CN106208036A - Parallel operation system optimal point based on efficiency and current-sharing performance area and maximum determines method - Google Patents

Abstract

The present invention relates to determine method based on efficiency and current-sharing performance area and maximum parallel operation system optimal point, present invention expression formula Γ=Ψ (i) between the expression formula η=Φ (i) obtained respectively between efficiency eta with power module load current i and current-sharing standard deviation inverse Γ and power module load current i and corresponding optimum pointWithOn the basis of.With the area of η=Φ (i) and Γ=Ψ (i) with as object function, ask forWithBetween electric current I_refSo thatThere is reliability high, the feature such as practical；Can effectively take into account parallel operation system current-sharing performance and efficiency index, improve performance driving economy and the reliability of system, provide Reliable guarantee for parallel operation security of system, Effec-tive Function.

Description

Parallel operation system optimal point based on efficiency and current-sharing performance area and maximum determines Method

Technical field

The present invention relates to parallel operation system optimal point based on efficiency and current-sharing performance area and maximum and determine method, energy Quickly determine efficiency and the current-sharing integrated performance index optimal working point of parallel operation system, realize high for parallel operation system Efficiency and current-sharing composite of performance index optimal control provide and support, when the method is equally applicable to other electronic equipment parallel runnings Efficiency and the determination of current-sharing integrated performance index optimal working point.

Background technology

High-power parallel operation power supply its be multiple power module Parallel opertation structures, due to possess compatible strong, can N+m A series of advantages such as redundancy backup, highly reliable, cost performance is high, design difficulty is relatively low, be easily managed, become solution high-power defeated One of preferred option going out Power Management Design, equal Flow Technique has become the core technology of parallel operation.All Flow Technique refer to multiple During power module parallel operation, on the premise of meeting output voltage stable state accuracy and dynamic response, there is the uniform of degree of precision Distribute each power module load current.So, the height of parallel operation system current-sharing performance is directly connected to machine system Safe and reliable and high performance operation.

Owing to parallel operation system load electric current has time variation and randomness, cause using tradition sharing control scheme (i.e. on-line operation power module quantity is constant, and the output electric current being regulated each power module by sharing control algorithm reaches equal Stream target and load coupling target scheme) parallel operation system in power module working range contain underloading, semi-load, specified The operating modes such as load and overload.On the one hand, when under different loads operating mode, parallel operation system is run, its system current-sharing performance exists one Determine difference, it is impossible to guarantee that system all has higher current-sharing performance under different loads current conditions；On the other hand, power module In the case of different loads, its work efficiency is the most different, can not guarantee that system is respectively provided with more efficient under different loads operating mode Rate.It is therefore desirable to a kind of method determines parallel operation system effectiveness and current-sharing composite of performance index optimal working point.

Existing parallel operation system sharing control strategy can guarantee that parallel operation system load electric current all lineman It is distributed equally as power module.But there is following two problem: one, can not realize at parallel operation system current-sharing performance In preferable state；Two, parallel operation system can not realize higher efficiency.In order to realize parallel operation system in different loads feelings Efficiency and current-sharing effect integrated performance index under condition, primary prerequisite is to must determine parallel operation system effectiveness and current-sharing effect The really operating point under integrated performance index optimal situation, it is parallel operation system effectiveness and the control of current-sharing overall performance optimization Premise.

But, find by retrieving existing paper and patent, not yet find a kind of reliable and practical parallel operation system System operating point when system optimum point determines method for determining efficiency and current-sharing combination property optimum.Thus parallel operation to be realized System effectiveness and the optimal control of current-sharing performance, a kind of take into account efficiency and current-sharing performance and reliable and practical parallel operation system System optimal working point determines that method is just particularly important, and it optimizes reliability service for parallel operation system and has important Impact.

Summary of the invention

It is an object of the invention to overcome above-mentioned weak point, it is proposed that based on efficiency and current-sharing performance area and maximum Parallel operation system optimal point determines method.

The technical scheme is that a kind of based on efficiency and current-sharing performance area with maximum parallel operation system optimal Point determines method, and its step is as follows:

(1) the parallel operation system load electric current I for interval time, K power module formed with cycle T_outAccording to step Input isEtc. up regulation；I & lt program-control electronic load current value is labeled asCorresponding i-th Secondary source module current-sharing desired value is labeled asI is that Current electronic loads number of times；

(2) with cycle T_sFor interval, parallel operation system power supply module output current, output voltage and input power are entered Row gathers；

(3) the power module output current array of the parallel operation system being made up of K × U × V element is set up {Data_curr(m) (i) (j) }, power module output voltage array { Data_volt(m) (i) (j) } and power module input power number Group { P (m) (i) (j) }, wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；K is parallel operation system power supply mould Number of blocks, it is the positive integer more than 1；U is the regulation number of times of program-control electronic load operating current；V is electronic load each time In the case of need to gather power module output current, output voltage and the number of times of input power, it is the positive integer more than 1；M is for working as Front power module sequence number, i is that Current electronic loads number of times, and j is current times of collection.

(4) power module obtaining serial number m is I at current-sharing expectation electric current_refRelative deviation time (i)Wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；

(5) power module obtaining serial number m is I at current-sharing expectation electric current_refThe expectation of δ (m) (i) (j) time (i)

(6) standard deviation of power module δ (m) (i) (j) of serial number m is obtained

(7) obtaining K power module at current-sharing expectation electric current is I_refMean standard deviation time (i)And Corresponding inverse

(8) to U data pointI ∈ [1, U] carries out processing the relation Γ=Ψ (i) drawing between Γ and i；

(9) in allowing output current scope, obtain and meetMaximum

(10) the power module efficiency of serial number m is obtained Wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；

(11) mathematic expectaion of power module η (m) (i) (j) of serial number m is obtainedWherein m =1,2 ... K, i=1,2 ... U；

(12) to U data pointI ∈ [1, U] carries out process and draws efficiency eta and power module load current i Between relation η=Φ (i)；

(13) in allowing output current scope, obtain and meetMaximum

(14) obtainWithBetween meetMaximum electric current I_ref:Wherein:?WithBetween, this I_refValue is The efficiency of parallel operation system and current-sharing integrated performance index optimal working point.

In step (2).The jth electricity that the power module of m-th sequence number is sampled in the case of i & lt electronic load current Flow data is labeled as Data_curr(m)(i)(j)；The power module of m-th sequence number is adopted in the case of i & lt electronic load current The jth voltage data of sample is labeled as Data_volt(m)(i)(j)；By the power module of m-th sequence number at i & lt electronic load The jth input power data markers of current conditions down-sampling is P (m) (i) (j)；By the power module of m-th sequence number i-th The jth efficiency calculated in the case of secondary electronic load current is designated as η (m) (i) (j)；By the power module of m-th sequence number i-th In the case of secondary electronic load current, efficiency meansigma methods is labeled as η_mi；By the power module of m-th sequence number at i & lt electronic load electricity The jth current data of sampling and power module current-sharing desired value I in the case of stream_refI () relative deviation is labeled as δ (m) (i) (j)；By the power module of m-th sequence number current data in the case of i & lt electronic load current and power module current-sharing mesh The mathematic expectaion of scale value relative deviation is labeled as E_mi；By the power module of m-th sequence number in the case of i & lt electronic load current The standard deviation of current data and power module current-sharing desired value relative deviation be labeled as S_mi；

In step (8), application fitting of a polynomial, curve matching, interpolating method are to U data pointi∈[1,U] Process.

In step (12), application fitting of a polynomial, curve matching, interpolating method are to U data pointi∈[1, U] process.

The principle of the present invention mainly comprises with lower part: first, obtain parallel operation system power supply module average efficiency η with Expression formula η=the Φ (i) of power module load current i, and ask for load current corresponding during Φ (i) maximumSecondly, obtain Take the expression formula Γ=Ψ between parallel operation system power supply module mean standard deviation inverse Γ and power module load current i (i), and ask for load current corresponding during Ψ (i) maximumFinally, existWithBetween ask for optimal current I_refMeet Area andOwing to its characteristic of power module of same size totally keeps consistent, thus Existed by the parallel operation system measuring K (the big I of K is determined, it is 10 that K of the present invention fixes tentatively) individual power module composition by user Current-sharing performance indications under different loads electric current can obtain the parallel operation system of any N number of power module composition and bear in difference Current-sharing performance indications in the case of load.

Present invention have the advantage that

(1) present invention covers load current full operating range operating mode, there is wide applicability；

(2) present invention can comprehensively take into account parallel operation system effectiveness and current-sharing performance indications, have significant economy and System reliability；

(3) present invention is in the expression formula η=Φ (i) obtained respectively between efficiency eta and power module load current i and current-sharing Expression formula Γ=Ψ (i) between standard deviation inverse Γ with power module load current i and corresponding optimum pointWithBase On plinth.With the area of η=Φ (i) and Γ=Ψ (i) with as object function, ask forWithBetween electric current I_refSo thatWherein:?WithBetween.This value characterizes During parallel operation system current-sharing efficiency and current-sharing integrated performance index optimum and corresponding at wire module load current value, Foundation is provided for parallel operation system effectiveness and current-sharing optimal control.

(4) of the present invention based on efficiency and current-sharing performance area with the maximum parallel operation system optimal point side of determination The features such as it is high that method has reliability, practical；Can effectively take into account parallel operation system current-sharing performance and efficiency index, improve system The performance driving economy of system and reliability, provide Reliable guarantee for parallel operation security of system, Effec-tive Function.

Accompanying drawing explanation

Fig. 1 is parallel operation system effectiveness and current-sharing comprehensive performance testing system structure chart.

Fig. 2 is efficiency and current-sharing combination property area and schematic diagram.

Detailed description of the invention

Below for accompanying drawing, embodiments of the invention are described further:

The invention provides parallel operation system optimal point based on efficiency and current-sharing performance area and maximum and determine method. Fig. 1 show parallel operation system effectiveness and current-sharing combination property optimal working point test system structure figure, and Fig. 2 is that efficiency is with equal Stream combination property area and schematic diagram.Fig. 1 function is the functional relationship η=Φ obtaining parallel operation system effectiveness with load current (i) and the module current-sharing standard deviation mathematical relationship Γ=Ψ (i) with load current reciprocal, and determine respective optimal load electricity StreamWithOn this basis, with the area of η=Φ (i) and Γ=Ψ (i) with as object function, ask forWithBetween Electric current I_refSo thatSo that it is determined that efficiency and current-sharing Load current I during combination property optimum_ref.Fig. 1 mainly includes host computer (PC), program-control electronic load, power module, merit Rate meter etc..Host computer (PC) major function is for obtaining online module I P address, input power, module output current, output work Rate, control program-control electronic load operating current, calculating η=Φ (i), Γ=Ψ (i) and optimal load electric current I_ref；Program-controlled electronic Load is for regulating the load current of parallel operation system；Power module mainly realizes receiving IP and sets, receives host computer order Data export electric current, output to host computer with uploading；Energy meter is mainly used in measuring the input power of online power module. Fig. 2 gives I_refPlace is interval wherein guarantees efficiency and current-sharing combination property area and maximum schematic diagram.Current-sharing regulatory function Realization with or without communication bus autonomous equalizing current mode and have the equal stream mode of communication bus, special flow equalizing function module realize, The present invention does not repeats.

Parallel operation system effectiveness and current-sharing comprehensive performance testing system variable declaration are as follows: K is that parallel operation tests system Power module quantity, the occurrence of K can set according to practical situation.I_NFor power module rated current；For parallel operation system System output-current rating, meetsU is load current point quantity, i.e. parallel operation system load electric current I_outFromAccording to being spaced apartEquidistantly change to(containing underloading, semi-load, specified load and overload conditions, U must Must be the positive integer not less than 20, user can determine according to the maximum load current value of system work)；For electronics Output electric current when being supported at i-th, wherein: U >=i >=1；M is power module sequence number, meets: the IP of K power module is according to from little It is mapped as m=1 to big order, 2 ... K, i.e. m=1 are the power module sequence number that IP is minimum, and m=2 is IP minimal power module Sequence number ..., m=K is the power module sequence number that IP is maximum by that analogy；V is that parallel operation system is in a certain load current point Time need to be to when single online power module output current, output voltage and input power data sampling quantity, V can be according to actual need It is sized.Data_currM () (i) (j), (K >=m >=1, U >=i >=1, V >=j >=1) is that the power module of serial number m existsUnder the conditions of jth current sampling data；Data_volt(m) (i) (j), (K >=m >=1, U >=i >=1, V >=j >=1) Power module for serial number m existsUnder the conditions of jth output voltage sampled data；P (m) (i) (j), (K >=m >=1, U >=i >=1, V >=j >=1) it is that the power module of serial number m existsUnder the conditions of jth input power hits According to；η (m) (i) (j), (K >=m >=1, U >=i >=1, V >=j >=1) is that the power module of serial number m existsUnder the conditions of The jth efficiency data calculated, meets:η_mi Power module for serial number m existsUnder the conditions of the mathematic expectaion of V η (m) (i) (j), meet:I_refI () is that power module existsUnder the conditions of current-sharing target reference, meet:Wherein: U >=i >=1；η_iFor K power module at current-sharing expectation electric current it isOperating mode under Average efficiency, meets:δ (m) (i) (j) is that the power module of serial number m existsUnder the conditions of jth Individual sample rate current and current-sharing reference target electric currentRelative standard deviation values, meet:E_miPower module for serial number m existsUnder the conditions of V δ (m) I the mathematic expectaion of () (j), meets:S_miPower module for serial number m existsBar The standard deviation of V δ (m) (i) (j) under part, meets:S_iFor K power module mark Quasi-deviation average, meetsΓ_iFor S_iInverse, meet:

Definition t=0 is the last moment of parallel operation system no-load running；T is adjacent two load current interval times； Then t ∈ ((i-1) T, iT], (U >=i >=1) is parallel operation system load electric currentThe operation time.Due toRunning needs each power module is gathered 3V sample data, thus, host computer need to gather 3 altogether × K × V data.The time assuming one data of host computer collection is T₁, then system works inState needs T_total=3 × K × V × T₁Time, thus it must is fulfilled for T >=T_total.Again due to current-sharing performance data reliability and sampling number With sampling time T₁Relevant, thus T and T need to be considered according to the actual requirements₁Size, it is ensured that the reliability of current-sharing performance indications.

First, from controlling engineering knowledge, evaluate the overshoot that the performance of system can be responded by system step, adjust Time and steady-state deviation index are weighed.Thus, parallel operation system electronic load byStep isTime, we again may be by measuring the electric current of power module and export between current-sharing target reference Dynamic response evaluates the current-sharing performance of power module.From mathematical statistics knowledge, parallel system current-sharing standard deviation characterizes Be system current-sharing dynamic response process relative overshoot size, embody the collection of output electric current during its current-sharing step response Moderate, can reflect power module current-sharing performance indications；Secondly, parallel operation system, should while meeting current-sharing performance indications This takes into account the economic benefit that system is run；Finally, by ask for expression formula η between efficiency eta and power module load current i= Expression formula Γ=Ψ (i) and corresponding optimum between Φ (i) and standard deviation inverse Γ with power module load current i are born Carry electric currentWithOn the basis of, with the area of η=Φ (i) and Γ=Ψ (i) with as object function, ask forWithBetween Electric current I_refSo thatSo that it is determined that efficiency and current-sharing Load current I during combination property optimum_ref, its physical significance shows that parallel operation system is in efficiency under which kind of load current Best with current-sharing combination property.

T ∈ ((i-1) T, iT], (U >=i >=1), electronic load current isThe then current-sharing of power module Target reference current is:

$I_{ref}\left(i\right)=\frac{\frac{i}{20}{I}_N^p}{K}=\frac{i}{20}{I}_N,U\≥i\≥1,---\left(1\right)$

Obtain the power module output current sampled data data of serial number m: Data_curr(m) (i) (j), (K >=m >=1, U >=i >=1, V >=j >=1), thus, its current-sharing relative deviation δ (m) (i) (j) is:

$\mathrm{\δ}\left(m\right)\left(i\right)\left(j\right)=\frac{{\mathrm{Data}}_{curr}\left(m\right)\left(i\right)\left(j\right)-I_{ref}\left(i\right)}{I_{ref}\left(i\right)},---\left(2\right)$

The power module asking for serial number m existsUnder the conditions of relative deviation δ (m) (i) (j) about the mathematics of j Expect E_miFor:

$E_{mi}=\frac{1}{V}\underset{j=1}{\overset{V}{\Σ}}\mathrm{\δ}\left(m\right)\left(i\right)\left(j\right),---\left(3\right)$

The power module asking for serial number m existsUnder the conditions of relative deviation δ (m) (i) (j) about the standard of j Deviation S_miFor:

$S_{mi}=\sqrt{\frac{1}{V-1}\underset{j=1}{\overset{V}{\Σ}}{\left(\mathrm{\δ}\right(m\left)\right(i\left)\right(j)-E_{mi})}^2},---\left(4\right)$

S_miPhysical significance be: the power module of serial number m existsUnder the conditions of the standard deviation of relative deviation Difference, S_miThe least show power moduleUnder the conditions of current-sharing concentration degree performance the best.

Calculating K power module at current-sharing expectation electric current isTime mean standard deviation:

$S_i=\frac{1}{K}\underset{m=1}{\overset{K}{\Σ}}{S}_{mi},---\left(5\right)$

Calculate S_iΓ reciprocal_i, meet:

${\mathrm{\Γ}}_i=\frac{1}{S_i},---\left(6\right)$

Γ_iPhysical significance be: Γ_iShow the most greatly power moduleUnder the conditions of current-sharing concentration degree performance The best.

Application Related Computational Methods (such as fitting of a polynomial, curve matching, interpolating method etc.) is to U data pointCarry out processing the expression formula drawn between Γ and power module load current i:

Γ=Ψ (i), (7)

In allowing output current scope, solve load currentMeet:

$\mathrm{\Ψ}\left(I_{ref}^2\right)\≥\mathrm{\Ψ}\left(i\right),---\left(8\right)$

The power module asking for serial number m existsCondition efficiency eta (m) (i) (j) is:

$\mathrm{\η}\left(m\right)\left(i\right)\left(j\right)=\frac{{\mathrm{Data}}_{curr}\left(m\right)\left(i\right)\left(j\right)\×{\mathrm{Data}}_{volt}\left(m\right)\left(i\right)\left(j\right)}{P\left(m\right)\left(i\right)\left(j\right)}\×100\%,---\left(9\right)$

The power module asking for serial number m existsUnder the conditions of η (m) (i) (j) about mathematic expectaion η of j_mi For:

${\mathrm{\η}}_{mi}=\frac{1}{V}\underset{j=1}{\overset{V}{\Σ}}\mathrm{\η}\left(m\right)\left(i\right)\left(j\right),---\left(10\right)$

η_miPhysical significance be: the power module of serial number m existsUnder the conditions of the meansigma methods of efficiency, η_mi Show the most greatly power moduleUnder the conditions of economic performance the best, the most energy-conservation；

Calculating K power module at current-sharing expectation electric current isOperating mode under average efficiency:

${\mathrm{\η}}_i=\frac{1}{K}\underset{m=1}{\overset{K}{\Σ}}{\mathrm{\η}}_{mi},---\left(11\right)$

Application Related Computational Methods (such as fitting of a polynomial, curve matching, interpolating method etc.) is to U data pointCarry out processing the expression formula drawn between efficiency eta and power module load current i:

η=Φ (i), (12)

In allowing output current scope, solve load currentMeet:

$\mathrm{\Φ}\left(I_{ref}^1\right)\≥\mathrm{\Φ}\left(i\right),---\left(13\right)$

With the area of η=Φ (i) and Γ=Ψ (i) with as object function, calculate optimal load electric current I_ref, meet:

$\left(\right|{\∫}_{I_{ref}^1}^{I_{ref}}\mathrm{\Φ}\left(i\right)di|+|{\∫}_{I_{ref}}^{I_{ref}^2}\mathrm{\Ψ}\left(i\right)di\left|\right)\≥\left(\right|{\∫}_{I_{ref}^1}^i\mathrm{\Φ}\left(i\right)di|+|{\∫}_i^{I_{ref}^2}\mathrm{\Ψ}\left(i\right)di\left|\right),---\left(14\right)$

Wherein: I_refIt is inWithBetween,It is inWithBetween；

I_refPhysical significance be: the parallel operation system effectiveness being made up of K power module and current-sharing combination property are optimum Time load current.

The invention provides parallel operation system optimal point based on efficiency and current-sharing performance area and maximum and determine method, Comprise the steps:

(1) the parallel operation system load electric current I for interval time, K power module formed with cycle T_outAccording to step Input isEtc. up regulation；First time program-control electronic load current value is labeled asCorresponding the One secondary module current-sharing desired value is labeled asCurrent electronic load number of times is i, makes i=1；

(2) with cycle T_sFor interval, parallel operation system power supply module output current, output voltage and input power are entered Row gathers.First current data mark that the power module of first sequence number is sampled in the case of first time electronic load current It is designated as Data_curr(1)(1)(1)；That the power module of first sequence number is sampled in the case of first time electronic load current One voltage data is labeled as Data_volt(1)(1)(1)；By the power module of first sequence number in first time electronic load current In the case of sampling first input power data markers be P (1) (1) (1)；By the power module of first sequence number for the first time First efficiency calculated in the case of electronic load current is designated as η (1) (1) (1)；By the power module of first sequence number first In the case of secondary electronic load current, efficiency meansigma methods is labeled as η₁₁；By the power module of first sequence number at first time electronic load First current data of current conditions down-sampling and module current-sharing desired value I_ref(1) relative deviation is labeled as δ (1) (1) (1)； By the power module of first sequence number current data in the case of first time electronic load current and module current-sharing desired value phase The mathematic expectaion of deviation is labeled as E₁₁；By the power module of first sequence number electricity in the case of first time electronic load current Flow data is labeled as S with the standard deviation of module current-sharing desired value relative deviation₁₁；Current power module serial number m, makes m=1； Current electronic load number of times is i, makes i=1；The current current times that gathers is j, makes j=1；

(3) the parallel operation system module output current several group { Data being made up of K × U × V element is set up_curr(m) (i) (j) }, power module output voltage array { Data_volt(m) (i) (j) } and power module input power array { P (m) (i) (j) }, wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；K is parallel operation system power supply module number, and it is Positive integer more than 1；U is the regulation number of times of program-control electronic load operating current.Underloading, semi-load, specified is covered for satisfied evaluation Load and overload situations, the value of U is more than 20；V exports electric current, output voltage for needing acquisition module in the case of electronic load each time With the number of times of input power, it is the positive integer more than 1；M is current power module sequence number, and i is that Current electronic loads number of times, j For current times of collection.

(4) power module solving serial number m is I at current-sharing expectation electric current_refRelative deviation time (i)Wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；

(5) power module solving serial number m is I at current-sharing expectation electric current_refThe expectation of δ (m) (i) (j) time (i)

(6) standard deviation of power module δ (m) (i) (j) of serial number m is solved

(7) calculating K power module at current-sharing expectation electric current is I_refMean standard deviation time (i)And Corresponding inverse

(8) application Related Computational Methods (such as fitting of a polynomial, curve matching, interpolating method etc.) is to U data pointI ∈ [1, U] carries out processing the expression formula Γ=Ψ (i) drawing between Γ and i；

(9) in allowing output current scope, solveMeetMaximum；

(10) the power module efficiency of serial number m is solved Wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；

(11) mathematic expectaion of power module η (m) (i) (j) of serial number m is solvedWherein m =1,2 ... K, i=1,2 ... U；(η(m)(i)(j),η_miShow that the most greatly the efficiency performance of module is the best)；

(12) application Related Computational Methods (such as fitting of a polynomial, curve matching, interpolating method etc.) is to U data pointI ∈ [1, U] carries out processing the expression formula η=Φ (i) drawn between efficiency eta and power module load current i；

(13) in allowing output current scope, solveMeetMaximum；

(14) ask forWithBetween electric current I_ref, meetIt is maximum, it may be assumed thatWherein:?WithBetween.This I_refValue is The efficiency of parallel operation system and current-sharing integrated performance index optimal working point；

(15) parallel operation system optimal point determines end.

Embodiment is not construed as the restriction invented, but any spiritual improvements introduced based on the present invention, all Ying Ben Within the protection domain of invention.

Claims (4)

1. parallel operation system optimal point based on efficiency and current-sharing performance area and maximum determines a method, and its feature exists In: its step is as follows:

(1) the parallel operation system load electric current I for interval time, K power module formed with cycle T_outAccording to stepping-in amount ForEtc. up regulation；I & lt program-control electronic load current value is labeled asCorresponding i & lt electricity Source module current-sharing desired value is labeled asI is that Current electronic loads number of times；

(2) with cycle T_sFor interval, parallel operation system power supply module output current, output voltage and input power are adopted Collection；

(3) the power module output current array { Data of the parallel operation system being made up of K × U × V element is set up_curr(m) (i) (j) }, power module output voltage array { Data_volt(m) (i) (j) } and power module input power array { P (m) (i) (j) }, wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；K is parallel operation system power supply module number, and it is Positive integer more than 1；U is the regulation number of times of program-control electronic load operating current；V is for needing in the case of electronic load each time to gather The number of times of power module output current, output voltage and input power, it is the positive integer more than 1；M is current power module sequence Number, i is that Current electronic loads number of times, and j is current times of collection.

(4) power module obtaining serial number m is I at current-sharing expectation electric current_refRelative deviation time (i)Wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；

(5) power module obtaining serial number m is I at current-sharing expectation electric current_refThe expectation of δ (m) (i) (j) time (i)

(6) standard deviation of power module δ (m) (i) (j) of serial number m is obtained

(7) obtaining K power module at current-sharing expectation electric current is I_refMean standard deviation time (i)And it is corresponding Inverse

(8) to U data pointI ∈ [1, U] carries out processing the relation Γ=Ψ (i) drawing between Γ and i；

(9) in allowing output current scope, obtain and meetMaximum

(10) the power module efficiency of serial number m is obtained Wherein m=1,2 ... K, i=1,2 ... U, j=1,2 ... V；

(11) mathematic expectaion of power module η (m) (i) (j) of serial number m is obtainedWherein m=1, 2 ... K, i=1,2 ... U；

(12) to U data pointI ∈ [1, U] carries out process and draws between efficiency eta and power module load current i Relation η=Φ (i)；

(13) in allowing output current scope, obtain and meetMaximum

(14) obtainWithBetween meetMaximum electric current I_ref:Wherein:?WithBetween, this I_refValue is The efficiency of parallel operation system and current-sharing integrated performance index optimal working point.

Parallel operation system optimal point based on efficiency and current-sharing performance area and maximum the most according to claim 1 determines Method, it is characterised in that: in step (2).The power module of m-th sequence number is sampled in the case of i & lt electronic load current Jth current data be labeled as Data_curr(m)(i)(j)；By the power module of m-th sequence number at i & lt electronic load electricity In the case of stream, the jth voltage data of sampling is labeled as Data_volt(m)(i)(j)；By the power module of m-th sequence number i-th In the case of secondary electronic load current, the jth input power data markers of sampling is P (m) (i) (j)；Power supply by m-th sequence number The jth efficiency that module calculates in the case of i & lt electronic load current is designated as η (m) (i) (j)；Power supply by m-th sequence number Module efficiency meansigma methods in the case of i & lt electronic load current is labeled as η_mi；By the power module of m-th sequence number in i & lt The jth current data of sampling and power module current-sharing desired value I in the case of electronic load current_refI () relative deviation is labeled as δ(m)(i)(j)；By the power module of m-th sequence number current data in the case of i & lt electronic load current and power module The mathematic expectaion of current-sharing desired value relative deviation is labeled as E_mi；By the power module of m-th sequence number in i & lt electronic load current In the case of the standard deviation of current data and power module current-sharing desired value relative deviation be labeled as S_mi。

Parallel operation system optimal point based on efficiency and current-sharing performance area and maximum the most according to claim 1 determines Method, it is characterised in that: in step (8), application fitting of a polynomial, curve matching, interpolating method are to U data pointI ∈ [1, U] processes.

Parallel operation system optimal point based on efficiency and current-sharing performance area and maximum the most according to claim 1 determines Method, it is characterised in that: in step (12), application fitting of a polynomial, curve matching, interpolating method are to U data pointI ∈ [1, U] processes.