CN106787845A - A kind of Pulse rectifier low switching frequency model prediction power control algorithm - Google Patents

Abstract

The invention discloses a kind of Pulse rectifier low switching frequency model prediction power control algorithm, under low switching frequency, the power prediction value of active power and reactive power is calculated by model prediction power algorithm, and then calculate component of the optimum control amount for causing evaluation function minimum under rotating coordinate system, the modulating wave after weber compensates is tried to achieve by voltage-second balance backoff algorithm again, pulsewidth modulation strategy generating switch controlling signal is finally combined, the control of Pulse rectifier is completed.The present invention improves the control accuracy of Pulse rectifier model prediction power control algorithm, without Calculation Estimation function repeatedly, reduces the complexity of algorithm, control and sample frequency is improve, with good dynamic and steady-state behaviour；Algorithm flexibility ratio is high, can be coordinated with control algolithm using different modulation strategies, meets the demand of different application occasion.

Description

A kind of Pulse rectifier low switching frequency model prediction power control algorithm

Technical field

The present invention relates to single-phase PWM converter control system technical field in power electronics and power drives field, specifically It is a kind of Pulse rectifier low switching frequency model prediction power control algorithm.

Background technology

Pulse rectifier has energy capable of bidirectional flowing, and voltage on line side, electric current keep unity power factor, DC side The advantages of voltage constant, so it is widely used in high-power rail traction transmission system, ups power etc..At present, single-phase rectifier Device control algolithm mainly includes current indirect control, hysteretic loop current control, transient current testing and dq shaft current uneoupled controls.

In order to improve the control accuracy and dynamic property of rectifier, it is adaptable to the direct Power Control algorithm of three-phase rectifier Based on instantaneous power theory, active power and reactive power are directly controlled, pulse rectifier is reached net side unit power The performance indications such as factor, DC voltage be constant.Traditional direct Power Control algorithm using stagnant ring switch list control it is active and Reactive power, fast response time, but its switching frequency are not fixed, harmonic wave is widely distributed, be unfavorable for net side wave filter designs, so The algorithm is seldom used in real system；In this regard, the power control algorithm based on model prediction is proposed, the pre- measurement of power of conventional model Rate control algolithm selects suitable on off state by calculating evaluation function corresponding with each on off state is compared repeatedly, And then generate the gate pole control signal of rectifier；But the switching frequency of conventional model PREDICTIVE CONTROL is uncertain, harmonic wave distribution is wide It is general, constrain its application in Practical Project.

To solve the model prediction unfixed problem of direct Power Control switching frequency, there is scholar to propose that two vector models are pre- Survey direct Power Control, fixed-frequency control realized by optimizing duty ratio method, however the algorithm still need calculate repeatedly and Compare the evaluation function size under different on off states, therefore increased the amount of calculation of algorithm, to overcome Model Predictive Control meter The big shortcoming of calculation amount, related scholar proposes the model prediction power control strategy based on optimal modulation ripple, is adjusted with reference to pulse width Frequency Model Predictive Control is determined in system (PWM), realization.Although the algorithm can realize the Model Predictive Control of fixed switching frequency, so And the Model Predictive Control under low switching frequency still cannot be realized, control accuracy will rapidly be disliked with the reduction of switching frequency Change.

The content of the invention

Regarding to the issue above, it is an object of the invention to propose a kind of low switching frequency model prediction power control algorithm, The algorithm, using model prediction thought, directly calculates optimal modulation function, by voltage-second balance without Calculation Estimation function repeatedly The weber energy imbalance that backoff algorithm causes to low switching frequency is compensated, and then realizes that high-precision model prediction is direct Power Control, while reducing the complexity of control algolithm.Technical scheme is as follows：

A kind of Pulse rectifier low switching frequency model prediction power control algorithm, comprises the steps of：

Step 1：The instantaneous value of active power and reactive power in next dutycycle update cycle is estimated by formula (1)：

Wherein, T_sIt is switch periods, n represents half switch periods (T_s/ number 2), ω is line voltage angular frequency, u_abdWith u_abqD axle component and q axle component of the rectifier input voltage under d-q synchronous rotating frames are represented respectively；u_mRepresent net side electricity Pressure amplitude value, L is net side inductance value；A₁And A₂It is initial parameter, by n-th half switch period start time active-power Ps and idle The instantaneous value of power Q determines that it is expressed as

Wherein, T_cControlling cycle is represented, k is represented closest to n-th controlling cycle numbers of half switch periods；

Step 2：Defining evaluation function J is：

J={ P_ref[(n+1)T_s/2]-P[(n+1)T_s/2]}²+{Q_ref[(n+1)T_s/2]-Q[(n+1)T_s/2]}² (3)

Wherein, P_ref[(n+1)T_s/ 2] and Q_ref[(n+1)T_s/ 2] reference being respectively in (n+1)th half switch periods has The set-point of work(and reactive power；P[(n+1)T_s/ 2] and Q [(n+1) T_s/ 2] it is the wattful power in (n+1)th half switch periods The predicted value of rate and reactive power, these power prediction values are solved by formula (1) and obtained；

Step 3：To cause evaluation function J minimum, then u_abdAnd u_baqMeet formula (4)：

Formula (1) and formula (3) are substituted into formula (4), obtains causing the optimum control amount u that evaluation function is minimum_abαIn rotational coordinates Component u under system_abdWith u_abqCalculating formula be：

Wherein：u_abd(nT_s/ 2) and u_abq(nT_s/ 2) represent lower half switch periods internal modulation ripple in synchronous rotating frame (d-q) the optimum control component used under；T_cControlling cycle is represented, it is close to n-th half switch periods；

Step 4：By rotating coordinate transformation, by optimal controlled quentity controlled variable u_abdAnd u_abqUnder convert to static coordinate system (alpha-beta) α axle components, realize that voltage-second balance under low switching frequency is compensated by such as formula (6)：

Wherein, u_abα ^*It is by the modulating wave after voltage-second balance compensation；

Step 5：Modulating wave is modulated, by u_abα ^*Optimum control pulse is converted to be controlled rectifier.

The beneficial effects of the invention are as follows：Algorithm of the invention may operate at low switching frequency occasion, be thought using model prediction Think, realize the direct Power Control of traction rectifier device, improve Pulse rectifier model prediction power control algorithm Control accuracy；Without Calculation Estimation function repeatedly, the complexity of algorithm is reduced, control and sample frequency are improve, with good Good dynamic and steady-state behaviour；Algorithm flexibility ratio is high, can be coordinated with control algolithm using different modulation strategies, and meeting difference should With the demand of occasion.

Brief description of the drawings

Fig. 1 is that single phase model predicts direct Power Control system function division block diagram.

Fig. 2 is single-phase phase-locked loop system.

Fig. 3 is that single-phase instantaneous power estimates block diagram.

Specific embodiment

Technical scheme and technique effect are done further specifically with specific embodiment below in conjunction with the accompanying drawings.

Fig. 1 shows single-phase no phase-locked loop direct Power Control system function division block diagram, and whole system can be divided into optimal Controlled quentity controlled variable is calculated, and weber compensates and rotating coordinate transformation, pulsewidth modulation strategy, voltage on line side current acquisition, single-phase phase-locked loop, wink When power calculation, voltage PI outer shrouds control seven parts.The particular content of wherein major part is：

(1) optimum control amount is calculated：By Model Predictive Control thought, optimal solution is asked for power estimation function, pass through It is zero to the local derviation of controlled quentity controlled variable to make evaluation function, calculates optimum control amount.

(2) voltage-second balance compensation and rotating coordinate transformation：By voltage-second balance computing formula, to traditional rotating coordinate transformation Compensate, realize the weber compensation under low switching frequency, improve control accuracy.

(3) pulse width modulation：Modulating wave is modulated, different pulse trains is generated based on voltage-second balance principle, Driving switch pipe, is allowed to be switched on or off according to specified rule.

(4) single-phase phase-locked loop：By second order improper integral algorithm, generation net positive pressure hands over coordinate components, and then realizes single-phase System phase-lock-loop algorithm, obtains voltage on line side phase and amplitude information, while netting pressure quadrature component will be used for single-phase instantaneous work( Instantaneous power is calculated in rate algorithm for estimating.

(5) instantaneous power is calculated：The voltage on line side orthogonal vectors and rotational coordinates obtained by single-phase phase-locked loop module The orthogonal modulation wave vector that conversion module is obtained, calculates current on line side quadrature component, and calculate single-phase according to instantaneous power theory The instantaneous power of system.

Under low switching frequency, by model prediction power algorithm and voltage-second balance backoff algorithm, with reference to pulsewidth modulation plan Switch controlling signal is slightly generated, the control of Pulse rectifier is completed, specifically comprised the steps of：

Step 1：Estimate active power and reactive power in next duty by high-precision power prediction algorithm shown in formula (1) Than the instantaneous value of update cycle.

Wherein：T_sIt is switch periods, n represents half switch periods (T_s/ number 2), ω is line voltage angular frequency, u_abdWith u_abqD axle component and q axle component of the rectifier input voltage under rotating coordinate system are represented respectively.u_mVoltage on line side amplitude is represented, L is net side inductance value.A₁And A₂It is initial parameter, by n-th half switch period start time active-power Ps and reactive power Q Instantaneous value determines that it is represented by

Wherein, T_cControlling cycle is represented, k is represented closest to n-th controlling cycle numbers of half switch periods.

Step 2：To weigh the combination property of control algolithm, defining evaluation function J is

J={ P_ref[(n+1)T_s/2]-P[(n+1)T_s/2]}²+{Q_ref[(n+1)T_s/2]-Q[(n+1)T_s/2]}² (3)

Wherein P_ref[(n+1)T_s/ 2] and Q_ref[(n+1)T_s/ 2] reference being respectively in (n+1)th half switch periods has The set-point of work(and reactive power, the value and currency approximately equal.P[(n+1)T_s/ 2] and Q [(n+1) T_s/ 2] it is (n+1)th The predicted value of active power and reactive power in individual half switch periods, can be by high-precision power prediction algorithm shown in formula (1) Can be in the hope of.

Step 3：To cause that evaluation function is minimum, the u for being used_abdAnd u_baqShould meet：

And then, formula (1) and formula (3) are substituted into formula (4), obtain causing the optimum control amount u that evaluation function is minimum_abαIn rotation Turn the component u under coordinate system_abdWith u_abqCalculating formula be：

Wherein：u_abd(nT_s/ 2) and u_abq(nT_s/ 2) represent lower half switch periods internal modulation ripple in synchronous rotating frame (d-q) the optimum control component used under；T_cControlling cycle is represented, k represents the kth secondary control cycle, and it is close to n-th half Switch periods；Net pressure angular frequency is estimated to obtain by software phase-lock loop.

Step 4：By rotating coordinate transformation, by optimal controlled quentity controlled variable u_abdAnd u_abqUnder convert to static coordinate system (alpha-beta) α axle components, by the weber compensation policy as shown in formula (6), realize the voltage-second balance of pulsewidth modulation under low switching frequency.

Wherein, u_abα ^*It is the modulating wave after weber compensates.

Step 5：By pulsewidth modulation strategy, different pulse trains are generated based on voltage-second balance principle, driving switch pipe, It is allowed to be switched on or off according to specified rule, by u_abα ^*Optimum control pulse is converted to be controlled rectifier.

Application example：

Fig. 2 gives single-phase phase-locked loop system schematic.Voltage on line side vector is counted by second order improper integral (SOGI) Quadrature voltage component is calculated, then by the off line amount of pressing to of orthogonal rest frame to calculating net pressure component by rotating coordinate transformation Q axle components under synchronous rotating frame (d-q), passing ratio integral controller controls it as zero, completes net pressure lock phase Function, while realizing asking for for net pressure virtual orthographic component.

Fig. 3 shows, instantaneous power computing block diagram.By voltage on line side quadrature component, and the modulation after rotating coordinate transformation The input quantity that ripple quadrature component is calculated as imaginary axis feedback instantaneous power, by instantaneous power algorithm for estimating, completes instantaneous work( Rate is estimated.

Claims (1)

1. a kind of Pulse rectifier low switching frequency model prediction power control algorithm, it is characterised in that comprising following step Suddenly：

Step 1：The instantaneous value of active power and reactive power in next dutycycle update cycle is estimated by formula (1)：

$\left\{\begin{array}{c}P(n\frac{T_s}{2}+t)=A_1\left(n\frac{T_s}{2}\right)cos\mathrm{\ω}t+A_2\left(n\frac{T_s}{2}\right)sin\mathrm{\ω}t-\frac{u_m}{2\mathrm{\ω}L}{u}_{abq}\left(n\frac{T_s}{2}\right)\\ Q(n\frac{T_s}{2}+t)=A_1\left(n\frac{T_s}{2}\right)sin\mathrm{\ω}t-A_2\left(n\frac{T_s}{2}\right)cos\mathrm{\ω}t+\frac{u_m^2}{2\mathrm{\ω}L}-\frac{u_m}{2\mathrm{\ω}L}{u}_{abd}\left(n\frac{T_s}{2}\right)\end{array}\right.,(0\≤t\≤\frac{T_s}{2})---\left(1\right)$

Wherein, T_sIt is switch periods, n represents half switch periods (T_s/ number 2), ω is line voltage angular frequency, u_abdAnd u_abq D axle component and q axle component of the rectifier input voltage under d-q synchronous rotating frames are represented respectively；u_mRepresent voltage on line side Amplitude, L is net side inductance value；A₁And A₂It is initial parameter, by n-th half switches period start time active-power P and idle work( The instantaneous value of rate Q determines that it is expressed as

$\left\{\begin{array}{c}{A}_1\left(n\frac{T_s}{2}\right)=\frac{u_m{u}_{abq}\left(n\frac{T_s}{2}\right)}{2\mathrm{\ω}L}+P\left({\mathrm{kT}}_c\right)\\ A_2\left(n\frac{T_s}{2}\right)=\frac{u_m^2}{2\mathrm{\ω}L}-\frac{u_m{u}_{abd}\left(n\frac{T_s}{2}\right)}{2\mathrm{\ω}L}-Q\left({\mathrm{kT}}_c\right)\end{array}\right.---\left(2\right)$

Wherein, T_cControlling cycle is represented, k is represented closest to n-th controlling cycle numbers of half switch periods；

Step 2：Defining evaluation function J is：

J={ P_ref[(n+1)T_s/2]-P[(n+1)T_s/2]}²+{Q_ref[(n+1)T_s/2]-Q[(n+1)T_s/2]}² (3)

Wherein, P_ref[(n+1)T_s/ 2] and Q_ref[(n+1)T_s/ 2] be respectively reference in (n+1)th half switch periods it is active and The set-point of reactive power；P[(n+1)T_s/ 2] and Q [(n+1) T_s/ 2] be active power in (n+1)th half switch periods and The predicted value of reactive power, these power prediction values are solved by formula (1) and obtained；

Step 3：To cause evaluation function J minimum, then u_abdAnd u_baqMeet formula (4)：

$\left\{\begin{array}{c}\frac{\∂J}{\∂u_{abd}}=0\\ \frac{\∂J}{\∂u_{abq}}=0\end{array}\right.---\left(4\right)$

Formula (1) and formula (3) are substituted into formula (4), obtains causing the optimum control amount u that evaluation function is minimum_abαUnder rotating coordinate system Component u_abdWith u_abqCalculating formula be：

$\left\{\begin{array}{c}{u}_{abd}\left(n\frac{T_s}{2}\right)=u_m-\frac{\mathrm{\ω}L}{u_m}\[Q_{ref}\left({\mathrm{kT}}_c\right)+Q\left({\mathrm{kT}}_c\right)\]-\frac{\mathrm{\ω}L\sin \left({\mathrm{\ωT}}_s/2\right)}{u_m\[1-\cos \left({\mathrm{\ωT}}_s/2\right)\]}\[P_{ref}\left({\mathrm{kT}}_c\right)-P\left({\mathrm{kT}}_c\right)\]\\ u_{abq}\left(n\frac{T_s}{2}\right)=-\frac{\mathrm{\ω}L}{u_m}\[P_{ref}\left({\mathrm{kT}}_c\right)+P\left({\mathrm{kT}}_c\right)\]+\frac{\mathrm{\ω}L\sin \left({\mathrm{\ωT}}_s/2\right)}{u_m\[1-\cos \left({\mathrm{\ωT}}_s/2\right)\]}\[Q_{ref}\left({\mathrm{kT}}_c\right)-Q\left({\mathrm{kT}}_c\right)\]\end{array}\right.---\left(5\right)$

Wherein：u_abd(nT_s/ 2) and u_abq(nT_s/ 2) represent lower half switch periods internal modulation ripple at synchronous rotating frame (d-q) Lower used optimum control component；T_cControlling cycle is represented, it is close to n-th half switch periods；

Step 4：By rotating coordinate transformation, by optimal controlled quentity controlled variable u_abdAnd u_abqα axles under convert to static coordinate system (alpha-beta) Component, realizes that the voltage-second balance under low switching frequency is compensated by such as formula (6)：

${u_{ab\mathrm{\α}}}^{*}=\frac{2u_{abd}}{{\mathrm{\ωT}}_s}\[cos\frac{{\mathrm{n\ωT}}_s}{2}-cos\frac{(n+1){\mathrm{\ωT}}_s}{2}\]+\frac{2u_{abq}}{{\mathrm{\ωT}}_s}\[sin\frac{(n+1){\mathrm{\ωT}}_s}{2}-sin\frac{{\mathrm{n\ωT}}_s}{2}\]---\left(6\right)$

Wherein, u_abα ^*It is by the modulating wave after voltage-second balance compensation；

Step 5：Modulating wave is modulated, by u_abα ^*Optimum control pulse is converted to be controlled rectifier.