Improved sliding mode active disturbance rejection control method for two-stage energy storage converter

Abstract

The present invention discloses an improved sliding mode anti-disturbance control method for a two-stage energy storage converter, specifically, step 1, PCS controller design: step 1.1, calculating the converter AC side differential equation and AC side instantaneous power in the dq coordinate system, and thereby calculating the relationship between d, q axis current and power; step 1.2, finite time observer design; step 1.3, deriving the PCS comprehensive control law according to the equivalent control law and the switching control law; step 2, bidirectional DC / DC control design. The present invention solves the problem of poor transient performance of a two-stage energy storage converter when power fluctuates and DC bus voltage fluctuates.

Description

Technical Field

[0001] The invention belongs to the technical field of power electronic converters, and in particular relates to an improved sliding mode anti-disturbance control method for a two-stage energy storage converter. Background Art

[0002] In the energy storage system, the power conversion system (PCS) is an indispensable part. Commonly used energy storage converters are unipolar and two-stage. Among them, the unipolar PCS has a simple structure and the control strategy is relatively easy to implement. However, it requires a large number of energy storage batteries, is difficult to manage batteries, and lacks the function of raising and lowering the battery pack voltage. Adding a bidirectional Buck / Boost converter to the front DC side of the unipolar PCS to form a two-stage PCS can improve the voltage level of the energy storage system application. In addition, the two-stage PCS controls the charging and discharging of the energy storage battery by adding a bidirectional DC / DC converter, which can also ensure the stable operation of the DC bus voltage and improve the system reliability. The stable operation of the energy storage system is the premise for new energy power generation to play an important role in peak shaving and valley filling, source-load balance, etc. In the two-stage PCS, the bidirectional DC / DC converter and PCS are the core parts of the entire system. Therefore, the study of the control strategy of this part is the key to ensure the stable operation of the energy storage system. Among the existing PCS control strategies, most are traditional linear proportional integral (PI) control, but PI control has certain shortcomings, such as large overshoot, long transient time, and poor anti-disturbance ability.

[0003] Among the existing nonlinear controls, there are model predictive control, fuzzy control, active disturbance rejection control, sliding mode control, etc. Among them, although model predictive control and fuzzy control can handle nonlinear systems well and have good robustness, they have large computational complexity or require many data training times, are highly dependent on the model and also rely on the subjective experience of the designer. Active disturbance rejection control (ADRC) is based on the extended state observer, does not rely on the precise modeling of the system, and has good application prospects in nonlinear systems. Sliding mode control has rapid convergence and strong robustness and has attracted widespread attention. Summary of the invention

[0004] The purpose of the present invention is to provide an improved sliding mode anti-disturbance control method for a two-stage energy storage converter, which solves the problem of poor transient performance of the two-stage energy storage converter when power fluctuates and DC bus voltage fluctuates.

[0005] The technical solution adopted by the present invention is an improved sliding mode anti-disturbance control method for a two-stage energy storage converter, which is specifically implemented according to the following steps:

[0006] Step 1: PCS controller design

[0007] Step 1.1, calculate the differential equation of the converter AC side and the instantaneous power of the AC side in the dq coordinate system, and calculate the relationship between the d-axis and q-axis currents and the power;

[0008] Step 1.2, finite time observer design;

[0009] Step 1.3, derive the PCS comprehensive control law based on the equivalent control law and the switching control law;

[0010] Step 2: Bidirectional DC / DC control design

[0011] The DC / DC converter adopts dual-loop control of voltage outer loop and current inner loop. The design process of current inner loop controller is the same as that of voltage outer loop controller. Voltage outer loop controller adopts improved sliding mode anti-disturbance control. The uncertain part of the system is estimated through finite time observer, and its estimated value is used to compensate for disturbance. Improved integral sliding mode is used in feedback control.

[0012] The present invention is also characterized in that:

[0013] Step 1.1 is as follows:

[0014] The energy storage system connects the input voltage to the two-stage PCS converter through a DC / DC bidirectional converter, and supplies the AC load by controlling the power output of the converter. The variable relationship of PCS in the abc coordinate system is:

[0015]

[0016] In formula (1), u a ,u b ,u c is the three-phase voltage on the grid side, i a ,i b ,i c is the three-phase current on the grid side, u ga ,u gb ,u gc is the AC grid voltage, L, C are the filter parameters, R is the parasitic parameter of the filter inductor, i dc is the stabilizing capacitor current, C dc is the voltage stabilizing capacitor, u dc is the DC bus voltage, i r is the battery current, s j is the switching function that characterizes the on-off state of the switch tubes in each bridge arm of the converter, j = a, b, c,

[0017]

[0018] After coordinate transformation of equation (1), the differential equation of the converter AC side in the dq coordinate system is obtained as follows:

[0019]

[0020] In formula (3), ω is the grid voltage angular frequency, u dr =s d u dc ,u qr =s q u dc ,s d 、s q are the d and q axis switching functions, u d ,u q are the d-axis and q-axis input voltages respectively, i d ,i q are the d and q axis input currents respectively;

[0021] On the basis of the three-phase balance of the grid voltage, the d-axis direction of the grid voltage is taken as the direction of the voltage vector, that is, u d =u g ,u q =0, according to the instantaneous power theory, the instantaneous power on the AC side is:

[0022]

[0023] In formula (4), P is the active power output on the AC side, and Q is the reactive power output on the AC side;

[0024] The relationship between d-axis current and q-axis power is calculated by equation (4):

[0025]

[0026] Step 1.2 is as follows:

[0027] Converting equation (3) into the ADRC paradigm is:

[0028]

[0029] In formula (6), ω is the grid voltage angular frequency, b d 、b q are the d-axis and q-axis control gain, respectively, and f d 、f q are the equivalent d-axis and q-axis lumped disturbances respectively. The lumped disturbance includes the unmodeled part of the system, the coupled part, and the internal and external disturbances of the system. The lumped disturbance expression is:

[0030]

[0031] Let the state variable x d1 =i d , x d2 =f d , xq1 =i q , x q2 =f q , then formula (6) can be written as follows:

[0032]

[0033] According to formula (8), the form of the finite time observer is:

[0034]

[0035] In formula (9), l 1,2,3,4 >0 is the observer gain, 0.5<α1=α3<1, α2=α4=2α, z d1 、z d2 、z q1 、z q2 x d1 、x d2 、x q1 、x q2 Observed value of .

[0036] Step 1.3 is as follows:

[0037] Let i after coordinate transformation d ,i q The deviations e1 and e2 from the reference values ​​are:

[0038]

[0039] In formula (10), i dref for i d The reference value of i qref for i q Reference value of

[0040] Construct the d-axis and q-axis integral sliding mode functions s1 and s2 as:

[0041]

[0042] In formula (11), λ1, λ2, β1, β2, a1, a2 are sliding surface coefficients, and are all greater than or equal to 0;

[0043] make Substituting it into equation (6), we can get the equivalent control law u for the d and q axes: dr,eq 、u qr,eq :

[0044]

[0045] The improved compound speed reaching law is used as the switching control law of sliding mode anti-disturbance control, and the d- and q-axis switching control law u d,sw、u q,sw as follows:

[0046]

[0047] In formula (13), δ is a variable exponential parameter;

[0048] δ=γ+(ε-γ)e -μ|s| (14)

[0049] In formula (14), γ>1, 0<ε<1, μ>0, s is the sliding mode function, and s takes s1 and s2 according to the corresponding relationship between δ and the d and q axes respectively;

[0050] Therefore, the PCS comprehensive control law u on the d and q axes is obtained: dr 、u qr for:

[0051]

[0052] The design of the voltage outer loop controller in step 2 is as follows:

[0053] The power switch tubes of the DC / DC converter are VT1 and VT2. d and d′ are the duty ratios of VT1 and VT2 respectively. Since VT1 and VT2 are complementary to each other, i.e., d+d′=1, it is only necessary to control the on / off state of one of the switch tubes and the other switch tube to obtain the state at the same time. When the switch tube VT1 is turned on and VT2 is turned off, the DC / DC converter works in Buck mode; when the switch tube VT1 is turned off and VT2 is turned on, the converter works in Boost mode. According to the state space averaging method, the state equation of the DC / DC converter is:

[0054]

[0055] In formula (16), u in is the discharge voltage of the energy storage battery, and u in As the input voltage of the DC converter, L dc , C dc are filter inductance and capacitance respectively, f v (t) is the unknown disturbance of the voltage outer loop, f c (t) is the unknown disturbance of the current inner loop, i0 is the DC side input current of the energy storage converter;

[0056] The voltage equation in equation (16) is converted into ADRC form:

[0057]

[0058] In formula (17), b1 is the voltage loop control gain, f1 is the equivalent lumped disturbance outside the control input;

[0059] Let the state variable x of the voltage outer loop be v1 =u dc ,x v2 =f1, then the voltage outer loop observer is in the form of:

[0060]

[0061] In formula (18), the voltage outer loop observer gain l v1,v2 >0,z v1 、z v2 x v1 、x v2 Observed value of

[0062] The reference value of the DC / DC converter output voltage is u dcref , the voltage deviation between the output voltage and the reference value e v for:

[0063] e v =u dc -u dcref (19)

[0064] The differential of voltage deviation with respect to time is:

[0065]

[0066] In formula (20), b1 is the voltage loop control gain, and u1 represents the voltage outer loop control law;

[0067] Design the voltage outer loop integral sliding mode function s v for:

[0068] s v =e v +λ3∫e v dt+β3 tanh(a3e v ) (twenty one)

[0069] In formula (21), λ3, β3, and a3 are all coefficients of the voltage outer loop sliding mode surface, and are all greater than 0;

[0070] The reaching law of the voltage outer loop sliding mode feedback is:

[0071]

[0072] In formula (22), k7, k8, and k9 are all reaching law coefficients, and are all greater than 0;

[0073] Combining equations (17) to (22), the voltage outer loop control law is:

[0074]

[0075] The beneficial effects of the present invention are:

[0076] The invention discloses an improved sliding mode anti-disturbance control method for a two-stage energy storage converter, applies ADRC to the constant power control of a PCS and the dual-loop control of a front-stage DC / DC converter, takes ADRC as the main body of the controller, introduces a finite time observer and sliding mode control into the ADRC to improve the error convergence speed and control accuracy, thereby improving the anti-disturbance capability of the entire controller; solves the problems of poor transient performance and strong coupling of the two-stage energy storage converter when the power fluctuates and the DC bus voltage fluctuates; the control method of the invention has good coupling suppression effect, reduces the grid-connected current harmonics, and improves the transient response capability of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 It is the main circuit diagram of the two-stage energy storage converter;

[0078] Figure 2 It is a PCS control block diagram of the improved sliding mode anti-disturbance control method of the two-stage energy storage converter of the present invention;

[0079] Figure 3 It is a DC / DC converter voltage outer loop control block diagram of the improved sliding mode anti-disturbance control method of the two-stage energy storage converter of the present invention;

[0080] Figure 4 It is a comparison chart of PCS active power response simulation under different control strategies;

[0081] Figure 5 This is a comparison chart of PCS reactive power response simulation under different control strategies;

[0082] Figure 6 This is the current harmonic analysis diagram under PI control;

[0083] Figure 7 It is a current harmonic analysis diagram when the control method of the present invention is controlled;

[0084] Figure 8 is the DC bus voltage response diagram under different control strategies;

[0085] FIG9 is a transient response diagram of active power of PCS under different control strategies, wherein FIG9(a) is a transient response diagram of PI control strategy, and FIG9(b) is a transient response diagram under control of the present invention;

[0086] FIG10 is a transient response diagram of the active power direction of the PCS under different control strategies when switching, wherein FIG10(a) is a transient response diagram of the PI control strategy, and FIG10(b) is a transient response diagram under the control of the present invention;

[0087] FIG11 is a transient response diagram of PCS reactive power under different control strategies, wherein FIG11(a) is a transient response diagram of PI control strategy, and FIG11(b) is a transient response diagram under control of the present invention;

[0088] FIG12 is a transient response diagram of the DC bus voltage under different control strategies when the PCS output power suddenly changes, wherein FIG12(a) is a transient response diagram of the PI control strategy, and FIG12(b) is a transient response diagram under the control of the present invention;

[0089] FIG13 is a transient response diagram of the DC bus voltage under different control strategies when the PCS output power direction is switched, wherein FIG13(a) is a transient response diagram of the PI control strategy, and FIG13(b) is a transient response diagram under the control of the present invention. DETAILED DESCRIPTION

[0090] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0091] Example 1

[0092] This embodiment provides an improved sliding mode auto-disturbance rejection control method for a two-stage energy storage converter, which is specifically implemented according to the following steps:

[0093] Step 1: PCS controller design

[0094] Step 1.1, calculate the differential equation of the converter AC side and the instantaneous power of the AC side in the dq coordinate system, and calculate the relationship between the d-axis and q-axis currents and the power;

[0095] Step 1.2, finite time observer design;

[0096] Step 1.3, derive the PCS comprehensive control law based on the equivalent control law and the switching control law;

[0097] Step 2: Bidirectional DC / DC control design

[0098] The DC / DC converter adopts dual-loop control of voltage outer loop and current inner loop. The design process of current inner loop controller is the same as that of voltage outer loop controller. Voltage outer loop controller adopts improved sliding mode anti-disturbance control. The uncertain part of the system is estimated through finite time observer, and its estimated value is used to compensate for disturbance. Improved integral sliding mode is used in feedback control.

[0099] Example 2

[0100] This embodiment provides an improved sliding mode anti-disturbance control method for a two-stage energy storage converter. Based on Embodiment 1, step 1 is specifically as follows:

[0101] Step 1.1, calculate the differential equation of the converter AC side and the instantaneous power of the AC side in the dq coordinate system, and calculate the relationship between the d-axis and q-axis currents and the power;

[0102] The circuit structure of the two-stage PCS is as follows: Figure 1 As shown in the figure, the energy storage system connects the input voltage to the two-stage PCS converter through the DC / DC bidirectional converter, and supplies the AC load by controlling the converter to output a certain power. When the power output of the converter exceeds the load requirement, the excess can be connected to the grid for use. Conversely, when the converter output power is insufficient to meet the load requirement, the power difference is provided by the grid. In addition, when the load is low, the AC meter can charge the energy storage system through the converter to achieve efficient use of electrical energy. Figure 1 It can be seen that the variable relationship of PCS in the abc coordinate system is:

[0103]

[0104] In formula (1), u a ,u b ,u c is the three-phase voltage on the grid side, i a ,i b ,i c is the three-phase current on the grid side, u ga ,u gb ,u gc is the AC grid voltage, L, C are the filter parameters, R is the parasitic parameter of the filter inductor, i dc is the stabilizing capacitor current, C dc is the voltage stabilizing capacitor, u dc is the DC bus voltage, i r is the battery current, s j is the switching function that characterizes the on-off state of the switch tubes in each bridge arm of the converter, j = a, b, c,

[0105]

[0106] After coordinate transformation of equation (1), the differential equation of the converter AC side in the dq coordinate system is obtained as follows:

[0107]

[0108] In formula (3), ω is the grid voltage angular frequency, u dr =s d u dc ,u qr =s q u dc ,s d 、s q are the d and q axis switching functions, u d ,uq are the d-axis and q-axis input voltages respectively, i d ,i q are the d and q axis input currents respectively;

[0109] On the basis of the three-phase balance of the grid voltage, the d-axis direction of the grid voltage is taken as the direction of the voltage vector, that is, u d =u g ,u q =0, according to the instantaneous power theory, the instantaneous power on the AC side is:

[0110]

[0111] In formula (4), P is the active power output on the AC side, and Q is the reactive power output on the AC side;

[0112] The relationship between d-axis current and q-axis power is calculated by equation (4):

[0113]

[0114] Step 1.2, finite time observer design;

[0115] Converting equation (3) into the ADRC paradigm is:

[0116]

[0117] In formula (6), ω is the grid voltage angular frequency, b d 、b q are the d-axis and q-axis control gain, respectively, and f d 、f q are the equivalent d-axis and q-axis lumped disturbances respectively. The lumped disturbance includes the unmodeled part of the system, the coupled part, and the internal and external disturbances of the system. The lumped disturbance expression is:

[0118]

[0119] Let the state variable x d1 =i d , x d2 =f d , x q1 =i q , x q2 =f q , then formula (6) can be written as follows:

[0120]

[0121] According to formula (8), the form of the finite time observer is:

[0122]

[0123] In formula (9), l 1,2,3,4 >0 is the observer gain, 0.5<α1=α3<1, α2=α4=2α, z d1 、z d2 、z q1 、z q2 x d1 、x d2 、x q1 、x q2 Observed value of .

[0124] Step 1.3, derive the PCS comprehensive control law based on the equivalent control law and the switching control law;

[0125] Let i after coordinate transformation d ,i q The deviations e1 and e2 from the reference values ​​are:

[0126]

[0127] In formula (10), i dref for i d The reference value of i qref for i q Reference value of

[0128] Construct the d-axis and q-axis integral sliding mode functions s1 and s2 as:

[0129]

[0130] In formula (11), λ1, λ2, β1, β2, a1, a2 are sliding surface coefficients, and are all greater than or equal to 0;

[0131] The sliding mode function can track the expected value quickly and smoothly, and also has good control accuracy and control stability under disturbance;

[0132] make Substituting it into equation (6), we can get the equivalent control law u for the d and q axes: dr,eq 、u qr,eq :

[0133]

[0134] The equivalent control law can ensure that the control trajectory of the system moves on the sliding surface. In addition, the control law needs to be switched to ensure that the movement of the control trajectory of the system does not leave the sliding surface.

[0135] The switching control law is generally performed by the sliding mode approaching law, and the commonly used ones are the constant speed approaching law and the exponential approaching law. However, the constant speed approaching law has a slow approaching speed, and although the exponential approaching law has a fast approaching speed, it will produce a large chattering when the system motion trajectory reaches the vicinity of the sliding mode surface. Therefore, the improved compound variable speed approaching law is used as the switching control law of the sliding mode anti-disturbance, and the d and q axis switching control law u d,sw 、u q,sw as follows:

[0136]

[0137] In formula (13), δ is a variable exponential parameter;

[0138] δ=γ+(ε-γ)e -μ|s| (14)

[0139] In formula (14), γ>1, 0<ε<1, μ>0, s is the sliding mode function, s takes s1 and s2 according to the corresponding relationship between δ and d and q axes, and γ=1.5, ε=0.5, μ=1;

[0140] The improved reaching law is a composite of the exponential reaching law and the variable power reaching law. The variable exponential parameter δ can be adaptively changed according to the control trajectory of the system. When the motion trajectory of the controlled object is far away from the sliding surface, the value of δ increases to speed up the speed of the controlled trajectory approaching the sliding surface; when the distance between the motion trajectory and the sliding surface is small, δ takes a smaller value to weaken the chattering phenomenon in the approaching motion.

[0141] Therefore, the PCS comprehensive control law u on the d and q axes is obtained: dr 、u qr for:

[0142]

[0143] like Figure 2 Shown is the PCS control block diagram.

[0144] Example 3

[0145] This embodiment provides an improved sliding mode anti-disturbance control method for a two-stage energy storage converter. Based on Embodiment 1-2, the voltage outer loop controller in step 2 is specifically designed as follows:

[0146] Depend on Figure 1It can be seen that the power switch tubes of the DC / DC converter are VT1 and VT2, and d and d′ are the duty ratios of VT1 and VT2 respectively. Since VT1 and VT2 are complementary to each other, that is, d+d′=1, it is only necessary to control the on-off state of one of the switch tubes and the other switch tube to obtain the state at the same time. When the switch tube VT1 is turned on and VT2 is turned off, the DC / DC converter works in Buck mode; when the switch tube VT1 is turned off and VT2 is turned on, the converter works in Boost mode. According to the state space averaging method, the state equation of the DC / DC converter is:

[0147]

[0148] In formula (16), u in is the discharge voltage of the energy storage battery, and u in As the input voltage of the DC converter, L dc is the filter inductance, f v (t) is the unknown disturbance of the voltage outer loop, f c (t) is the unknown disturbance of the current inner loop, i0 is the DC side input current of the energy storage converter;

[0149] The DC / DC converter uses voltage and current dual-loop control to improve the control accuracy and disturbance suppression capability of the system. The voltage outer loop also uses the improved sliding mode anti-disturbance control. The uncertain part of the system is estimated by the finite time observer, and its estimated value is used to compensate for the disturbance. In the feedback control, the improved integral sliding mode is used to improve the response speed and anti-disturbance capability of the controller. The control purpose of the voltage outer loop is to provide a reference value for the inner loop inductor current, so the control quantity of the voltage outer loop is the inductor current;

[0150] The voltage equation in equation (16) is converted into ADRC form:

[0151]

[0152] In formula (17), b1 is the voltage loop control gain, f1 is the equivalent lumped disturbance outside the control input;

[0153] Let the state variable x of the voltage outer loop be v1 =u dc ,x v2 =f1, then the voltage outer loop observer is in the form of:

[0154]

[0155] In formula (18), the voltage outer loop observer gain l v1,v2 >0,z v1 、z v2 x v1 、x v2 Observed value of

[0156] The reference value of the DC / DC converter output voltage is u dcref , the voltage deviation between the output voltage and the reference value e v for:

[0157] e v =u dc -u dcref (19)

[0158] The differential of voltage deviation with respect to time is:

[0159]

[0160] In formula (20), b1 is the voltage loop control gain, and u1 represents the voltage outer loop control law;

[0161] Design the voltage outer loop integral sliding mode function s v for:

[0162] s v =e v +λ3∫e v dt+β3tanh(a3e v ) (twenty one)

[0163] In formula (21), λ3, β3, and a3 are all coefficients of the voltage outer loop sliding mode surface, and are all greater than 0;

[0164] The reaching law of the voltage outer loop sliding mode feedback is:

[0165]

[0166] In formula (22), k7, k8, and k9 are all reaching law coefficients, and are all greater than 0;

[0167] Combining equations (17) to (22), the voltage outer loop control law is:

[0168]

[0169] The design process of the current inner loop controller is the same as that of the voltage outer loop. The current inner loop is based on the inductor current i in the DC / DC converter. L is the state variable, according to i L The i obtained by equation (23) in the above voltage outer loop Lref The controller is designed based on the current deviation between them to obtain the control signal (i.e. duty cycle d) that is finally passed into the DC / DC converter, thereby ensuring the stable operation of the DC / DC module.

[0170] Simulation and Experimental Analysis

[0171] (I) Analysis of simulation results

[0172] In order to verify the feasibility of the control method of the present invention, a simulation circuit was built in MATLAB / Simulink software, and a DC / DC dual-loop PI controller and a PI decoupling controller of the converter were built for comparison, and then the advantages of the control method of the present invention were analyzed.

[0173] Table 1 System model parameters

[0174]

[0175] Table 2 Controller parameters

[0176]

[0177] In order to verify the transient advantage of the control method of the present invention compared with the traditional control strategy, the corresponding active power transient mutation was made in the simulation, and the results are as follows: Figure 4 .

[0178] In the simulation, the active power of the PCS part was set to change three times: the active power changed from 20kW to 40kW in 0.15s, from 40kW to 10kW in 0.20s, and from 10kW to -20kW in 0.25s ("+" indicates PCS output power, "-" indicates PCS absorption power). Figure 4 It can be seen that under PI control, a certain amount of overshoot is generated after the power jump, and it takes a certain period of transient process to recover to steady-state operation; and the control method of the present invention has a smaller overshoot than PI control, and a shorter transient time to recover to steady state. Figure 4 The part in the middle frame is the coupling effect caused by the jump of reactive power at 0.3s. The control method of the present invention has obvious coupling suppression effect.

[0179] In the simulation, the reactive power is set to jump once. The reactive power remains at zero from the start time to 0.30s, and jumps to 30kVar at 0.30s. Figure 5 As shown, when reactive power changes suddenly, the control method of the present invention has smaller power fluctuations due to coupling effects, in addition to the fact that the power fluctuations during the jump and the power ripples during the steady state are smaller than those of the PI control.

[0180] To determine whether the power quality of the energy storage converter meets the grid connection requirements during grid-connected operation, a Fourier analysis of the grid-connected current at the PCS output active power of 20kW was performed in Simulink for two steady-state cycles. Figure 6 , Figure 7As shown, the harmonic distortion rates under PI control and the control method of the present invention are 4.52% and 1.63% respectively. Although both meet the grid-connected requirement of less than 5%, the control method of the present invention has less harmonic content, so the control method of the present invention can provide higher power quality.

[0181] PCS is directly connected to DC / DC output, so the PCS system can be compared to the load part of DC / DC. When the power output of PCS changes, the corresponding DC / DC load changes. Maintaining the stability of DC output voltage is the premise of stable operation of PCS. Figure 4 It can be seen that the active power output of the PCS has three jumps, and two representative jumps are selected for comparative analysis of the control effect of the DC output voltage.

[0182] The DC output voltage fluctuation caused by the jump from 20kW to 40kW at 0.15s and the jump from 10kW to -20kW at 0.25s is as follows: Figure 5 As shown in the figure, the voltage fluctuation amplitude under the PI control strategy reaches 27V and 45V respectively, and the transient time to restore to the given voltage value is 0.012s and 0.015s respectively; while under the control method of the present invention, the voltage fluctuation amplitude is 5.5V, and the transient time is 0.003s and 0.004s respectively; in addition, the ripple voltage generated by the control method of the present invention is also significantly smaller than that of the PI control. Therefore, from the simulation results, it can be concluded that the control method of the present invention is superior in terms of rapidity and anti-disturbance.

[0183] (II) Analysis of experimental results

[0184] In order to verify the feasibility of the control method of the present invention when running close to the actual situation, a hardware-in-the-loop experimental platform was built based on HIL6020, and PI control and the control method of the present invention were applied thereto for comparative analysis.

[0185] Since it is inconvenient to measure the PCS output power during the experiment, it can be seen from formula (3) that the output power has a direct relationship with the d and q axis voltage and current after coordinate transformation, and the d axis voltage is constant, and its value is basically stable at 310V. Therefore, the transformation of the output power can be reflected by observing the d and q axis currents. d 、i q They are d and q axis circuits, u b 、i b are the voltage and current of phase b measured by the PCS AC. From formula (3), it can be obtained that when the PCS outputs 20kW of active power and 0 reactive power, i d 43A, i q is 0.

[0186] As shown in Figure 9(a), when the PCS output active power jumps from 20kW to 60kW, i d Jump from 43A to 129A. Under the PI control strategy, the transition time from the start of the jump to the recovery of the steady state is 28ms; the transition time from 60kW to 20kW is 32ms. Both power jumps produced a certain amount of overshoot. In addition, due to the coupling between the powers, the transient process of reactive power was also caused when the active power suddenly changed, as shown in the framed part in Figure 9(a). It can be seen from Figure 9(b) that under the control method of the present invention, the transition times of the two power jumps were 8ms and 9ms respectively, and the jump period was relatively smooth without overshoot. It is known from the above simulation that the coupling effect under the control of the control method of the present invention is significantly smaller than that of the PI control. At the same time, in the experimental results, the control method of the present invention has a good decoupling effect.

[0187] When the PCS output power jumps from 20kW to -20kW, that is, the PCS changes from output power to absorption power, i d Jump from 43A to -43A. As shown in Figure 10(a), the transition time of the two jumps under PI control is 28ms and 27ms respectively. As shown in Figure 10(b), the transition time of the power direction switching under the control method of the present invention is 11ms and 12ms respectively. Similarly, the control method of the present invention is also better than PI control in suppressing overshoot and coupling.

[0188] In Figure 11, the active power is maintained at 20kW. When the reactive power jumps from 0 to 30kVar, q Jump from 0 to -64A. Under the PI control strategy, it takes 21ms to recover from the jump from reactive power. At the same time, a certain overshoot is generated during the transient period and the active power fluctuates due to the coupling between the two. In the control method of the present invention, the transient transition time is only 9ms and no overshoot is generated.

[0189] The experiment sets the DC bus voltage to 800V in steady state, as shown in Figure 12(a). dc The accuracy is 100V / grid. Under the PI control strategy, when the PCS output power jumps from 20kW to 60kW, the DC bus voltage overshoots by 40V and recovers to steady-state operation after a transient process of 25ms. When the output power jumps from 60kW to 20kW, the overshoot caused by voltage fluctuation is 38V, and the transition time is 23ms. In Figure 12(b), u dc The accuracy is 50V / grid. The voltage overshoot caused by two power jumps in the control method of the present invention is 10V and 14V respectively, and the transition time of the two transient processes is 5ms and 7ms respectively.

[0190] When the PCS changes from output power to absorption power, the previous DC / DC module changes from a discharge state to a charge state. In Figure 13(a), under the PI control strategy, when the output power changes from 20kW to -20kW, the overshoot caused by the voltage fluctuation is 38V, and the transient time is 28ms; when the output power jumps from -20kW to 20kW, the voltage overshoot is 31V, and it returns to a stable state after a transition time of 27ms. As shown in Figure 13(b), when the direction of the output power changes, the voltage fluctuation of the control method of the present invention is smaller and the transient time is shorter.

[0191] Through the above simulation and experimental analysis, we can know that:

[0192] 1) The improved sliding mode anti-disturbance control method of the two-stage energy storage converter of the present invention uses the ADRC theory to classify the coupling part of the PCS into the lumped disturbance, and the lumped disturbance is quickly observed and compensated by a finite time observer, thereby achieving the suppression of the PCS coupling.

[0193] 2) When the PCS output power suddenly changes and the power direction switches, the sliding mode anti-disturbance control controller designed by the improved sliding mode anti-disturbance control method of the two-stage energy storage converter of the present invention reduces the overshoot, improves the response speed and has better transient performance. In addition, the control method of the present invention also reduces the current harmonic content and improves the power quality.

Claims

1. An improved sliding mode anti-disturbance control method for a two-stage energy storage converter, characterized in that: Follow the steps below to implement it: Step 1: PCS controller design Step 1.1, calculate the differential equation of the converter AC side and the instantaneous power of the AC side in the dq coordinate system, and calculate the relationship between the d-axis and q-axis currents and the power; The energy storage system connects the input voltage to the two-stage PCS converter through a DC / DC bidirectional converter, and supplies the AC load by controlling the power output of the converter. The variable relationship of PCS in the abc coordinate system is: (1) In formula (1), u a , u b , u c is the three-phase voltage on the grid side, i a , i b , i c is the three-phase current on the grid side, u ga , u gb , u gc is the AC grid voltage, L is the filtering parameter, is the parasitic parameter of the filter inductor, i dc is the stabilizing capacitor current, C dc is the voltage stabilizing capacitor, u dc is the DC bus voltage, i r is the battery current, s j To characterize the switching function of the on-off state of each bridge arm switch tube in the converter, j = a,b,c , (2) After coordinate transformation of equation (1), the differential equation of the converter AC side in the dq coordinate system is obtained as follows: (3) In formula (3), w is the grid voltage angular frequency, u dr = s d u dc , u qr = s q u dc , s d , s q are the d-axis and q-axis switching functions, u d , u q are the d-axis and q-axis input voltages respectively, i d , i q are the d and q axis input currents respectively; On the basis of the three-phase balance of the grid voltage, the d-axis direction of the grid voltage is taken as the direction of the voltage vector, that is, u d = u g , u q =0, according to the instantaneous power theory, the instantaneous power on the AC side is: (4) In formula (4), P is the active power output on the AC side, and Q is the reactive power output on the AC side; The relationship between d-axis current and q-axis power is calculated by equation (4): (5)； Step 1.2, finite time observer design; Converting equation (3) into the ADRC paradigm is: (6) In formula (6), ω is the grid voltage angular frequency, b d , b q are the d-axis and q-axis control quantity gains respectively, f d , f q are the equivalent d-axis and q-axis lumped disturbances, respectively. The lumped disturbance includes the unmodeled part of the system, the coupled part, and the internal and external disturbances of the system. The lumped disturbance expression is: (7) Let the state variable x d1 = i d , x d2 = f d , x q1 =i q , x q2 = f q , then formula (6) can be written as follows: (8) According to formula (8), the form of the finite time observer is: (9) In formula (9), l 1,2,3,4 >0 is the observer gain, 0.5<α1=α3<1, α2=α4=2α, z d1 , z d2 , z q1 , z q2 They are x d1 , x d2 , x q1 , x q2 Observed value of Step 1.3, derive the PCS comprehensive control law based on the equivalent control law and the switching control law; After coordinate transformation i d , i q The deviations e1 and e2 from the reference values ​​are: (10) In formula (10), i dref for i d The reference value of i qref for i q Reference value of Construct d and q axis integral sliding mode functions s 1. s 2 is: (11) In formula (11), , , , , , All are sliding surface coefficients, and are greater than or equal to 0; make , and substitute it into equation (6) to obtain the equivalent control law of d and q axes: u dr,eq , u qr,eq : (12) The improved compound speed reaching law is used as the switching control law of sliding mode anti-disturbance control, and the d and q axis switching control law u d,sw , u q,sw as follows: (13) In formula (13), δ is a variable index parameter; (14) In formula (14), γ>1, 0< ε <1, μ >0, s is the sliding mode function, s according to δ The corresponding relationship with the d and q axes is s 1. s 2; Therefore, the comprehensive control law of PCS in d and q axes is obtained: u dr , u qr for: (15)； Step 2: Bidirectional DC / DC control design The DC / DC converter adopts dual-loop control of voltage outer loop and current inner loop. The design process of current inner loop controller is the same as that of voltage outer loop controller. Voltage outer loop controller adopts improved sliding mode anti-disturbance control. The uncertain part of the system is estimated through finite time observer, and its estimated value is used to compensate for disturbance. Improved integral sliding mode is used in feedback control.

2. The improved sliding mode auto-disturbance rejection control method for a two-stage energy storage converter according to claim 1 is characterized in that: The design of the voltage outer loop controller in step 2 is as follows: The power switch tubes of the DC / DC converter are VT1, VT2, d , d′ are the duty cycles of VT1 and VT2 respectively. Since VT1 and VT2 are complementary, d + d′ =1, so we only need to control the on / off state of one switch tube and the other switch tube to obtain the state equation of the DC / DC converter at the same time. When the switch tube VT1 is turned on and VT2 is turned off, the DC / DC converter works in Buck mode; when the switch tube VT1 is turned off and VT2 is turned on, the converter works in Boost mode. According to the state space averaging method, the state equation of the DC / DC converter is: (16) In formula (16), u in is the discharge voltage of the energy storage battery, and u in As the input voltage of the DC converter, L dc is the filter inductor, f v ( t ) is the unknown disturbance of the voltage outer loop, f c ( t ) is the unknown disturbance in the current inner loop, i 0 is the DC side input current of the energy storage converter; The voltage equation in equation (16) is converted into ADRC form: (17) In formula (17), b 1 is the voltage loop control gain, f 1 is the equivalent lumped disturbance outside the control input; Let the state variable of the voltage outer loop be , then the voltage outer loop observer is in the form of: (18) In formula (18), the voltage outer loop observer gain is , z v1 , z v2 They are x v1 , x v2 Observed value of The DC / DC converter output voltage reference value is u dcref , the voltage deviation between the output voltage and the reference value e v for: (19) The differential of voltage deviation with respect to time is: (20) In formula (20), b 1 is the voltage outer loop control gain, u 1 indicates the voltage outer loop control law; Design the voltage outer loop integral sliding mode function S v for: (21) In formula (21), , , They are all the voltage outer loop sliding surface coefficients, and are all greater than 0; The reaching law of the voltage outer loop sliding mode feedback is: (22) In formula (22), k 7. k 8. k 9 are all reaching law coefficients, and are all greater than 0; Combining equations (17) to (22), the voltage outer loop control law is: (23)。

Images