CN102305890B - Direct-current voltage detection method of submodule of flexible direct-current transmission system - Google Patents

Abstract

The invention provides a direct-current voltage detection method of a submodule of a flexible direct-current transmission system, belonging to the technical field of direct-current detection of the flexible direct-current transmission system. In the method disclosed by the invention, circuits matched with a sampling resistor, a linear optocoupler, a high-precision operational amplifier and a high-precision analog/digital conversion chip have high topology practicalities, and can be applied to various occasions in which the direct-current voltage detection is required; and detection voltage has good filtering effect and linearity degree through software filtering processing. In the method, parameter requirement and model selection methods of various devices in the circuit are introduced, which can be used for guiding technical personnel to efficiently finish the voltage detection design of the submodule of a modularized multi-level flexible direct-current power transmission system. The method comprises sampling resistance design, linear optocoupling design, operational amplifier design, analog/digital conversion chip design and a filtering algorithm. The detection method has the advantages of high linearity degree and high precision within the voltage range from 0 to 5000V. According to the method, the cost of the whole system is lowest.

Description

A kind of DC voltage detection method of flexible DC power transmission system submodule

Technical field

The invention belongs to flexible DC power transmission system DC voltage detection technique field, specifically relate to a kind of DC voltage detection method of flexible DC power transmission system submodule.

Background technology

Technology of HVDC based Voltage Source Converter can provide good dynamic reactive to support to wind energy turbine set, avoids the reactive-load compensation equipment investment of wind energy turbine set; Excellent grid-connected performance is provided simultaneously, prevents the impact of the voltage fluctuation of wind energy turbine set on AC system, and improve the antijamming capability of wind energy turbine set to system fluctuation simultaneously.Due to voltage support effect can be provided, it can also significantly promote the low voltage ride-through capability that wind energy turbine set breaks down in situation in AC system; In addition, because flexible DC power transmission is not subject to distance limit, be also therefore the unique selection of external large-scale remote offshore grid-connected wind farm.Based on above significant advantage, flexible DC power transmission has become the wind farm grid-connected best-of-breed technology scheme of generally acknowledging in the world at present.

The submodule voltage detecting of flexible DC power transmission system is one of gordian technique of flexible DC power transmission engineering.Voltage detecting value is the basis of various algorithms, and its voltage detecting value is directly connected to capacitance balancing and controls, and system is controlled the enforcement of scheduling algorithm.

DC voltage detects, and generally considers detection voltage range, accuracy of detection, and the parameters such as the linearity, according to the parameter request of various devices, select best model, make voltage detecting scheme reach economically the optimization with detected parameters.

Traditional DC voltage detects, and general range is narrow, and ratio of precision is lower, and most of scheme adopts the devices such as Hall element, and its cost is high, and the linearity of Hall element is not high.In modular multilevel flexible DC power transmission system, the voltage range wide (all needing to provide high precision detected value from 0-5000 volt) of submodule, simultaneously all higher to the linearity and accuracy requirement, therefore traditional measuring method cannot reach modular multilevel flexible DC power transmission system for the voltage detecting parameter request of submodule.

Summary of the invention

The object of the invention is to provide a kind of submodule DC voltage detection method of modular multilevel formula flexible DC power transmission system, the method adopts sampling resistor, linear optical coupling, high-operational amplifier, the novel circuit topological that high precision mould/number conversion chip matches, by the processing of software, make to detect voltage and there is better filter effect and the linearity, detection method of the present invention has high linearity and high-precision advantage in voltage range is 0-5000V, make system-wide expense economize most simultaneously, can be used for the submodule voltage detecting design of instructing person skilled to complete efficiently modular multilevel formula flexible DC power transmission system.

For achieving the above object, the technical solution used in the present invention is:

A DC voltage detection method for flexible DC power transmission system submodule, its improvements are that described method comprises the steps:

1) circuit devcie is selected, and comprises the selection of sampling resistor, linear optical coupling, operational amplifier and mould/number conversion chip;

Sampling resistor is selected: sampling resistor resistance R is greater than the resistance of 10 times of each submodule equalizing resistances of flexible DC power transmission system, sampling resistor power

wherein U is each submodule magnitude of voltage of flexible DC power transmission system, sampling resistor withstand voltage > U, sampling resistor precision is better than the precision that flexible DC power transmission system antithetical phrase module voltage detects, sampling resistor equivalent inductance value < 10nH;

Linear optical coupling is selected: linear optocoupler nonlinearity < 0.01%, the transfer ratio error of the former limit of linear optical coupling secondary within ± 5%, isolation voltage 5kVrms/1min, bandwidth range > 1MHz;

Operational amplifier is selected: operational amplifier power < 1mW, and precision > 95%, amplifier speed > 1MHz;

Mould/number conversion chip selection: mould/number conversion chip < switching time 500ns, sample frequency > 1MHz, port number >=1, figure place >=12;

2) design of circuit;

Described method is based on sampling resistor, linear optical coupling, operational amplifier, the circuit topology that mould/number conversion chip matches;

3) extract voltage detecting value

Magnitude of voltage to 1 circuit in step 2 carries out continuous sampling, using the mean value of sample voltage value as voltage detecting value U _{inspection i};

4) filtering, correction;

The voltage detecting value U that step 3 is obtained _{inspection i}through secondary filtering and linearity parameter correction;

5) the voltage detecting value through filtering and calibration sends to higher controller.

Another optimal technical scheme of the present invention is: voltage detecting value U in described step 3 _{inspection i}testing process be:

Sample voltage value is through operational amplifier conditioning by linear optical coupling isolation, and the voltage signal that becomes 0～5V after rear class modulate circuit and filtering is sent into microprocessor and used as controlled quentity controlled variable;

In microprocessor, mould/number conversion result is provided with to 8 grades of FIFO buffer memorys, can store continuous 10 sampled result simultaneously, while depositing new data at every turn, can data the oldest in FIFO be extruded to queue simultaneously, the data of the minimax of storing in new fifo queue are rejected, and done sums on average to remaining 8 sampled result, obtain voltage detecting value U _{inspection i}.

Another optimal technical scheme of the present invention is: described step 4 neutral line degree parameter correction process is:

If straight-line equation is suc as formula (1),

Y＝A+BX (1)

Wherein: A, B are any real numbers;

X is trial voltage;

The digital quantity of Y for obtaining in theory after analog to digital conversion;

Y ^*for the real figure amount obtaining after mould/number conversion;

Application principle of least square method, by measured value Y ^*with the deviation (Y that utilizes (formula 1-1) calculated value Y=A+BX ^*-Y) quadratic sum minimum is " optimization criterion ", order:

In wushu (1) substitution formula (2):

Wherein: X _iit is the trial voltage detecting for the i time;

Y ^* _ifor what obtain, detect real figure amount the i time after mould/number conversion;

When

hour, function

a, B are asked to partial derivative, make these two partial derivatives equal zero;

That is:

$\mathrm{mA}+\left(\mathrm{\&Sigma;}{X}_i\right)B=\mathrm{\&Sigma;}{Y}_i^{*}---\left(6\right)$

$\left(\mathrm{\&Sigma;}{X}_i\right)A+\left(\mathrm{\&Sigma;}{X}_i^2\right)B=\mathrm{\&Sigma;}\left(X_i{Y}_i^{*}\right)---\left(7\right)$

The system of equations of separating (6) and (7) two equations draws:

$A=\left(\mathrm{\&Sigma;}{Y}_i^{*}\right)/m-B\left(\mathrm{\&Sigma;}{X}_i\right)/m---\left(8\right)$

$B=\left[(\mathrm{\&Sigma;}{X}_i{Y}_i^{*}-\left(\mathrm{\&Sigma;}{X}_i\mathrm{\&Sigma;}{Y}_i^{*}\right)/m)\right]/[\mathrm{\&Sigma;}{X}_i^2-{\left(\mathrm{\&Sigma;Xi}\right)}^2/m]---\left(9\right)$

At this moment in A, B substitution formula (1), obtain linear function expression formula;

Select some theoretical voltage in test specification, as work as X ₁=500V, X ₂=1000V, X ₃=1500V, X ₄=2500V, X ₅=4000V, according to actual measurement magnitude of voltage can obtain linear function expression formula coefficient A and B;

$U_c=\frac{V_{\mathrm{ref}}}{4096}*(A+\mathrm{BX})$

Wherein: U _cfor the actual DC voltage obtaining after 12 mould/number conversions;

V _refit is the reference voltage of analog to digital conversion circuit.

Owing to having adopted technique scheme, compared with prior art, the obtained beneficial effect of the present invention comprises:

1, design of integer electro-circuit practicality is high

Detection method of the present invention adopts: sampling resistor, and linear optical coupling, high-operational amplifier, the circuit topology that high precision mould/number conversion chip matches, circuit design is novel, can be applicable to the multiple occasion that needs DC voltage to detect;

2, software algorithm

By the processing of software, make to detect voltage and there is better filter effect and the linearity;

3, there is high precision, high linearity, wide region detects voltage, cheaply advantage

Detection voltage range is 0-5000V, and the nonlinearity parameter of linear optocoupler is lower than 0.01%;

4, there is directive significance

The present invention has introduced parameter request and the selection method of each device of DC voltage testing circuit targetedly, can instruct other designer to complete similar DC voltage detection scheme.

Accompanying drawing explanation

Below in conjunction with accompanying drawing, the present invention is further described.

Fig. 1 is the circuit design of DC voltage detection method;

Fig. 2 is the design cycle of DC voltage detection method;

Fig. 3 is the software algorithm of DC voltage detection method;

Reference numeral is as follows:

R1, R2---sampling resistor;

R3, R5---gain build-out resistor;

R4---current-limiting resistance;

R6, R7---impedance matching resistance;

C---submodule side DC capacitor;

C1, C2---filter capacitor;

U1, U2, U4, U5---operational amplifier;

U3---linear light is even;

U6---analog to digital converter;

U7---microprocessor;

GND---submodule side publicly;

DGND---digital circuit side publicly;

VCC---power supply;

Vin---input sample voltage;

Vout---sampling and outputting voltage;

I _f---the LED current of flowing through;

I _pD1---the photodiode current of flowing through;

I _pD2---the photodiode current of flowing through;

Linear optical coupling is inner, I _pD1=V _in/ R1, outgoing side V _out=I _pD2* R2, so, can obtain,

wherein

for the ratio amplifying parameters of linear light idol self, be generally 1.

The proportionate relationship of sampled voltage and output voltage can be obtained thus, thereby virtual voltage size can be calculated.

Embodiment

Below in conjunction with example, the present invention will be described in detail.

In order to realize the submodule voltage detecting of modular multilevel formula flexible DC power transmission system, the method need to solve sampling resistor design, linear optical coupling design, Design of Amplifiers, mould/number conversion chip design, the technical matterss such as filtering algorithm.

(1) sampling resistor design

For modular multilevel flexible DC power transmission system submodule voltage detection method, sampling resistor is one of its Primary Component.The design of sampling resistor need to consider the impacts of parameter on modular multilevel flexible DC power transmission system such as its resistance value, power, precision, equivalent inductance.The selection of the resistance of sampling resistor will be considered the equalizing resistance design of modular multilevel flexible DC power transmission system modules, and the resistance of sampling resistor will be much larger than the resistance of equalizing resistance; By system, the voltage detecting precision for submodule decides the accuracy selection of sampling resistor; The parameter of equivalent inductance need to be as far as possible little, to guarantee the sampling accuracy of system.

(2) linear optical coupling design

For modular multilevel flexible DC power transmission system submodule voltage detecting scheme, in order to realize in whole reference voltage detection, there is high linearity, need to consider the parameters such as the precision of linear optical coupling, temperature are floated, the linearity.Selecting a suitable linear optical coupling is the key that guarantees the detection scheme linearity.

(3) Design of Amplifiers

Submodule voltage, after sampling resistor, need to carry out impedance matching through operation amplifier circuit.The parameters such as the precision of operational amplifier are also one of the emphasis parameters that need to consider of circuit design.

(4) mould/number conversion chip design

Mould/the number conversion of sampled voltage is undertaken by mould/number conversion chip, need to consider the switching rate of mould/number conversion chip, sampling rate, port number, the parameters such as figure place.According to system, the specific requirement of voltage detecting is selected the mould/number conversion chip matching.

(5) filtering algorithm

Mould/number conversion chip uploads to voltage detecting value after control chip, by software filtering, magnitude of voltage is carried out to filtering and linearity correction, to reach the parameter request of system to voltage detecting.

In a word, the submodule voltage detecting of modular multilevel flexible DC power transmission system, above technology point is the key issue that needs solution in this invention.

The submodule voltage detecting scheme of a kind of modular multilevel formula flexible DC power transmission system of the present invention, comprises sampling resistor type selecting, linear optical coupling type selecting, amplifier type selecting, mould/number conversion chip type selecting, the steps such as software filtering.Design of integer electro-circuit as shown in Figure 1.

(1) sampling resistor type selecting:

1.1 sampling resistor resistance R

According to the equalizing resistance of each submodule of modular multilevel flexible DC power transmission system, sampling resistor resistance R is at the more than 10 times of voltage sharing resistance value.

1.2 sampling resistor power P

According to the magnitude of voltage U of each submodule of modular multilevel flexible DC power transmission system, according to P=U*U/R, select the power of sampling resistor.

1.3 sampling resistor withstand voltages

According to the magnitude of voltage U of each submodule of modular multilevel flexible DC power transmission system, make the withstand voltage of sampling resistor be greater than U.

1.4 sampling resistor precision

The modular multilevel flexible DC power transmission system of accuracy requirement according to to(for) submodule voltage detecting, determines the precision of sampling resistor, and the precision of sampling resistor should be better than the voltage detecting precision of system.

The parameters such as 1.5 sampling resistor equivalent inductances

Select low sense or noninductive resistance, adopt equivalent resistance inductance value < 10nH, make the parameters such as equivalent inductance minimum for the impact of voltage detecting.

(2) linear optical coupling type selecting:

2.1 linear optical coupling nonlinearities

Choose compared with the device of low nonlinearity degree parameter, as far as possible lower than 0.01% as far as possible.

2.2 linear optical coupling transfer ratios

The transfer ratio error of the former limit secondary of linear optical coupling is within ± 5%.

2.3 bandwidth

Owing to being that DC voltage detects, not strict for bandwidth requirement, general bandwidth range > 1MHz, general model all can meet.

(3) operational amplifier type selecting:

Select low-power consumption, high precision, two-forty amplifier, general operational amplifier power < 1mW, precision > 95%, and amplifier speed > 1MHz makes operational amplification circuit can meet system for the parameter request of voltage detecting.

(4) mould/number conversion chip type selecting:

Mould/number conversion chip is mainly considered the parameters such as slewing rate, sample frequency, port number, figure place.According to the voltage detecting parameter request of system, select suitable mould/number conversion chip.Conventionally mould/number conversion chip < switching time 500ns, adopts sample frequency to be greater than 1MHz, port number >=1, the model of figure place >=12.

(5) software and linearity correction

Voltage detecting value is converted to after digital signal, and control chip carries out filtering and linearity correction by Weighted Average Algorithm and least-squares algorithm to magnitude of voltage, and as shown in Figure 2, software algorithm flow process as shown in Figure 3 for hardware design flow process.

From V _inobtain V _outdetailed process as follows:

If X is trial voltage, the digital quantity of Y for obtaining in theory after analog to digital conversion, Y ^*for the real figure amount obtaining, during mutual relationship between two variablees (X, Y), conventionally can obtain a series of paired data (X after analog to digital conversion ₁, Y ₁, X ₂, Y ₂, X ₃, Y ₃... X _m, Y _m); M is number of samples, and these points, near straight line, can make this straight-line equation suc as formula (1),

Y＝A+BX (1)

Wherein: A, B are any real numbers.

For setting up straight-line equation, will determine A and B, application principle of least square method, by measured value Y ^*with the deviation (Y that utilizes (formula 1-1) calculated value Y=A+BX ^*-Y) quadratic sum

minimum is " optimization criterion ".Order:

In wushu (1) substitution formula (2):

When

hour, available functions

a, B are asked to partial derivative, make these two partial derivatives equal zero.

That is:

$\mathrm{mA}+\left(\mathrm{\&Sigma;}{X}_i\right)B=\mathrm{\&Sigma;}{Y}_i^{*}---\left(6\right)$

$\left(\mathrm{\&Sigma;}{X}_i\right)A+\left(\mathrm{\&Sigma;}{X}_i^2\right)B=\mathrm{\&Sigma;}\left(X_i{Y}_i^{*}\right)---\left(7\right)$

Obtain two two system of equations that are unknown number about A, B, separate these two system of equations and draw:

$A=\left(\mathrm{\&Sigma;}{Y}_i^{*}\right)/m-B\left(\mathrm{\&Sigma;}{X}_i\right)/m---\left(8\right)$

$B=\left[(\mathrm{\&Sigma;}{X}_i{Y}_i^{*}-\left(\mathrm{\&Sigma;}{X}_i\mathrm{\&Sigma;}{Y}_i^{*}\right)/m)\right]/[\mathrm{\&Sigma;}{X}_i^2-{\left(\mathrm{\&Sigma;Xi}\right)}^2/m]---\left(9\right)$

At this moment in A, B substitution formula (1), formula now (1) is exactly the first linear equation returning, i.e. linear function expression formula.

Select some theoretical voltage in test specification, as work as X ₁=500V, X ₂=1000V, X ₃=1500V, X ₄=2500V, X ₅=4000V, according to actual measurement magnitude of voltage

can obtain linear function expression formula coefficient A and B.

Making the digital quantity obtaining after analog to digital conversion is D _u, D _u=Y=A+BX obtains actual DC voltage and is after 12 analog to digital conversion: v _refit is the reference voltage of analog to digital conversion circuit.

According to specific exemplary embodiment, invention has been described herein.It will be apparent under not departing from the scope of the present invention, carrying out to one skilled in the art suitable replacement or revise.Exemplary embodiment is only illustrative, rather than the restriction to scope of the present invention, and scope of the present invention is defined by appended claim.

Claims (3)

1. a DC voltage detection method for flexible DC power transmission system submodule, is characterized in that described method comprises the steps:

1) circuit devcie is selected, and comprises the selection of sampling resistor, linear optical coupling, operational amplifier and mould/number conversion chip;

Sampling resistor is selected: sampling resistor resistance R is greater than the resistance of 10 times of each submodule equalizing resistances of flexible DC power transmission system, sampling resistor power

wherein U is each submodule magnitude of voltage of flexible DC power transmission system, sampling resistor withstand voltage >U, sampling resistor precision is better than the precision that flexible DC power transmission system antithetical phrase module voltage detects, sampling resistor equivalent inductance value <10nH;

Linear optical coupling is selected: linear optocoupler nonlinearity <0.01%, the transfer ratio error of the former limit of linear optical coupling secondary within ± 5%, isolation voltage 5kVrms/1min, bandwidth range >1MHz;

Operational amplifier is selected: operational amplifier power <1mW, precision >95%, amplifier speed >1MHz;

Mould/number conversion chip selection: mould/number conversion chip < switching time 500ns, sample frequency >1MHz, port number >=1, figure place >=12;

2) design of circuit;

Described method is based on sampling resistor, linear optical coupling, operational amplifier, the circuit topology that mould/number conversion chip matches;

3) extract voltage detecting value

To step 2) in the magnitude of voltage of 1 circuit carry out continuous sampling, using the mean value of sample voltage value as voltage detecting value U _{inspection i};

4) filtering, correction;

The voltage detecting value U that step 3) is obtained _{inspection i}through secondary filtering and linearity parameter correction;

5) the voltage detecting value through filtering and calibration sends to higher controller.

2. the DC voltage detection method of a kind of flexible DC power transmission system submodule as claimed in claim 1, is characterized in that voltage detecting value U in described step 3) _{inspection i}testing process be:

Sample voltage value is through operational amplifier conditioning by linear optical coupling isolation, and the voltage signal that becomes 0～5V after rear class modulate circuit and filtering is sent into microprocessor and used as controlled quentity controlled variable;

In microprocessor, mould/number conversion result is provided with to 8 grades of FIFO buffer memorys, can store continuous 10 sampled result simultaneously, while depositing new data at every turn, can data the oldest in FIFO be extruded to queue simultaneously, the data of the minimax of storing in new fifo queue are rejected, and done sums on average to remaining 8 sampled result, obtain voltage detecting value U _{inspection i}.

3. the DC voltage detection method of a kind of flexible DC power transmission system submodule as claimed in claim 1, is characterized in that described step 4) neutral line degree parameter correction process is:

If straight-line equation is suc as formula (1),

Y=A+BX （1）

Wherein: A, B are any real numbers;

X is trial voltage;

The digital quantity of Y for obtaining in theory after analog to digital conversion;

Y ^*for the real figure amount obtaining after mould/number conversion;

Application principle of least square method, by measured value Y ^*with the deviation (Y that utilizes (formula 1) calculated value Y=A+BX ^*-Y) quadratic sum minimum is " optimization criterion ", order:

In wushu (1) substitution formula (2):

Wherein: X _iit is the trial voltage detecting for the i time;

Y ^* _ifor what obtain, detect real figure amount the i time after mould/number conversion;

When

hour, function

a, B are asked to partial derivative, make these two partial derivatives equal zero;

That is:

$\mathrm{mA}+\left(\mathrm{\&Sigma;}{X}_i\right)B=\mathrm{\&Sigma;}{Y}_i^{*}---\left(6\right)$

$\left(\mathrm{\&Sigma;}{X}_i\right)A+\left(\mathrm{\&Sigma;}{X}_i^2\right)B=\mathrm{\&Sigma;}\left(X_i{Y}_i^{*}\right)---\left(7\right)$

The system of equations of separating (6) and (7) two equations draws:

$A=\left(\mathrm{\&Sigma;}{Y}_i^{*}\right)/m-B\left(\mathrm{\&Sigma;}{X}_i\right)/m---\left(8\right)$

$B=\left[(\mathrm{\&Sigma;}{X}_i{Y}_i^{*}-\left(\mathrm{\&Sigma;}{X}_i\mathrm{\&Sigma;}{Y}_i^{*}\right)/m)\right]/[\mathrm{\&Sigma;}{X}_i^2-{\left(\mathrm{\&Sigma;Xi}\right)}^2/m]---\left(9\right)$

At this moment in A, B substitution formula (1), obtain linear function expression formula;

Select some theoretical voltage in test specification, as work as X ₁=500V, X ₂=1000V, X ₃=1500V, X ₄=2500V, X ₅=4000V, according to actual measurement magnitude of voltage

can obtain linear function expression formula coefficient A and B;

$U_c=\frac{V_{\mathrm{ref}}}{4096}*(A+\mathrm{BX})$

Wherein: U _cfor the actual DC voltage obtaining after 12 mould/number conversions;

V _refit is the reference voltage of analog to digital conversion circuit.

Images