CN109274116A - A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method - Google Patents

A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method Download PDF

Info

Publication number
CN109274116A
CN109274116A CN201811064879.0A CN201811064879A CN109274116A CN 109274116 A CN109274116 A CN 109274116A CN 201811064879 A CN201811064879 A CN 201811064879A CN 109274116 A CN109274116 A CN 109274116A
Authority
CN
China
Prior art keywords
micro
capacitance sensor
node
direct current
exchange
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201811064879.0A
Other languages
Chinese (zh)
Inventor
彭寒梅
苏敏
刘聪
王迪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xiangtan University
Original Assignee
Xiangtan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xiangtan University filed Critical Xiangtan University
Priority to CN201811064879.0A priority Critical patent/CN109274116A/en
Publication of CN109274116A publication Critical patent/CN109274116A/en
Pending legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/381—Dispersed generators
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/02—Circuit arrangements for ac mains or ac distribution networks using a single network for simultaneous distribution of power at different frequencies; using a single network for simultaneous distribution of ac power and of dc power
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/04—Circuit arrangements for ac mains or ac distribution networks for connecting networks of the same frequency but supplied from different sources
    • H02J3/06—Controlling transfer of power between connected networks; Controlling sharing of load between connected networks
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/388—Islanding, i.e. disconnection of local power supply from the network

Abstract

The invention discloses a kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation methods: forming 2 kinds of alternating current-direct current mixing micro-capacitance sensor operating system types and 9 kinds of node types, solve the initial point of AC and DC micro-capacitance sensor subsystem and alternating current-direct current interconnection micro-capacitance sensor continuous tide respectively using band nonmonotone line search LM (Levenberg-Marquardt with a nonmonotone line search, LMNL) algorithm and the alternating iteration method based on LMNL algorithm；Predict that bearing calibration carries out alternating current-direct current interconnection micro-capacitance sensor prediction correction calculating using the bidirectional iteration for combining tangential method to predict and combine Newton method correction.The conventional Load Flow method for solving based on LMNL algorithm proposed, the conventional Load Flow suitable for different type micro-capacitance sensor calculate；The bidirectional iteration of it is proposed predicts bearing calibration, solves the problems, such as alternating current-direct current power there are coupled relation and the varying loading complicated variety that continuous tide calculates under different AC and DC sides.The continuous tide that the method for the present invention is suitable for alternating current-direct current mixing micro-capacitance sensor under the various methods of operation calculates, and has good engineering application value.

Description

A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method
Technical field
The present invention relates to Power System Analysis fields, calculate more particularly to a kind of alternating current-direct current mixing micro-capacitance sensor continuous tide Method.
Background technique
Micro-capacitance sensor presses the power supply electrical energy form of master network, can be divided into exchange micro-capacitance sensor, direct-current grid and alternating current-direct current mixing Micro-capacitance sensor.Alternating current-direct current mixing micro-capacitance sensor refers to containing AC and DC bus and connects the connection-transformation device of AC and DC bus (interlinking converter, ILC), not only can directly to AC load power supply again can directly to DC load power supply Micro-capacitance sensor.It is the power network simultaneously containing exchange micro-capacitance sensor and direct-current grid that some scholars, which define alternating current-direct current mixing micro-capacitance sensor, Network.With single power-supplying forms exchange micro-capacitance sensor and direct-current grid compares, alternating current-direct current mixing micro-capacitance sensor have it is various Characteristic and advantage.Alternating current-direct current mixing micro-capacitance sensor will be Efficient Development and the preferred micro-capacitance sensor mould for utilizing distribution type renewable energy Formula, and alternating current-direct current mixing micro-capacitance sensor is the important composition form of the following distribution system.Therefore, carry out alternating current-direct current mixing micro-capacitance sensor And its correlative study to the utilization efficiency for promoting China's distribution type renewable energy and promotes electric power development in science and technology with important Theoretical and realistic meaning.
Continuation method is produced into Continuation Method in conjunction with electric system conventional Load Flow, it is with containing parameter trend side The known solution point of certain of journey is initial point, tracks the process for the solution curve that it is mapped out in hyperspace.Continuous tide is to calculate The effective tool of power system steady state voltage stability limit point；Meanwhile conventional Load Flow equation Jacobian matrix can also be overcome unusual Brought dyscalculia is to solve the problems, such as one of method of Abnormal Load Flow of Power Systems.Distributed generation resource (distributed Generator, DG) access so that alternating current-direct current mixing micro-capacitance sensor becomes a complicated more electric power networks, and due to it The particularity of internal element and its close to the characteristic of power load, structure is more fragile compared with bulk power grid, so as to cause new Voltage's distribiuting and Voltage-stabilizing Problems；In addition, method of operation multiplicity, AC/DC power can compared to traditional exchange micro-capacitance sensor It is proposed newly with coordination and interaction and the limitation of connection-transformation device transimission power to the stable operation of alternating current-direct current mixing micro-capacitance sensor system Challenge.Therefore, alternating current-direct current mixing micro-capacitance sensor continuous tide, which calculates, has good research significance and application value, can be system Static frequency Voltage Stability Analysis and control provide effective tool, and the conventional tide of the alternating current-direct current mixing micro-capacitance sensor being directed to Stream calculation is the basis of the analysis of alternating current-direct current mixing micro-capacitance sensor systematic steady state, planning and operation study.
Currently, the research of alternating current-direct current mixing micro-capacitance sensor also in the desk study stage, it is studied both at home and abroad with practice according to It is so insufficient.Has the control that overall operation control model, connection-transformation device are focused primarily upon to the research of alternating current-direct current mixing micro-capacitance sensor It system, running Optimization and distributes rationally etc..However, both at home and abroad to the Continuation Power Flow Model of alternating current-direct current mixing micro-capacitance sensor with The research of derivation algorithm is few.
Alternating current-direct current mixing micro-capacitance sensor is by exchange micro-capacitance sensor subsystem and accesses public network, has the mode that is incorporated into the power networks and isolated island fortune Row mode.In conjunction with connection-transformation device connection, alternating current-direct current mixing micro-capacitance sensor has 4 kinds of basic methods of operation: joint is grid-connected Mode is handed over and directly from mode, joint off-network mode and difference off-network mode.6 kinds of operation systems can be formed under 4 kinds of different running methods System: grid-connected/isolated island alternating current-direct current interconnection micro-capacitance sensor, grid-connected/isolated island exchange micro-capacitance sensor subsystem, grid-connected/isolated island DC micro-electric net System；The Baseline Control Strategy that micro-capacitance sensor uses mainly has master & slave control, equity control and integrated control strategy；In addition, grid-connected/ Isolated island alternating current-direct current interconnect the power in micro-capacitance sensor between alternating current-direct current subsystem can two-way flow, each subsystem can independent control Coordinated control, the alternating current-direct current mixing micro-capacitance sensor method of operation is versatile and flexible as a result, can be using not under certain basic method of operation Same control strategy.It mainly includes that initial point calculates, prediction link calculates and correction link calculates 3 steps that continuous tide, which calculates, Suddenly, the essence that initial point solves is that conventional Load Flow calculates.The conventional Load Flow that micro-capacitance sensor is exchanged under no balance nodes calculates and contains DG AC distribution net there is difference substantially, conventional electrical distribution net conventional Load Flow algorithm and its deformation will be no longer applicable in；Without balance Not preassigned, the unknown state variable of the steady frequency of node exchange micro-capacitance sensor, and system frequency can be with load Parameters variation and change, frequency departure rated value will cause system crash when excessive, then its continuous tide need to calculate power versus frequency Curve；In addition, in alternating current-direct current interconnection micro-capacitance sensor when ILC is the power supply of a subsystem, while being also another subsystem Load, alternating current-direct current power is calculated there are coupled relation, the initial point of continuous tide and correction link calculates and is different from single confession The micro-capacitance sensor of electric form.The continuous tide of alternating current-direct current mixing micro-capacitance sensor calculates complicated multiplicity under different running method as a result,.
Summary of the invention
According to above background technique, the present invention proposes a kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method: shape At 2 kinds of alternating current-direct current mixing micro-capacitance sensor operating system types and 9 kinds of node types；Using band nonmonotone line search LM (Levenberg-Marquardt with a nonmonotone line search, LMNL) algorithm solves the micro- electricity of AC and DC The initial point of net continuous tide solves alternating current-direct current using the alternating iteration method based on LMNL algorithm and interconnects micro-capacitance sensor continuous tide Initial point；Predict that link uses localized parameterization method and tangential method, correction link using hypersphere parametric method and Newton method is combined, the prediction correction link for forming AC and DC micro-capacitance sensor continuous tide calculates and predicts school based on bidirectional iteration The prediction correction link of the alternating current-direct current interconnection micro-capacitance sensor continuous tide of correction method calculates.
The technical solution of the present invention is as follows:
A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method, including form 2 kinds of alternating current-direct current mixing micro-capacitance sensor operations System type and 9 kinds of node types；LMNL algorithm solves AC and DC micro-capacitance sensor continuous tide initial point and based on LMNL algorithm Alternating iteration method solves alternating current-direct current and interconnects micro-capacitance sensor continuous tide initial point；Prediction link is using localized parameterization method and cuts It is pre- to form AC and DC micro-capacitance sensor continuous tide using hypersphere parametric method and combination Newton method for line method, correction link Correction link is surveyed to calculate and predict that the alternating current-direct current interconnection micro-capacitance sensor continuous tide of bearing calibration predicts corrector loop based on bidirectional iteration Section calculates.
Wherein, 2 kinds of alternating current-direct current mixing micro-capacitance sensor operating system types of the formation and 9 kinds of node types, specific as follows:
Alternating current-direct current mixing micro-capacitance sensor is by exchange micro-capacitance sensor subsystem and accesses public network, has the mode that is incorporated into the power networks and isolated island fortune Row mode.Incorporating interconnecting converter connection, alternating current-direct current mixing micro-capacitance sensor have 4 kinds of basic methods of operation: combining grid-connected side Formula (method of operation 1) is handed over and directly from mode (method of operation 2), joint off-network mode (method of operation 3) and difference off-network mode (method of operation 4).Whether interconnected according to AC and DC micro-capacitance sensor subsystem, 4 kinds of differences of alternating current-direct current mixing micro-capacitance sensor are run substantially The operating system formed under mode is divided into 2 seed types.Alternating current-direct current mixing micro-capacitance sensor operating system type I: alternating current-direct current interconnects micro- electricity Net.Alternating current-direct current mixing micro-capacitance sensor operating system Type II: the exchange micro-capacitance sensor subsystem and DC micro-electric net system not interconnected. Alternating current-direct current mixing micro-capacitance sensor is type I under the basic method of operation 1,3；It is Type II under the basic method of operation 2,4.
According to the control method of DG device and connection-transformation device under 4 kinds of basic methods of operation, carry out at DG device node Reason.There is the control method of DG device in exchange micro-capacitance sensor subsystem: constant voltage constant frequency control, power limitation control and P- ω/Q-U are sagging Control.DG apparatus control method in DC micro-electric net system has: Isobarically Control, power limitation control and the sagging control of P-U.Mutually Even converter is used to support DC micro-electric net system under the basic method of operation 1, generally uses Isobarically Control；Substantially it is running Coordinated control AC and DC micro-capacitance sensor subsystem is used under 3 mode of mode, it is general using the sagging control of coordination.The basic method of operation 1, under 3, it is 1 exchange node that exchange micro-capacitance sensor subsystem, which is handled connection-transformation device, and defines the node as ILC friendship Flow node；It is 1 DC node by the processing of connection-transformation device for DC micro-electric net system, and defining the node is ILC DC node.9 kinds of node types, including 5 kinds of exchange node types are formed as a result: exchange balance nodes, PQ node, PV section Point, the sagging node of exchange exchange node with ILC；Including 4 kinds of DC node types: DC balance node, direct current invariable power node, The sagging node of direct current and ILC DC node.Exchange balance nodes, DC balance node and PQ node, PV node, direct current invariable power Node is similar to corresponding node type in conventional electric power system.
The equivalent power supply of the sagging node of AC and DC is active and reactive power is respectively
In formula, PDGa、QDGaEquivalent power supply to exchange sagging node is active and reactive power, ω, Ua、ω0、Ua0Respectively Exchange the virtual voltage frequency (system frequency) and amplitude and floating voltage frequency and amplitude of sagging node, Kpa -1、KQa -1It is corresponding The sagging coefficient of active and reactive power；PDGdFor the equivalent power supply active power of the sagging node of direct current, Ud、Ud0For the sagging section of direct current The virtual voltage and floating voltage of point, Kpd -1For the sagging coefficient of corresponding active power.
The equivalent power supply of ILC node is active and reactive power is
In formula, PILC、QILCThe equivalent power supply of respectively ILC node is active and reactive power；ω′,U′ILCdcRespectively ILC Side frequency, value of the virtual voltage after normalized of ILC DC side are exchanged, variation range is [- 1,1]；UILCac,0、 UILCacRespectively exchange side floating voltage and virtual voltage amplitude；KPILC、KQILCFor the active and idle control coefrficient of ILC.
Wherein, the LMNL algorithm solves AC and DC micro-capacitance sensor continuous tide initial point and the alternating based on LMNL algorithm Alternative manner solves alternating current-direct current and interconnects micro-capacitance sensor continuous tide initial point, specific as follows:
Continuous tide, which is calculated, to be calculated by initial point, predicts that link calculates and correction link calculates 3 steps and forms.Initial point It is the systematic steady state flow solution under given initial load parameter, solving essence is that conventional Load Flow calculates.Using alternating iteration side Method solves the conventional Load Flow of alternating current-direct current interconnection micro-capacitance sensor, and its essence is solve iteration mistake in exchange micro-capacitance sensor subsystem conventional Load Flow It is embedded in a complete DC micro-electric net system conventional Load Flow calculating process in journey, is handed under different running method as a result, straight The key that stream mixing micro-capacitance sensor conventional Load Flow calculates is attributed to the conventional Load Flow solution with/without balance nodes AC and DC micro-capacitance sensor. When using node power as injection rate, power flow equation is one group of nonlinear equation, can make the tide model shape of AC and DC micro-capacitance sensor Formula is consistent, is convenient for analytical calculation.It establishes and conventional Load Flow is unified with/without balance nodes AC and DC micro-capacitance sensor based on node power Model, application matrix form can be abbreviated as
F (x)=0, x=[U, δ, ω, xILC]∈Rn (4)
In formula, F (x) is that node in addition to balance nodes is active and reactive power equilibrium equation；X is the known variables of system Vector, n are the number of known variables；U, δ is unknown voltage magnitude vector sum voltage phase angle vector；xILCIt exchanges and saves for ILC The unknown state variable of point, ILC DC node.For having balance nodes system x without ω, for without ILC DC node and The system x that ILC exchanges node is free of xILC, x is free of for the DC micro-electric net system x without ILC DC nodeILC、δ、ω。
Further, F (x) can be written as
In formula, FP(xP) it is active power nonlinear function, FQ(xQ) it is exchange reactive power nonlinear function；PGFor node Equivalent power supply active power, QGTo exchange node equivalent power supply reactive power, PLFor node duty value active power, QLTo hand over Flow node duty value reactive power, PiActive power, Q are injected for nodeiReactive power is injected for exchange node；xPdc、xDdc、 xILCdcThe respectively unknown state vector of the sagging node of direct current invariable power node, direct current, ILC DC node；xPQ、xPV、xDac、 xILCacRespectively exchange the unknown state vector of PQ node, PV node, the sagging node of exchange, ILC exchange node.
LM (Levenberg Marquardt) method is suitable for solving Nonlinear System of Equations, and does not require Jacobian matrix Nonsingular, and initial value nonsingular in equation Jacobian matrix has local Quadratic Convergence close to accurate take off, but does not have Global convergence.Linear search technique is one of the main method for obtaining LM method global convergence.Linear search has dull line Property search and nonmonotone line search two ways.The shortcomings that dull linear search is the step-length very little sometimes obtained, especially When direction and the negative gradient direction that algorithm generates close to it is vertical when, and nonmonotone line search can overcome this defect, raising calculation Method speed.And nonmonotone line search not find a function the every single-step iteration of value all monotonic decreasings, so that the selection of step factor is more Tool elasticity.
Define LMNL algorithm: the three step LM algorithms with nonmonotone line search.LMNL algorithm is used to solve with/without balance The unified conventional Load Flow model of node AC and DC micro-capacitance sensor solves AC and DC micro-capacitance sensor continuous tide using LMNL algorithm Initial point.If F:R in formula (4)n→RnIt is continuously differentiable function, and there is tide under given service condition and under computational accuracy Stream solution x*.LMNL algorithm solves formula (4), first finds out d1k:
In formula, I is unit matrix, βkFor non-negative parameter；Fk=F (xk)、Jk=J (xk) be respectively in formula (4) F (x) working as The value and Jacobian matrix of preceding iteration point x.
Enable yk=xk+d1k, obtain d2k:
Enable zk=yk+d2k, obtain d3k:
βkShown in update mode such as formula (9):
Enable dLMNLk=d1k+d2k+d3k, judge whether formula (10) is true.
ψ(xk+dk)≤ρψ(xk)-k1||d1k||2-k2||d2k||2-k3||d3k||2 (10)
In formula, ψ (xk)=‖ F (xk)‖2, ρ, k1,k2,k3>0。
If formula (10) is set up, next iteration point xk+1=xk+dLMNLk, otherwise xk+1=xk+αkd1k+αk 2d2k+αk 3d3k, αkIt is updated by nonmonotone line search:
αk=max { 1, γ, γ2..., γ ∈ (0,1) (11)
αk=γi(i=0,1,2 ...) formula (12) need to be met.
In formula, σ1,σ2,σ3,σ4﹥ 0,
It repeats the above process, until meeting ‖ Jk TFk‖≤ε (the conventional Load Flow computational accuracy that ε is setting), obtains formula (4) Flow solution.
To formula (4), it is continuous to make hypothesis 1:F (x), J (x) Lipschitz.Make assume 2: ‖ F (x) ‖ N (x*, b) (b ∈ [0, 1] there are one in) to be weaker than the nonsingular local error bound of Jacobian matrix.Provable: if 1) assuming, 1 is set up, LMNL algorithm It is global convergence；If 2) assume, 2 are set up, and LMNL algorithm is third order dispersive.
The initial point of alternating current-direct current interconnection micro-capacitance sensor continuous tide is solved using the alternating iteration method based on LMNL algorithm.Base In the alternating iteration method of LMNL algorithm: the state variable frequencies omega of setting exchange micro-capacitance sensor subsystem and exchange node first The initial value (initial value for exchanging node voltage need to be only arranged in the case of having balance nodes in exchange micro-capacitance sensor subsystem) of voltage, makees ω For known quantity, DC micro-electric net system trend is solved using LMNL algorithm, obtains the transmitting active power of connection-transformation device PILCdc；By PILCac=PILCdcExchange micro-capacitance sensor subsystem node power equation is substituted into, into exchange micro-capacitance sensor subsystem trend Iterative process is solved, ω and exchange node voltage are calculated；Direct-current grid is carried out further according to the frequency correction value being calculated Subsystem Load flow calculation, continuous alternating iteration obtain the flow solution of alternating current-direct current mixing micro-capacitance sensor until convergence.The reality of this method Matter is to be embedded in a complete DC micro-electric net system trend in exchange micro-capacitance sensor subsystem Load Flow Solution iterative process Calculating process, and AC and DC micro-capacitance sensor subsystem all uses LMNL algorithm to carry out Load Flow Solution.
Wherein, the prediction link uses localized parameterization method and tangential method, and correction link uses hypersphere parameter Change method and combination Newton method form AC and DC micro-capacitance sensor continuous tide prediction correction link and calculate and formed based on two-way Iteration predicts that the alternating current-direct current interconnection micro-capacitance sensor continuous tide prediction correction link of bearing calibration calculates, specific as follows:
It is identical as the essence that the continuous tide of direct-current grid prediction correction link calculates to exchange micro-capacitance sensor, and prediction link is adopted With localized parameterization method and tangential method, correction link is using hypersphere parametric method and combines Newton method, as a result, shape It strikes a bargain, direct-current grid continuous tide prediction correction link calculates.
Continuous tide calculating is to track it in hyperspace using solution point known to certain of containing parameter power flow equation as initial point The process of the solution curve mapped out.AC and DC micro-capacitance sensor power flow equation containing load parameter based on node power can be abbreviated as
F (x)=0, x=[U, δ, ω, xILC,λ]∈Rn+1 (14)
In formula, f (x) is that node in addition to balance nodes is active and reactive power equation；λ is reflection system load level Load parameter, the same formula of the meaning of other variables (4).
Predict that link uses localized parameterization method and tangential method.Power flow equation containing load parameter is taken in datum Total differential has
In formula, the same formula of meaning (14) of f (x), x；xjFor the datum mark of jth time prediction, df (xj) it is f (x) in xjThat locates is micro- Point, d xjIt is x in xjThe differential at place, i.e. tangent vector.
It is local parameter that λ is selected when the 1st prediction, and the last prediction tangent vector interior joint voltage magnitude of hereafter selection falls most Serious quantity of state is local parameter.After obtaining determining tangent vector, future position is obtained by formula (16).
x′j+1=xj+αdxj (16)
In formula, x 'j+1For future position；α is the step-length of setting；When the 1st prediction, datum mark xjFor initial point.
Correction link is using hypersphere parametric method and combination Newton method.Correction link is with future position for approximation Solution obtains the determination solution point on solution curve by correction iteration, and its essence is using future position as initial value, solve containing parameter tide The accurate solution of flow equation.The unknown quantity number of containing parameter power flow equation need to use parametric method structural map 1 more than equation number Shape augments equation, and the power flow equation that is expanded is
In formula, g (x)=0 is supplement equation.
Supplement equation is obtained using hypersphere parametric method.The figure of this method construction are as follows: cross datum mark and tangent line is pre- Future position under survey method, and using this two o'clock line as the hypersphere of diameter, the equation of supplement is
The essence of extended power flow equations is one group of Nonlinear System of Equations.It is proposed combination Newton Algorithm extended power flow equations: It is first corrected using traditional Newton method, cannot still be restrained when it iterates to calculate the maximum number of iterations that number is more than setting When, then it is assumed that future position is far from solution curve, at this moment using band Armijo Linear search Newton method correction.Solve extension trend side The band Armijo Linear of journey (18) searches for Newton method and is
In formula, dANmFor iteration direction；xm+1For obtained value after the m times correction iterative calculation, when m=0, xmFor future position x′j+1； J′xmFor in Expanded Jacobian matrix J 'xIn xmThe value at place；M(xm) it is M (x) in xmThe value at place；γmFor the m times iteration Step-length is searched for by Armijo Linear and is determined.
The prediction correction link that alternating current-direct current interconnects micro-capacitance sensor continuous tide calculates the AC and DC for being different from single power-supplying forms Micro-capacitance sensor.The varying loading of alternating current-direct current interconnection micro-capacitance sensor continuous tide mainly divides 2 kinds of situations: varying loading is in exchange micro-capacitance sensor System side and varying loading are in DC micro-electric net system side.It is micro- that alternating current-direct current interconnection is carried out using bidirectional iteration prediction bearing calibration The prediction correction link of power grid continuous tide calculates: varying loading, only need to be micro- to exchanging when exchanging micro-capacitance sensor subsystem Power grid subsystem predicted and (is not required to predict DC micro-electric net system), and prediction result passes through frequency in correction link Rate influences connection-transformation device transimission power, and the Instable value for exchanging micro-capacitance sensor subsystem is transmitted to DC side, cross, straight Stream micro-capacitance sensor subsystem all needs to be corrected；Varying loading, only need to be to DC micro-electric in DC micro-electric net system Net system is predicted (be not required to exchange micro-capacitance sensor subsystem predict), and prediction result passes through ILC in correction link Node voltage influences its transimission power, and the Instable value of DC micro-electric net system is transmitted to exchange side, AC and DC Micro-capacitance sensor subsystem all needs to be corrected；Using localized parameterization method and tangential method is used, correction uses hypersphere for prediction Parametric method and combination Newton method.
A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method of the present invention, technical effect mainly have:
1. the present invention forms 2 kinds according to the 4 kinds of basic methods of operation and DG apparatus control method of alternating current-direct current mixing micro-capacitance sensor Alternating current-direct current mixing micro-capacitance sensor operating system type and 9 kinds of node types propose LMNL algorithm, the alternating iteration based on LMNL algorithm Method solves the initial point of AC and DC micro-capacitance sensor and alternating current-direct current interconnection micro-capacitance sensor continuous tide respectively, and prediction link is using part It is micro- to form AC and DC using hypersphere parametric method and combination Newton method for parametric method and tangential method, correction link Power grid continuous tide predicts that correction link calculates and predicts that the alternating current-direct current interconnection micro-capacitance sensor of bearing calibration is continuous based on bidirectional iteration Trend predicts that correction link calculates, and it is more to solve the complexity that alternating current-direct current mixing micro-capacitance sensor continuous tide under different running method calculates Sample problem has good engineering application value；
2. the conventional Load Flow method for solving proposed by the present invention based on LMNL algorithm, can handle multiple types node, it is applicable in It is calculated in the conventional Load Flow of the AC and DC micro-capacitance sensor with/without balance nodes, and is alternating current-direct current mixing micro-capacitance sensor conventional Load Flow meter The basis of calculation；
3. prediction link proposed by the present invention uses localized parameterization method and tangential method, correction link uses hypersphere Parametric method and combination Newton method, and bearing calibration is predicted based on this bidirectional iteration, it is suitable for alternating current-direct current and mixes micro- electricity The prediction of continuous tide, which corrects, under net different running method and different varying loadings calculates.
Detailed description of the invention
Fig. 1 is the step flow chart of the method for the present invention
Fig. 2 is the alternating current-direct current mixing micro-capacitance sensor continuous tide calculated result of embodiment
Fig. 3 is the alternating current-direct current mixing micro-capacitance sensor continuous tide calculated result 2 of embodiment
Fig. 4 is the alternating current-direct current mixing micro-capacitance sensor continuous tide calculated result 3 of embodiment
Fig. 5 is the alternating current-direct current mixing micro-capacitance sensor continuous tide calculated result 4 of embodiment
Specific embodiment
Technical solution of the present invention is described in further detail with reference to the accompanying drawing:
Fig. 1 is a kind of step flow chart of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method, comprising the following steps:
Step 1: the parameter and the method for operation of input alternating current-direct current mixing micro-capacitance sensor system line, load, DG device；Setting is negative Lotus changes node number, the corresponding λ of initial point0Value, varying loading growth pattern and default increment；Institute is differentiated according to the method for operation Belong to alternating current-direct current mixing micro-capacitance sensor operating system type, counts the number and corresponding node number of 9 kinds of node types.
Wherein, its affiliated alternating current-direct current mixing micro-capacitance sensor operating system type is differentiated according to the method for operation, specific as follows:
Alternating current-direct current mixing micro-capacitance sensor can form 2 kinds of operating system types under 4 kinds of basic methods of operation of difference.Alternating current-direct current is mixed Close micro-capacitance sensor operating system type I: alternating current-direct current interconnects micro-capacitance sensor.Alternating current-direct current mixing micro-capacitance sensor operating system Type II: it does not interconnect Exchange micro-capacitance sensor subsystem and DC micro-electric net system.Belong to alternating current-direct current mixing micro-capacitance sensor under the basic method of operation 1,3 Operating system type I；Belong to alternating current-direct current mixing micro-capacitance sensor operating system Type II under the basic method of operation 2,4.
Wherein, the number of 9 kinds of node types is counted, specific as follows:
5 kinds of exchange node types: exchange balance nodes, PQ node, PV node, the sagging node of exchange exchange node with ILC. 4 kinds of DC node types: DC balance node, direct current invariable power node, the sagging node of direct current and ILC DC node.For handing over Micro-capacitance sensor subsystem is flowed, the DG device processing of constant voltage constant frequency control is exchange balance nodes, is exchanged at power limitation control DG device Reason is PQ node or PV node, and to exchange sagging node, the processing of connection-transformation device is the sagging control DG device processing of P- ω/Q-U ILC exchanges node.For DC micro-electric net system, the DG device processing of DC constant voltage control is DC balance node, direct current The DG device processing of power limitation control is direct current invariable power node, and the DG device processing of the sagging control of P-U is the sagging node of direct current, The processing of connection-transformation device is ILC DC node.Node voltage size and the phase angle for exchanging balance nodes are given, DC balance section The node voltage size of point is given.
The node power equation of other 4 kinds of exchanges nodes can be uniformly written as
In formula, xacFor the unknown state vector of intercommunion subsystem；FPacm(xac)、FGacm(xac)、PGacm、QGacm、PLacm、 QLacm、 Pacm、QacmRespectively exchange active and reactive power nonlinear function, the active and idle function of equivalent power supply of node m Rate, duty value be active and reactive power and injects active and reactive power；SPQ、SPV、SDacRespectively PQ node, PV node, Exchange the node serial number set of sagging node.
The node power equation of other 3 kinds of DC nodes can be uniformly written as
Fdci(xdc)=PGdci-PLdci-Pdci=0 (2)
In formula, xdcFor the unknown state vector of direct current subsystem；Fdci(xdc)、PGdci、PLdci、PdciRespectively DC node Active power nonlinear function, equivalent power supply active power, duty value active power and the injection active power of i.
Step 2: judging whether to belong to alternating current-direct current mixing micro-capacitance sensor operating system type I? if so, 3 are entered step, if it is not, Then enter step 4.
Step 3: the continuous tide for carrying out exchange micro-capacitance sensor subsystem and DC micro-electric net system calculates:
1) initial point is solved using LMNL algorithm；
2) prediction link using localized parameterization method and tangential method, correction link using hypersphere parametric method and Newton method is combined, prediction correction link is carried out and calculates；
3) using predict tangent vector d λ symbol as critical point criterion, judge whether to meet critical pumping rate? if satisfied, then tying Beam continuous tide calculates, and calculated result is exported, if not satisfied, then going to step 2).
Wherein, LMNL algorithm solves initial point, specific as follows:
Set the corresponding λ of initial point0Value and AC and DC micro-capacitance sensor continuous tide initial point computational accuracy ε.It establishes based on section The AC and DC micro-capacitance sensor of point power unifies conventional Load Flow model F (x)=0.First find out d1k:
In formula, Fk=F (xk)、Jk=J (xk) it is respectively F (x) in current iteration point xkValue and Jacobian matrix；I is Unit matrix, βkFor non-negative parameter.
Enable yk=xk+d1k, obtain d2k:
Enable zk=yk+d2k, obtain d3k:
βkShown in update mode such as formula (6):
Enable dLMNLk=d1k+d2k+d3k, judge whether formula (7) is true.
ψ(xk+dk)≤ρψ(xk)-k1||d1k||2-k2||d2k||2-k3||d3k||2 (7)
In formula, ψ (xk)=‖ F (xk)‖2, ρ, k1,k2,k3>0。
If formula (7) is set up, next iteration point xk+1=xk+dLMNLk, otherwise xk+1=xk+αkd1k+αk 2d2k+αk 3d3k, αk It is updated by nonmonotone line search:
αk=max { 1, γ, γ2..., γ ∈ (0,1) (8)
αk=γi(i=0,1,2 ...) formula (9) need to be met.
In formula, σ1,σ2,σ3,σ4﹥ 0,
It repeats the above process, until meeting ‖ Jk TFk‖≤ε obtains flow solution xk, as given λ0Under computational accuracy ε Initial point.
Wherein, prediction link uses hypersphere parametrization side using localized parameterization method and tangential method, correction link Method and combination Newton method carry out prediction correction link and calculate, specific as follows:
Given varying loading growth pattern and default increment establish the AC and DC micro-capacitance sensor based on node power containing negative Lotus parameter power flow equation f (x)=0 enables prediction number j=0, carries out prediction link and calculates:
1. taking total differential in datum to containing parameter power flow equation f (x)=0:
In formula, xjFor the datum mark of jth time prediction, j=1,2 .., NP, NPTo predict total degree；df(xj) it is f (x) in xj The differential at place, dxjIt is x in xjThe differential at place, i.e. tangent vector.
2. using localized parameterization method, it is local parameter that when the 1st prediction, which selects λ, and the last prediction of hereafter selection is tangentially The quantity of state that amount interior joint voltage magnitude falls most serious is local parameter.After obtaining determining tangent vector, obtained by formula (12) Future position.
x′j+1=xj+αdxj (12)
In formula, x 'j+1For future position；α is the step-length of setting；When the 1st prediction, datum mark xjFor initial point.
Step-length α, correction computational accuracy ε are set2, traditional Newton method correction maximum number of iterations Nmax, and enable correction iteration Calculation times m=0, the initial value of extended power flow equations solution are x0=x 'j+1, it is corrected link calculating:
1. use hypersphere parametric method obtain supplement equation for
2. solving extended power flow equations using traditional Newton method；
3. if amount of unbalance | Δ M (xm)|≤ε2, then obtain determining solution point xm(i.e. xj+1), otherwise go to step 4.；
4. enabling m=m+1, m > N is differentiatedmaxIf going to step 5., if it is not, going to step 2.；
5. judging to calculate and restrain using combination Newton Algorithm extended power flow equations? if so, obtaining computational accuracy ε2 Under xm, if it is not, then terminating to calculate.
Step 4: it carries out alternating current-direct current interconnection micro-capacitance sensor continuous tide and calculates:
1) initial point is solved using the alternating iteration method based on LMNL algorithm；
2) it carries out prediction correction link using bidirectional iteration prediction bearing calibration to calculate, wherein prediction uses local parameter Change method and tangential method, correction is using hypersphere parametric method and combination Newton method；
3) using predict tangent vector d λ symbol as critical point criterion, judge whether to meet critical pumping rate? if satisfied, then tying Beam continuous tide calculates, and calculated result is exported, if not satisfied, then going to step 2).
Wherein, initial point is solved using the alternating iteration method based on LMNL algorithm, specific as follows:
1. the steady frequency initial value ω of exchange micro-capacitance sensor subsystem is arrangedk2, the initial value of unknown exchange node voltage, order k2=0；
2. carrying out DC micro-electric net system conventional Load Flow using LMNL algorithm to solve, at this time by steady frequency ωk2As Known quantity obtains calculation of tidal current xdck1And PILCdc；
3. by PILCac=PILCdcThe power flow equation of exchange micro-capacitance sensor subsystem is substituted into, and steady frequency is unknown state change Amount carries out exchange micro-capacitance sensor subsystem conventional Load Flow using LMNL algorithm and solves iteration, calculates the iterative value of steady frequency ωk2+1And the iterative value of unknown exchange node voltage；
4. terminating to calculate, output is handed over straight if exchange micro-capacitance sensor subsystem conventional Load Flow calculating meets its condition of convergence The conventional Load Flow solution x of stream interconnection micro-capacitance sensordck1、xack2, otherwise, enable k2=k2+ 1, it is transferred to step 2..
Wherein, prediction correction link is carried out using bidirectional iteration prediction bearing calibration to calculate, specific as follows:
Varying loading is carried out exchange prediction link and calculates and replaced using alternating current-direct current when exchanging micro-capacitance sensor subsystem The correction link of iteration calculates:
1. calculating the predicted value for exchanging micro-capacitance sensor subsystem with tangential method using localized parameterization method；
2. using the predicted value for exchanging micro-capacitance sensor subsystem as initial value, using hypersphere parametric method and combination newton side The correction that method carries out exchange micro-capacitance sensor subsystem calculates, and obtains independent accurate solution；
3. substituting into direct-current grid using the frequency correction value in the independent corrected value for exchanging micro-capacitance sensor subsystem as known quantity The correction link of subsystem calculates correction voltage value, amount of unbalance and the new transimission power of DC micro-electric net system；
4. judging that amount of unbalance maximum value is less than computational accuracy? if so, saving AC and DC micro-capacitance sensor subsystem 2. corrected value is transferred to step if it is not, then solving using what this correction calculated as the initial value for correcting calculating next time.
Varying loading is carried out the calculating of DC prediction link and is replaced using alternating current-direct current in DC micro-electric net system The correction link of iteration calculates:
1. calculating DC micro-electric net using localized parameterization method and tangential method with the initial value of initial calculation The predicted value of system；
2. using the predicted value of DC micro-electric net system as initial value, using hypersphere parametric method and combination newton side Method carries out the correction of DC micro-electric net system, obtains independent accurate solution；
3. substituting into exchange micro-capacitance sensor using the transmitting power in the independent corrected value of DC micro-electric net system as known quantity The correction link of system calculates corrected value, amount of unbalance and the new systematic steady state frequency of exchange micro-capacitance sensor subsystem.
4. judging whether amount of unbalance maximum value is less than computational accuracy, if so, saving AC and DC micro-capacitance sensor subsystem 2. corrected value is transferred to step if it is not, then solving using what this correction calculated as the initial value for correcting calculating next time.
The present invention uses low by the improved 38 node exchange micro-capacitance sensor subsystem of 33 node system of IEEE, Benchmark The alternating current-direct current mixing micro-capacitance sensor example system of the 17 node DC micro-electric net systems composition of pressure micro-capacitance sensor structure of modification is tested Card.
Parameter is set in LMNL algorithm: μ=10-6, ρ=0.8, k1=k2=k3=0.005, σ1=σ2=σ3=σ4= 0.005, γ=0.5, m=0.01, η=0.5, Load flow calculation precision ε=10-5.ε is set2=10-5, Nmax=10, λ0=0, become Change load to increase in a manner of constant power factor.
Alternating current-direct current mixing micro-capacitance sensor example system is set, wherein 33 node power distribution net system of IEEE node 8,18, 22,25 and 33 5 DG devices of access, constitute 38 node intercommunion subsystems；Benchmark low pressure micro-capacitance sensor structure is taken, in node 1,6,8 and 17 4 DG devices of access, are transformed into 17 node direct current subsystems；The interconnection for connecting ac bus and DC bus becomes Exchanger unit exchanges node for intercommunion subsystem processing for 1 ILC, and node serial number 39 is for direct current subsystem processes 1 ILC DC node, node serial number 18.
The continuous of operating system Type II is carried out with alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method of the invention Load flow calculation:
If alternating current-direct current mixing micro-capacitance sensor example system is under the basic method of operation 4, connection-transformation device does not work.Setting is handed over Micro-capacitance sensor subsystem: reference capacity 1MVA, reference frequency 50Hz is flowed, fixed phase angle is 0rad, and steady frequency range is [0.996,1.004] pu, node voltage amplitude per unit value range are [0.9400,1.0600] pu, the voltage phase angle of node 1 For system reference phase angle, all DG devices are all made of the sagging control of P- ω/Q-U, all sagging control DG devices it is specified defeated The per unit value of frequency and voltage magnitude is 1pu, 1pu out.DC micro-electric net system: reference power 100kVA is set, is owned DG device is all made of the sagging control of P-U.The quiescent voltage and static frequency characteristic of load are not considered, exchange micro-capacitance sensor subsystem The initial value of unknown frequency, unknown exchange node voltage amplitude and phase angle is taken as 1pu, 1pu, 0rad, direct-current grid respectively The initial value of subsystem unknown node voltage is taken as 1pu.
The continuous tide for exchanging micro-capacitance sensor subsystem calculates, and the setting of varying loading point obtains continuous tide on exchange node 7 P- ω curve, the P-U curve of node 5 of stream are as shown in Figure 2.Calculated result shows: 1. obtaining the frequency of continuous tide initial point For 0.9979pu, this is because all DG devices are all made of sagging control characteristic in system, and all sagging control DG devices Total active power is 2.8000pu, and the total burden with power of system is 3.7150pu, after system reaches stable state, sagging control The output power of DG device will meet theory analysis greater than its rated power to meet the equilibrium of supply and demand of system power.Verifying The correctness of the conventional Load Flow method for solving based on LMNL algorithm proposed, and this method can handle multiple types node, fit For the AC and DC micro-capacitance sensor conventional Load Flow calculating with/without balance nodes.2. the λ critical value of P-U curve and P- ω curve is all 56.5765 demonstrating the correctness of the method for the present invention.
The continuous tide of DC micro-electric net system calculates, and varying loading point is arranged on DC node 7, obtains continuous tide The P-U curve of the node 5 of stream is as shown in Figure 3.Calculated result shows that the λ critical value of P-U curve is 16.7693, demonstrates this hair The correctness of bright method.
The continuous tide of operating system type I is carried out with alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method of the invention Stream calculation:
If alternating current-direct current mixing micro-capacitance sensor example system is in the basic method of operation 3.ILC device parameter setting: UILCdc,max、 UILCdc,minRespectively 1.06pu and 0.94pu, ωmax、ωminRespectively 1.004pu and 0.996pu, KPILC=8, KQILC=5.
Varying loading obtains continuous tide on exchange node 7 in exchange micro-capacitance sensor subsystem side, the setting of varying loading point The P-U curve of different nodes is as shown in Figure 4.Calculated result shows: system change load parameter λ critical value is 62.0163, same DC node is identical with the P-U curve bifurcation point for exchanging node under varying loading node, i.e. λ critical value is equal, demonstrates this hair The correctness of bright method.
Varying loading is arranged on DC node 7 in DC micro-electric net system side, varying loading point, obtains continuous tide The P-U curve of different nodes is as shown in Figure 5.Calculated result shows: system change load parameter λ critical value is 49.7459, same DC node is identical with the P-U curve bifurcation point for exchanging node under varying loading node, demonstrates the correctness of the method for the present invention.

Claims (2)

1. a kind of alternating current-direct current mixing micro-capacitance sensor tidal current computing method, it is characterised in that: form 2 kinds of alternating current-direct current mixing micro-capacitance sensor operations System type and 9 kinds of node types propose band nonmonotone line search LM (Levenberg-Marquardt with a Nonmonotone line search, LMNL) algorithm solve AC and DC micro-capacitance sensor continuous tide initial point and based on LMNL calculate The alternating iteration method of method solves alternating current-direct current and interconnects micro-capacitance sensor continuous tide initial point, and prediction link uses localized parameterization method And it is continuously damp to form AC and DC micro-capacitance sensor using hypersphere parametric method and combination Newton method for tangential method, correction link Stream prediction correction link calculates and the alternating current-direct current interconnection micro-capacitance sensor continuous tide based on bidirectional iteration prediction bearing calibration predicts school Positive link calculates；
Wherein, 2 kinds of alternating current-direct current mixing micro-capacitance sensor operating system types of the formation and 9 kinds of node types, specific as follows:
Alternating current-direct current mixing micro-capacitance sensor is by exchange micro-capacitance sensor subsystem and accesses public network, has be incorporated into the power networks mode and isolated operation mould Formula, incorporating interconnecting converter connection, alternating current-direct current mixing micro-capacitance sensor have 4 kinds of basic methods of operation: joint grid-connected mode (fortune Line mode 1), hand over and directly from mode (method of operation 2), joint off-network mode (method of operation 3) and distinguish off-network mode (operation side Formula 4), whether interconnected according to AC and DC micro-capacitance sensor subsystem, it will be under 4 kinds of basic methods of operation of difference of alternating current-direct current mixing micro-capacitance sensor The operating system of formation is divided into 2 seed types, alternating current-direct current mixing micro-capacitance sensor operating system type I: alternating current-direct current interconnects micro-capacitance sensor, hands over straight Stream mixing micro-capacitance sensor operating system Type II: the exchange micro-capacitance sensor subsystem and DC micro-electric net system not interconnected, alternating current-direct current Mixing micro-capacitance sensor is type I under the basic method of operation 1,3, is Type II under the basic method of operation 2,4；
According to the control method of DG device and connection-transformation device under 4 kinds of basic methods of operation, DG device node processing is carried out, is handed over There is the control method of DG device in stream micro-capacitance sensor subsystem: constant voltage constant frequency control, power limitation control and the sagging control of P- ω/Q-U, DG apparatus control method in DC micro-electric net system has: Isobarically Control, power limitation control and the sagging control of P-U, and interconnection becomes Parallel operation is used to support DC micro-electric net system under the basic method of operation 1, Isobarically Control is generally used, in the basic method of operation Coordinated control AC and DC micro-capacitance sensor subsystem is used under 3 modes, it is general using the sagging control of coordination, the basic method of operation 1,3 Under, it is 1 exchange node that exchange micro-capacitance sensor subsystem, which is handled connection-transformation device, and defining the node is ILC exchange section The processing of connection-transformation device is 1 DC node for DC micro-electric net system by point, and defining the node is ILC direct current section Point forms 9 kinds of node types, including 5 kinds of exchange node types as a result: under exchange balance nodes, PQ node, PV node, exchange The node that hangs down exchanges node, including 4 kinds of DC node types with ILC: DC balance node, direct current invariable power node, direct current are sagging Node and ILC DC node, exchange balance nodes, DC balance node and PQ node, PV node, direct current invariable power node are similar Corresponding node type in conventional electric power system；
The equivalent power supply of the sagging node of AC and DC is active and reactive power is respectively
In formula, PDGa、QDGaEquivalent power supply to exchange sagging node is active and reactive power, ω, Ua、ω0、Ua0Respectively exchange The virtual voltage frequency (system frequency) and amplitude and floating voltage frequency and amplitude of sagging node, Kpa -1、KQa -1To have accordingly The sagging coefficient of function, reactive power, PDGdFor the equivalent power supply active power of the sagging node of direct current, Ud、Ud0For the sagging node of direct current Virtual voltage and floating voltage, Kpd -1For the sagging coefficient of corresponding active power；
The equivalent power supply of ILC node is active and reactive power is
In formula, PILC、QILCThe equivalent power supply of respectively ILC node is active and reactive power, ω ', U 'ILCdcRespectively ILC exchange Side frequency, value of the virtual voltage after normalized of ILC DC side, variation range are [- 1,1], UILCac,0、UILCacPoint Side floating voltage and virtual voltage amplitude, K Wei not exchangedPILC、KQILCFor the active and idle control coefrficient of ILC；
Wherein, the LMNL algorithm solves AC and DC micro-capacitance sensor continuous tide initial point and the alternating iteration based on LMNL algorithm Method solves alternating current-direct current and interconnects micro-capacitance sensor continuous tide initial point, specific as follows:
Continuous tide calculate by initial point calculate, prediction link calculate and correction link calculate 3 steps forms, initial point be to Determine the systematic steady state flow solution under initial load parameter, solving essence is that conventional Load Flow calculates, and is asked using alternating iteration method The conventional Load Flow for solving alternating current-direct current interconnection micro-capacitance sensor, its essence is solve in iterative process in exchange micro-capacitance sensor subsystem conventional Load Flow It is embedded in a complete DC micro-electric net system conventional Load Flow calculating process, alternating current-direct current is mixed under different running method as a result, Close micro-capacitance sensor conventional Load Flow calculate key be attributed to with/without balance nodes AC and DC micro-capacitance sensor conventional Load Flow solution, when with When node power is injection rate, power flow equation is one group of nonlinear equation, can make the tide model form one of AC and DC micro-capacitance sensor It causes, is convenient for analytical calculation, establish and conventional Load Flow model is unified with/without balance nodes AC and DC micro-capacitance sensor based on node power, Its application matrix form can be abbreviated as
F (x)=0, x=[U, δ, ω, xILC]∈Rn (4)
In formula, F (x) is that node in addition to balance nodes is active and reactive power equilibrium equation, x be system known variables to Amount, n are the number of known variables, and U, δ are unknown voltage magnitude vector sum voltage phase angle vector, xILCIt exchanges and saves for ILC The unknown state variable of point, ILC DC node, for having balance nodes system x without ω, for without ILC DC node and The system x that ILC exchanges node is free of xILC, x is free of for the DC micro-electric net system x without ILC DC nodeILC,δ,ω；
Further, F (x) can be written as
In formula, FP(xP) it is active power nonlinear function, FQ(xQ) it is exchange reactive power nonlinear function, PGFor node equivalence Power supply active power, QGTo exchange node equivalent power supply reactive power, PLFor node duty value active power, QLFor exchange section Point duty value reactive power, PiActive power, Q are injected for nodeiReactive power, x are injected for exchange nodePdc、xDdc、xILCdc The respectively unknown state vector of the sagging node of direct current invariable power node, direct current, ILC DC node, xPQ、xPV、xDac、xILCacPoint The unknown state vector of PQ node, PV node, the sagging node of exchange, ILC exchange node Wei not exchanged；
LM (Levenberg Marquardt) method is suitable for solving Nonlinear System of Equations, and does not require the non-surprise of Jacobian matrix Different, and initial value nonsingular in equation Jacobian matrix has local Quadratic Convergence close to accurate take off, but does not have the overall situation Convergence, linear search technique are one of the main methods for obtaining LM method global convergence, and linear search has dullness linearly to search The shortcomings that rope and nonmonotone line search two ways, dull linear search is the step-length very little sometimes obtained, especially works as calculation Direction and the negative gradient direction that method generates close to it is vertical when, and nonmonotone line search can overcome this defect, and raising algorithm is fast Degree, and nonmonotone line search not find a function the every single-step iteration of value all monotonic decreasings, so that the selection of step factor has more bullet Property；
LMNL algorithm: the three step LM algorithms with nonmonotone line search is defined, LMNL algorithm is used to solve with/without balance nodes The unified conventional Load Flow model of AC and DC micro-capacitance sensor solves the initial of AC and DC micro-capacitance sensor continuous tide using LMNL algorithm Point, if F:R in formula (4)n→RnIt is continuously differentiable function, and there are flow solutions under given service condition and under computational accuracy X*, LMNL algorithm solve formula (4), first find out d1k:
In formula, I is unit matrix, βkFor non-negative parameter, Fk=F (xk)、Jk=J (xk) be respectively in formula (4) F (x) change currently The value and Jacobian matrix of generation point x；
Enable yk=xk+d1k, obtain d2k:
Enable zk=yk+d2k, obtain d3k:
βkShown in update mode such as formula (9):
Enable dLMNLk=d1k+d2k+d3k, judge whether formula (10) is true,
ψ(xk+dk)≤ρψ(xk)-k1||d1k||2-k2||d2k||2-k3||d3k||2 (10)
In formula, ψ (xk)=‖ F (xk)‖2, ρ, k1,k2,k3>0；
If formula (10) is set up, next iteration point xk+1=xk+dLMNLk, otherwise xk+1=xk+αkd1k+αk 2d2k+αk 3d3k, αkBy non- Dull linear search updates:
αk=max { 1, γ, γ2..., γ ∈ (0,1) (11)
αk=γi(i=0,1,2 ...) need to meet formula (12),
In formula, σ1,σ2,σ3,σ4﹥ 0；
It repeats the above process, until meeting ‖ Jk TFk‖≤ε (the conventional Load Flow computational accuracy that ε is setting), obtains the trend of formula (4) Solution；
To formula (4), it is continuous to make hypothesis 1:F (x), J (x) Lipschitz, makees to assume 2: ‖ F (x) ‖ in N (x*, b) (b ∈ [0,1]) Interior there are one to be weaker than the nonsingular local error bound of Jacobian matrix, provable: if 1) assuming, 1 is set up, and LMNL algorithm is complete Office is convergent, and 2) if assuming, 2 are set up, LMNL algorithm is third order dispersive；
The initial point that alternating current-direct current interconnection micro-capacitance sensor continuous tide is solved using the alternating iteration method based on LMNL algorithm, is based on The alternating iteration method of LMNL algorithm: the state variable frequencies omega of setting exchange micro-capacitance sensor subsystem and exchange node voltage first Initial value (exchange micro-capacitance sensor subsystem has the initial value that exchange node voltage need to only be arranged in the case of balance nodes), using ω as The amount of knowing solves DC micro-electric net system trend using LMNL algorithm, obtains the transmitting active power P of connection-transformation deviceILCdc, will PILCac=PILCdcExchange micro-capacitance sensor subsystem node power equation is substituted into, into exchange micro-capacitance sensor subsystem Load Flow Solution iteration Process, calculates ω and exchange node voltage, carries out DC micro-electric net system tide further according to the frequency correction value being calculated Stream calculation, continuous alternating iteration obtain the flow solution of alternating current-direct current mixing micro-capacitance sensor, the essence of this method is to hand over until convergence It is embedded in a complete DC micro-electric net system Load flow calculation process in stream micro-capacitance sensor subsystem Load Flow Solution iterative process, And AC and DC micro-capacitance sensor subsystem all uses LMNL algorithm to carry out Load Flow Solution；
Wherein, the prediction link uses localized parameterization method and tangential method, and correction link uses hypersphere parametrization side Method and combination Newton method form AC and DC micro-capacitance sensor continuous tide and predict correction link calculating and formed to be based on bidirectional iteration Predict that the alternating current-direct current interconnection micro-capacitance sensor continuous tide prediction correction link of bearing calibration calculates, specific as follows:
It is identical as the essence that the continuous tide of direct-current grid prediction correction link calculates to exchange micro-capacitance sensor, and prediction link uses office Portion's parametric method and tangential method, correction link are formed as a result, using hypersphere parametric method and combination Newton method AC and DC micro-capacitance sensor continuous tide predicts that correction link calculates；
Continuous tide calculating is to track it using solution point known to certain of containing parameter power flow equation as initial point and map in hyperspace The process of solution curve out, the AC and DC micro-capacitance sensor power flow equation containing load parameter based on node power can be abbreviated as
F (x)=0, x=[U, δ, ω, xILC,λ]∈Rn+1 (14)
In formula, f (x) is that the node in addition to balance nodes is active and reactive power equation, λ are the load for reflecting system load level Parameter, the same formula of the meaning of other variables (4)；
Predict link use localized parameterization method and tangential method, to power flow equation containing load parameter datum take entirely it is micro- Dividing has
In formula, the same formula of meaning (14) of f (x), x, xjFor the datum mark of jth time prediction, df (xj) it is f (x) in xjThe differential at place, dxjIt is x in xjThe differential at place, i.e. tangent vector；
It is local parameter that λ is selected when the 1st prediction, and the last prediction tangent vector interior joint voltage magnitude of hereafter selection falls most serious Quantity of state be local parameter, after obtaining determining tangent vector, future position is obtained by formula (16)；
x′j+1=xj+αdxj (16)
In formula, x 'j+1For future position, α is the step-length of setting, when the 1st prediction, datum mark xjFor initial point；
For correction link using hypersphere parametric method and combination Newton method, correction link is led to using future position as approximate solution Overcorrect iteration obtains the determination solution point on solution curve, and its essence is solve containing parameter trend side using future position as initial value The accurate solution of journey, the unknown quantity number of containing parameter power flow equation 1 more than equation number, need to using parametric method constructing graphic come Equation is augmented, the power flow equation that is expanded is
In formula, g (x)=0 is supplement equation；
Supplement equation, the figure of this method construction are as follows: cross datum mark and tangent line predicted method are obtained using hypersphere parametric method Under future position, and using this two o'clock line as the hypersphere of diameter, the equation of supplement is
The essence of extended power flow equations is one group of Nonlinear System of Equations, proposes combination Newton Algorithm extended power flow equations: first adopting It is corrected with traditional Newton method, when it iterates to calculate the maximum number of iterations that number is more than setting and cannot still restrain, then Think that future position far from solution curve, at this moment using band Armijo Linear search Newton method correction, solves extended power flow equations (18) band Armijo Linear searches for Newton method and is
In formula, dANmFor iteration direction, xm+1For obtained value after the m times correction iterative calculation, when m=0, xmFor future position x 'j+1, J′xmFor in Expanded Jacobian matrix J 'xIn xmThe value at place, M (xm) it is M (x) in xmThe value at place, γmFor the m times iteration step length, by The search of Armijo Linear determines；
The prediction correction link that alternating current-direct current interconnects micro-capacitance sensor continuous tide calculates the micro- electricity of AC and DC for being different from single power-supplying forms Net, the varying loading of alternating current-direct current interconnection micro-capacitance sensor continuous tide mainly divide 2 kinds of situations: varying loading is in exchange micro-capacitance sensor subsystem Side and varying loading carry out alternating current-direct current using bidirectional iteration prediction bearing calibration and interconnect micro-capacitance sensor in DC micro-electric net system side The prediction correction link of continuous tide calculates: varying loading, only need to be to exchange micro-capacitance sensor when exchanging micro-capacitance sensor subsystem Subsystem predicted and (is not required to predict DC micro-electric net system), and prediction result passes through frequency pair in correction link Connection-transformation device transimission power is influenced, and the Instable value for exchanging micro-capacitance sensor subsystem is transmitted to DC side, AC and DC is micro- Power grid subsystem all needs to be corrected, and varying loading, only need to be to DC micro-electric net in DC micro-electric net system System is predicted (be not required to exchange micro-capacitance sensor subsystem predict), and prediction result passes through ILC node in correction link Voltage influences its transimission power, and the Instable value of DC micro-electric net system is transmitted to exchange side, the micro- electricity of AC and DC Net system all needs to be corrected, and using localized parameterization method and tangential method is used, correction uses hypersphere parameter for prediction Change method and combination Newton method.
2. alternating current-direct current mixing micro-capacitance sensor according to claim 1, feature alternating current-direct current mixing micro-capacitance sensor can be joint simultaneously Net mode (method of operation 1) is handed over and directly from mode (method of operation 2), joint off-network mode (method of operation 3) or difference off-network side Alternating current-direct current mixing micro-capacitance sensor under formula (method of operation 4).
CN201811064879.0A 2018-09-12 2018-09-12 A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method Pending CN109274116A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811064879.0A CN109274116A (en) 2018-09-12 2018-09-12 A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811064879.0A CN109274116A (en) 2018-09-12 2018-09-12 A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method

Publications (1)

Publication Number Publication Date
CN109274116A true CN109274116A (en) 2019-01-25

Family

ID=65188841

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811064879.0A Pending CN109274116A (en) 2018-09-12 2018-09-12 A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method

Country Status (1)

Country Link
CN (1) CN109274116A (en)

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110148946A (en) * 2019-05-24 2019-08-20 浙江大学 A kind of isolated island micro-capacitance sensor tidal current computing method solved based on two step of confactor
CN110336284A (en) * 2019-07-30 2019-10-15 湘潭大学 Isolated operation alternating current-direct current mixing micro-capacitance sensor static security methods of risk assessment
CN110474329A (en) * 2019-08-21 2019-11-19 山东劳动职业技术学院（山东劳动技师学院） A kind of continuous tide variable step size method and system based on polynomial regression technology
CN110707762A (en) * 2019-10-12 2020-01-17 国网山东省电力公司济南供电公司 Multi-energy complementary alternating current-direct current hybrid micro-grid load flow feasible region calculation method
CN111355241A (en) * 2019-12-03 2020-06-30 上海交通大学 Alternating current-direct current hybrid micro-grid droop coefficient selection method based on tidal current feasible region
CN111817359A (en) * 2020-06-28 2020-10-23 河海大学 Micro-grid solvable boundary analysis method based on equivalent circuit
CN113158573A (en) * 2021-04-29 2021-07-23 广西大学 Small hydropower station regional distribution network optimal power flow obtaining method based on deep learning

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104376156A (en) * 2014-10-31 2015-02-25 西南交通大学 Complex network construction method based on adjacent matrixes
CN105514971A (en) * 2015-11-02 2016-04-20 湘潭大学 Flow calculation method suitable for microgrids in various operation modes
CN106229987A (en) * 2016-09-14 2016-12-14 国网上海市电力公司 A kind of alternating current-direct current combined hybrid system tidal current computing method based on the LM algorithm improved
US20170250567A1 (en) * 2007-08-28 2017-08-31 Causam Energy, Inc. System and method for estimating and providing dispatchable operating reserve energy capacity through use of active load management
CN108493940A (en) * 2018-03-14 2018-09-04 广东电网有限责任公司汕头供电局 A kind of Ac/dc Power Systems trend alternating iteration computational methods containing flexible direct current

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170250567A1 (en) * 2007-08-28 2017-08-31 Causam Energy, Inc. System and method for estimating and providing dispatchable operating reserve energy capacity through use of active load management
CN104376156A (en) * 2014-10-31 2015-02-25 西南交通大学 Complex network construction method based on adjacent matrixes
CN105514971A (en) * 2015-11-02 2016-04-20 湘潭大学 Flow calculation method suitable for microgrids in various operation modes
CN106229987A (en) * 2016-09-14 2016-12-14 国网上海市电力公司 A kind of alternating current-direct current combined hybrid system tidal current computing method based on the LM algorithm improved
CN108493940A (en) * 2018-03-14 2018-09-04 广东电网有限责任公司汕头供电局 A kind of Ac/dc Power Systems trend alternating iteration computational methods containing flexible direct current

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
JIA CAO等: "A modified Levenberg-Marquardt approach to explore the limit operation state of AC/DC hybrid system", 《2015 IEEE POWER & ENERGY SOCIETY GENERAL MEETING》 *
彭寒梅等: "孤岛运行交直流混合微电网的潮流计算", 《电网技术》 *
曹佳等: "几种高阶收敛的Levenberg-Marquardt方法在潮流计算中的应用", 《电网技术》 *

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110148946A (en) * 2019-05-24 2019-08-20 浙江大学 A kind of isolated island micro-capacitance sensor tidal current computing method solved based on two step of confactor
CN110148946B (en) * 2019-05-24 2022-03-25 浙江大学 Island microgrid load flow calculation method based on two-step solution of auxiliary factors
CN110336284A (en) * 2019-07-30 2019-10-15 湘潭大学 Isolated operation alternating current-direct current mixing micro-capacitance sensor static security methods of risk assessment
CN110474329A (en) * 2019-08-21 2019-11-19 山东劳动职业技术学院（山东劳动技师学院） A kind of continuous tide variable step size method and system based on polynomial regression technology
CN110474329B (en) * 2019-08-21 2020-11-06 山东劳动职业技术学院（山东劳动技师学院） Continuous power flow step length changing method and system based on polynomial regression technology
CN110707762A (en) * 2019-10-12 2020-01-17 国网山东省电力公司济南供电公司 Multi-energy complementary alternating current-direct current hybrid micro-grid load flow feasible region calculation method
CN111355241A (en) * 2019-12-03 2020-06-30 上海交通大学 Alternating current-direct current hybrid micro-grid droop coefficient selection method based on tidal current feasible region
CN111355241B (en) * 2019-12-03 2023-08-08 上海交通大学 AC/DC hybrid micro-grid droop coefficient selection method based on tide feasible region
CN111817359A (en) * 2020-06-28 2020-10-23 河海大学 Micro-grid solvable boundary analysis method based on equivalent circuit
CN111817359B (en) * 2020-06-28 2021-11-02 河海大学 Micro-grid solvable boundary analysis method based on equivalent circuit
CN113158573A (en) * 2021-04-29 2021-07-23 广西大学 Small hydropower station regional distribution network optimal power flow obtaining method based on deep learning
CN113158573B (en) * 2021-04-29 2022-03-29 广西大学 Small hydropower station regional distribution network optimal power flow obtaining method based on deep learning

Similar Documents

Publication Publication Date Title
CN109274116A (en) A kind of alternating current-direct current mixing micro-capacitance sensor continuous tide calculation method
CN106877338B (en) The alternating current-direct current micro-capacitance sensor uncertain optimization operation method of the intermittent energy source containing high density
Xia et al. Direct nonlinear primal–dual interior-point method for transient stability constrained optimal power flow
CN105048468B (en) Power transmission-distribution network integrating voltage stability assessment method based on distributed calculation
CN109508499A (en) Multi-period more optimal on-positions of scene distribution formula power supply and capacity research method
CN101982917B (en) Calculation method of available transmission capacity for power grid scheduling
CN105958496B (en) The total transfer capability calculation method of system containing VSC based on continuous tide
CN110797874B (en) State estimation method for AC/DC hybrid power distribution network containing power electronic transformer
CN102184297B (en) Electrical/control system decoupling prediction method suitable for transient parallel simulation of micro-nets
CN106250640A (en) A kind of layering Dynamic Equivalence being applicable to area power grid
CN105046588B (en) A kind of calculation method of the improvement direct current Dynamic Optimal Power Flow Problem based on network loss iteration
CN103106314A (en) Time sequence probability modeling method for output power of solar photovoltaic power supply
CN103701116A (en) Novel intelligent electric quantity balance method adopting power real-time operation
CN102163845B (en) Optimal configuration method of distributed generations (DG) based on power moment algorithm
CN110688744A (en) Asynchronous distributed state estimation method applied to thermoelectric coupling network
CN103280845B (en) Water-fire-electricity system trans-provincial peak-regulating load distribution method
CN111355241A (en) Alternating current-direct current hybrid micro-grid droop coefficient selection method based on tidal current feasible region
CN113239512A (en) Toughness-considered screening method and system for AC/DC power distribution network planning scheme
CN108039720A (en) A kind of definite method and apparatus of alternating current-direct current combined hybrid system total transfer capability
CN107221935A (en) The AC/DC Hybrid Transmission System Interval Power Flow method converted based on confidence level
CN107968399A (en) A kind of method of fast search Static Voltage Stability Region Boundary
CN108988385A (en) A kind of grid-connected optimal location choosing method of distributed generation resource
CN110323779B (en) Method and system for dynamically aggregating power of distributed power generation and energy storage device
Zhang et al. An efficient approach for robust SCOPF considering load and renewable power uncertainties
CN105896547A (en) Large power network graded voltage control method under wind power access

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20190125