro-capacitance sensor of DG and SU high permeabilities, three phase fronts of design are pushed back for power flow algorithm.

Flow chart as shown in figure 3, the micro-capacitance sensor network prioritization scheme of high permeability to DG and SU, using Forward and backward substitution method Computing system trend, in view of the algorithm is weaker to the disposal ability of mesh, this module is improved to forward-backward sweep method, by weak ring Net is converted into Radial network, by the method in breakpoint both sides injecting compensating amount come the impact of equivalent loop.Additionally, according to each The interface mode of class distributed power source (wind-powered electricity generation, photovoltaic, fuel cell, miniature gas turbine etc.) and Power System Interconnection and they Operation and control mode, set up its Equivalent Model in Load flow calculation respectively, distributed power source are located in each iteration step and save Point is converted to PQ nodes, PI nodes or PV node.In calculating process, each element is participated in using the triphase flow in model library Calculate.Concrete grammar is：

(1) micro-capacitance sensor data initialization is carried out first, reads grid parameter and load parameter；

(2) form loop (loop) impedance matrix Z；

(3) Injection Current of each load bus is calculated according to the data read in；

(4) it is calculated as follows the equivalent Injection Current of DG and SU：

To i-th DG, the Injection Current phasor of its m phaseIt is expressed as：

In formula：G_PQThe set that the DG (i.e. PQ types) given by power is constituted；G_PVFor the DG that active and voltage magnitude gives The set constituted by (i.e. PV types)；G_PIThe set that the DG (i.e. PI types) given by active and current amplitude is constituted；P_gi.m、 Q_gi.m, andThe respectively active power of the m phases of i-th DG, reactive power and terminal voltage；For the node of PV types DG Resolve the difference of voltage and given voltage；Z_i.mFor the branch impedance sum that PV types DG are connected with source node；I_gi0.mFor PI types DG The given electric current of m phases；

To j-th SU, Injection Current amplitude I of its m phase_j.mIt is expressed as：

In formula：S_IAnd S_VThe set constituted respectively by current constant mode and by the SU of constant voltage mode energy storage；I_0.jFor energy storage The given charging current of unit；V_j.mAnd E_j.mThe respectively charging voltage and built-in potential of the m phases of energy-storage units j；R_jFor energy storage The charge circuit resistance of unit j.

(5) whether there is looped network loop in judging network, to the node superposition Injection Current that unlinks if it there is looped network loop；

(6) the three-phase Injection Current of DG and SU is superimposed, three phase fronts is carried out and is pushed back Load flow calculation；

(7) forward calculation is proceeded by from feeder terminal node, branch current is sued for peace, so as to obtain each bar branch road The three-phase current of top node；

(8) start to push back the voltage for calculating each branch road endpoint node from feeder line headend node, while the three-phase to each node Voltage is modified；

(9) reactive power of PV node injection is corrected, then judges whether the reactive power of PV node crosses the border, if sending out Life is crossed the border, and PV node is changed into PQ nodes proceeds to step (3) and re-start calculating, otherwise proceeds to (10)；

(10) whether default precision is met as the condition of convergence with the adjacent voltage amplitude of the iteration twice value difference of node, if meeting condition Then iteration convergence, proceeds to (11), otherwise corrects again the Injection Current of place node, then proceeds to step (3) and recalculate；

(11) calculate and terminate, export calculation of tidal current.

4. the discrete control variables in pair micro-capacitance sensor optimization case is obtained continuous target and is become using the process of history matching method Amount.Flow chart is as shown in Figure 4.Concrete steps include：

(1) discrete control variables to be adjusted is selected, is further appreciated that its position in micro-capacitance sensor and how to these changes Amount is adjusted；

(2) selection parameter method of adjustment, the present invention choose neighborhood method (a kind of random sampling algorithms)；

(3) automatic history matching (comparative observation data and prediction data, calculating misfit value, control undated parameter, up to Till data are coincide)；

(4) data analysis, if data results are unreasonable, returns the starting point of circulation, select other adjustment regions, Change the sample range of the parameter or modification parameter of adjustment；If result is rationally, and parameter converges to special value, then go to step (5)。

(5) optimal result fitting is got up, obtains a continuous optimal objective variable.

5. predictor-corrector interior point method is based on, by introducing continuous optimal objective variable and the utilization that history matching is obtained AMD algorithms carry out node optimizing code, to improve convergence of algorithm speed and global optimizing, the algorithm are used to solve micro- electricity The optimal power flow problems of net.Flow chart is as shown in Figure 5.Concrete steps include：

(1) data initialization, reads in the underlying parameter and service data of micro-capacitance sensor, at the same given original variable with to mutation The initial value of amount, and ensure slack variable u, l >=0, Lagrange multiplier y ≠ 0, z >=0, w >=0, obstruction factor μ >=0 is arranged just Beginning iterations k=0, maximum iteration time K_max, convergence difference ε₁、ε₂；

(2) needed for judging, whether adjustment control variables is discrete variable；If discrete variable, then history matching is carried out to which For continuous optimal objective variable, (3) are then proceeded to, (3) if continuous variable, are then directly proceeded to；

(3) using the higher limit according to node degree of AMD algorithms selecting to want numbered node principle, repair to improving interior point method The coefficient matrix of positive equation carries out node optimizing code；

(4) calculate maximum norm F of complementary clearance G ap and K-T conditions_max=max | | | L_x| |, | | L_y| |, | | L_z| |, | | L_w||}；

(5) Gap ＜ ε are judged₁And F_max＜ ε₂Whether set up, directly proceed to (6) of being false, optimum is exported if setting up and is unhitched Beam is calculated；

(6) calculate the Jacobin matrix of each class functionWith Hessian matrix

(7) solve the affine step-length of affine equation original variable and dual variableAnd it is mutual under radiation direction Mend clearance G ap^Aff；

(8) according to prediction Center Parameter σ=(Gap^Aff/Gap)³, (μ=σ × Gap/2m, m are to calculate obstruction factor μ The number of formula equation)；

(9) iterative equation is solved, obtains iteration direction Δ x, Δ y, Δ u, Δ w, Δ l, Δ z；

(10) solve the iteration step length step of original variable and dual variable_d、step_d, update original variable with to mutation Amount；

(11) judge current iteration number of times k whether less than maximum iteration time k_max, k=k+1 is put if meeting and proceed to (4), it is unsatisfactory for, " output is calculated and do not restrained " simultaneously terminates calculating.

Although the above-mentioned accompanying drawing that combines is described to the specific embodiment of the present invention, not to present invention protection model The restriction enclosed, one of ordinary skill in the art should be understood that on the basis of technical scheme those skilled in the art are not The various modifications made by needing to pay creative work or deformation are still within protection scope of the present invention.

Claims (9)

1. the micro-capacitance sensor tide optimization method of a kind of meter and Voltage Stability Constraints, is characterized in that, comprise the following steps：

Step one：With reference to the characteristic that is incorporated into the power networks of micro-capacitance sensor, voltage stability index is set up；

Step 2；Optimization aim is up to load balancing degree so that the active loss of micro-capacitance sensor network is minimum respectively, sets up micro-capacitance sensor The Mathematical Modeling of optimization；

Step 3：Determine the constraints of micro-capacitance sensor optimized mathematical model, including：Power-balance constraint, node voltage constraint, nothing The compensation constraint of work(compensation capacitor, transformer capacity constraint, the injecting power constraint of DG, the power B of SU_iWith energy S_iConstraint with And Voltage Stability Constraints；

Step 4：Optimize the discrete control variables in case to micro-capacitance sensor using the process of history matching method, obtain continuous optimum Target variable；

Step 5：Based on predictor-corrector interior point method, by introducing optimal objective variable, node optimization volume is carried out using AMD algorithms Number, the history matching predictor-corrector interior point method of meter and voltage stability constraint is constructed, optimum is carried out to micro-capacitance sensor optimization case Load flow calculation, obtain voltage stability index, and by calculated voltage stability index feed back to again history matching prediction- Correction interior point is optimized process, obtains the optimal solution of micro-capacitance sensor optimized mathematical model.

2. the micro-capacitance sensor tide optimization method of meter as claimed in claim 1 and Voltage Stability Constraints, is characterized in that, divide to containing The micro-grid system of cloth power supply and energy-storage units high permeability, is pushed back using three phase fronts and is converted into weakly loops for power flow algorithm Radial network, by the method in breakpoint both sides injecting compensating amount come the impact of the loop contained by equivalent micro-capacitance sensor；While root According to all kinds of distributed power sources and the interface mode and their operation and control mode of Power System Interconnection, which is set up respectively in trend Distributed power source place node is converted to PQ nodes, PI nodes or PV node in each iteration step by the Equivalent Model in calculating.

3. the micro-capacitance sensor tide optimization method of meter as claimed in claim 2 and Voltage Stability Constraints, is characterized in that, divide to containing The micro-grid system of cloth power supply and energy-storage units high permeability, the concrete grammar of Load flow calculation is：

(1) initial data of micro-capacitance sensor is initialized, including microgrid topology information, component parameters, load data；

(2) form loop-impedance matrice Z；

(3) calculate the Injection Current of each load bus；

(4) determine the equivalent Injection Current of distributed power source and energy-storage units；

(5) whether there is looped network loop in judging network, to the node superposition Injection Current that unlinks if it there is looped network loop；

(6) the three-phase Injection Current of energy-storage units and distributed power source is superimposed, three phase fronts is carried out and is pushed back Load flow calculation；

(7) forward calculation is proceeded by from feeder terminal node, branch current is sued for peace, so as to obtain each bar branch road top The three-phase current of node；

(8) start to push back the voltage for calculating each branch road endpoint node from feeder line headend node, while the three-phase voltage to each node It is modified；

(9) reactive power of PV node injection is corrected, then judges whether the reactive power of PV node crosses the border, if getting over PV node is then changed into PQ nodes and proceeds to step (3) and re-starts calculating by boundary, otherwise proceeds to (10)；

(10) whether default precision is met as the condition of convergence with the adjacent voltage amplitude of the iteration twice value difference of node, is changed if condition is met Withhold and hold back, proceed to (11), otherwise correct again the Injection Current of place node, then proceed to step (3) and recalculate；

(11) calculate and terminate, export calculation of tidal current.

4. the micro-capacitance sensor tide optimization method of meter as claimed in claim 1 and Voltage Stability Constraints, is characterized in that, the step In one, voltage stability index L_mgSpecially：

$\begin{array}{cc}{L}_{mg}=max|1-\frac{F_{ij}(\underset{i\∈S_G}{\mathrm{\Σ}}{\stackrel{\·}{V}}_i+\underset{k\∈S_S}{\mathrm{\Σ}}{\stackrel{\·}{V}}_k)}{{\stackrel{\·}{V}}_j}|& j\∈S_L\end{array};$

Wherein,To be connected with the voltage phasor of DG nodes；To be connected with the voltage phasor of SU nodes,For load bus voltage phase Amount；F_ijFor load participation factors；S_GTo be connected with the set of DG nodes；S_STo be connected with the set of SU nodes；S_LFor whole load buses Set.

5. the micro-capacitance sensor tide optimization method of meter as claimed in claim 1 and Voltage Stability Constraints, is characterized in that, the step In two, the Mathematical Modeling of the micro-capacitance sensor optimization of foundation is specially：

$\min f_1=\underset{(i,j)\∈B}{\Σ}\underset{m\∈M}{\Σ}{R}_{ij.m}\frac{P_{ij.m}^2+Q_{ij.m}^2}{V_{ij.m}^2};$

$\min f_2=\underset{(i,j)\∈B,m\∈M}{max}\frac{I_{ij.m}}{I_{N.ij.m}};$

Wherein, f₁For network active loss；f₂For branch road peak load rate；M={ A, B, C } is phase ordered sets；B is micro-capacitance sensor institute There is set of fingers；P_ij.m、Q_ij.mAnd I_ij.mActive power, reactive power and the electric current of branch road ij are flow through respectively；R_ij.mFor branch road The resistance of ij；V_ij.mFor the terminal voltage of branch road ij；I_N.ij.mFor the rated current of branch road ij.

6. the micro-capacitance sensor tide optimization method of meter as claimed in claim 1 and Voltage Stability Constraints, is characterized in that, the step In three,

(1) power-balance constraint is specially：

To any node i in micro-capacitance sensor, following condition need to be met：

$\underset{i\∈N}{\Σ}{P}_i=P_{loss}=\underset{i\∈N}{\Σ}{V}_i\underset{j\∈i}{\Σ}{V}_j\left(G_{ij}{\mathrm{cos\θ}}_{ij}\right);$

$\underset{i\∈N}{\Σ}{Q}_i=Q_{loss}=-\underset{i\∈N}{\Σ}{V}_i\underset{j\∈i}{\Σ}{V}_j\left(B_{ij}{\mathrm{cos\θ}}_{ij}\right);$

Wherein, P_iAnd Q_iThe respectively injection active power and reactive power of node i；P_loss、Q_lossRespectively micro-capacitance sensor is active Loss and reactive loss；V_iAnd V_jThe respectively voltage magnitude of node i and node j；G_ijAnd B_ijRespectively branch admittance Y_ijVoid Portion and real part；θ_ijFor node i and the phase angle difference of node j；N is micro-capacitance sensor node set；

(2) node voltage constraint is specially：

V_i.min≤V_i≤V_i.maxi∈N；

Wherein, V_i.minAnd V_i.maxThe respectively lower voltage limit and the upper limit of node i；V_iFor the voltage of node；

(3) reactive-load compensation capacitor compensation constraint is specially：

Q_cri.min≤Q_cri≤Q_cri.max i∈N_C；

Wherein, Q_criFor the reactive-load compensation amount of reactive-load compensation capacitor；Q_cri.max、Q_cri.minFor compensation rate Q_criBound；N_CFor Candidate compensation buses set.

7. the micro-capacitance sensor tide optimization method of meter as claimed in claim 1 and Voltage Stability Constraints, is characterized in that, the step In three,

(1) the injecting power constraint of DG is specially：

P_Gi.min≤P_Gi≤P_Gi.max i∈G_P；

Q_Gi.min≤Q_Gi≤Q_Gi.max i∈G_Q；

Wherein, P_GiAnd Q_GiThe active and idle output of respectively i-th DG；P_Gi.maxAnd P_Gi.minIt is respectively its corresponding wattful power Rate upper and lower limit；Q_Gi.maxAnd Q_Gi.minIt is respectively its corresponding reactive power upper and lower limit；G_PAnd G_QRespectively it is active reconcile it is idle Adjustable controllable DG set；

(2) the power B of SU_iWith energy S_iConstraint is specially：

P_charge≤B_i≤P_dischargei∈N；

S_low≤S_i≤S_highi∈N；

Wherein, P_dischargeFor SU maximum discharge powers；P_chargeFor SU maximum charge power；S_lowFor the least residue energy of SU, S_highFor the maximum residual energy of SU；

(3) Voltage Stability Constraints condition is specially：

L_mgi.min≤L_mgi≤L_mgi.maxi∈N；

Wherein, L_mgFor voltage stability index, L_mgi.maxAnd L_mgi.minRespectively L_mgiBound.

8. the micro-capacitance sensor tide optimization method of meter as claimed in claim 1 and Voltage Stability Constraints, is characterized in that, the step Four concrete grammar is：

(4-1) discrete control variables to be adjusted is selected, determines position of the discrete control variables in micro-capacitance sensor；

(4-2) setup parameter method of adjustment, carries out automatic history matching；

(4-3) data analysis, if data results are unreasonable, returns the starting point of circulation, selects other adjustment regions, changes Become the sample range of the parameter or modification parameter of adjustment；If result is rationally, and parameter converges to special value, then into next Step；

(4-4) optimal result fitting is got up, obtains a continuous optimal objective variable.

9. the micro-capacitance sensor tide optimization method of meter as claimed in claim 1 and Voltage Stability Constraints, is characterized in that, the step Five concrete grammar is：

(6-1) data initialization, reads in the underlying parameter and service data of micro-capacitance sensor, while given original variable and dual variable Initial value, and ensure slack variable u, l >=0, Lagrange multiplier y ≠ 0, z >=0, w >=0, obstruction factor μ >=0 arranges initial Iterations k=0, maximum iteration time K_max, convergence difference ε₁、ε₂；

(6-2) needed for judging, whether adjustment control variables is discrete variable；If discrete variable, then carrying out history matching to which is Continuous optimal objective variable, then proceeds to (6-3), if continuous variable, then directly proceeds to (6-3)；

(6-3) using AMD algorithms according to the higher limit of node degree selecting to want numbered node principle, to improving interior point method amendment side The coefficient matrix of journey carries out node optimizing code；

(6-4) calculate maximum norm F of complementary clearance G ap and K-T conditions_max=max | | L_x| |, | | L_y| |, | | L_z| |, | | L_w| |}；

(6-5) Gap ＜ ε are judged₁And F_max＜ ε₂Whether set up, directly proceed to (6-6) of being false, optimum is exported if setting up and is unhitched Beam is calculated；

(6-6) calculate the Jacobin matrix of each class functionWith Hessian matrix

(6-7) solve the affine step-length of affine equation original variable and dual variableAnd the complementation under radiation direction Clearance G ap^Aff；

(6-8) according to prediction Center Parameter σ=(Gap^Aff/Gap)³, calculate obstruction factor μ；

(6-9) iterative equation is solved, obtains iteration direction Δ x, Δ y, Δ u, Δ w, Δ l, Δ z；

(6-10) solve the iteration step length step of original variable and dual variable_d、step_d, update original variable and dual variable；

(6-11) judge current iteration number of times k whether less than maximum iteration time k_max, k=k+1 is put if meeting and proceed to (6- 4), it is unsatisfactory for, " output is calculated and do not restrained " simultaneously terminates calculating.