Decentralized voltage control method for microgrid based on nonlinear state observers

Abstract

A decentralized voltage control method for a microgrid based on nonlinear state observers, comprising the steps of step 10), establishing a large-signal model of distributed generations, a connection network and impedance-type loads in the microgrid; step 20), establishing a Luenberger-like nonlinear state observer for each distributed generation; step 30), estimating the dynamic characteristics of other distributed generations in real time based on the local measured values of each distributed generation; and step 40), implementing the decentralized voltage control based on the control requirements of reactive power sharing and voltage restoration. The control method realizes the voltage control of microgrid based on the decentralized state observers, which does not rely on communication transmission or remote measurement and avoids the adverse effects of communication latency and data drop-out on the control performance.

Claims

1. A method for controlling a decentralized voltage for a microgrid based on nonlinear state observers, characterized by comprising the following steps: step 10) scanning connection/disconnection statuses of all DG (Distributed Generation) units, loading and querying data; establishing a large-signal microgrid model with distributed generations, a connection network and impedance-type loads; step 20), establishing a nonlinear state observer for each of the distributed generations in a DSP (Digital Signal Processor): step 30), collecting output voltage V.sub.odi and output current i.sub.odi and i.sub.oqi from a local sensor, sending the output voltage V.sub.odi and output current i.sub.odi and i.sub.oqi to the nonlinear state observer; step 40), entering a microgrid voltage reference instruction by a human-machine interface, and sending the microgrid voltage reference instruction to the distributed generations via a 485 communication module that is a type of communication module; performing the decentralized voltage control by the distributed generations; wherein: the step 10) further comprising: there are N distributed generations; each distributed generation employs a droop control to generate the output voltage and frequency reference for individual inverters, as shown in formula (1): $\begin{array}{cc}\{\begin{array}{c}{}_i={}_n-m_{\mathrm{Pi}}{P}_i\\ k_{\mathrm{Vi}}{\stackrel{.}{V}}_{o,\mathrm{magi}}=V_n-V_{o,\mathrm{magi}}-n_{\mathrm{Qi}}{Q}_i\end{array}& \mathrm{formula}\left(1\right)\end{array}$ in formula (1), .sub.i denotes the local angular frequency of the i.sup.th distributed generation; .sub.n, denotes the local angular frequency reference value of the distributed generation: rad/s; m.sub.Pi denotes the frequency droop characteristic coefficient of the i.sup.th distributed generation: rad/s; P.sub.i denotes the actual output active power of the i.sup.th distributed generation: W; k.sub.Vi denotes a droop control gain of the i.sup.th distributed generation; {dot over (V)}.sub.o,magi denotes the change rate of the output voltage of the i.sup.th distributed generation: V/S; V.sub.n denotes the output voltage reference value of the distributed generation: V; V.sub.o,magi denotes the output voltage of the i.sup.th distributed generation: V; n.sub.Qi denotes the voltage droop characteristic coefficient of the i.sup.th distributed generation: V/Var; Q.sub.i denotes the actual output reactive power of the i.sup.th distributed generation: Var; the actual output active power P.sub.i and reactive power Q.sub.i of the i.sup.th distributed generation are obtained through a low-pass filter, as shown in formula (2): $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{P}}_i=-{}_{\mathrm{ci}}{P}_i+{}_{\mathrm{ci}}\left(V_{\mathrm{odi}}{i}_{\mathrm{odi}}+V_{\mathrm{oqi}}{i}_{\mathrm{oqi}}\right)\\ {\stackrel{.}{Q}}_i=-{}_{\mathrm{ci}}{Q}_i+{}_{\mathrm{ci}}\left(V_{\mathrm{oqi}}{i}_{\mathrm{odi}}-V_{\mathrm{odi}}{i}_{\mathrm{oqi}}\right)\end{array}& \mathrm{formula}\left(2\right)\end{array}$ in formula (2), {dot over (P)}.sub.i denotes the change rate of active power of the i.sup.th distributed generation: W/S; .sub.ci denotes the low-pass filter cutoff frequency of the i.sup.th distributed generation: rad/s; V.sub.odi denotes the d-axis component of the output voltage of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: V; V.sub.oqi denotes the q-axis component of the output voltage of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: V; i.sub.odi denotes the d-axis component of the output current of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: A; i.sub.oqi denotes the q-axis component of the output current of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: A; {dot over (Q)}.sub.i denotes the reactive power change rate of the i.sup.th distributed generation: Var/S; as a primary voltage control for each distributed generation enables the voltage magnitude on the q axis as zero, a secondary voltage control can be further obtained as formula (3): $\begin{array}{cc}\{\begin{array}{c}{k}_{\mathrm{Vi}}{\stackrel{.}{V}}_{\mathrm{odi}}=V_{\mathrm{ni}}-V_{\mathrm{odi}}-n_{\mathrm{Qi}}{Q}_i+u_i\\ V_{\mathrm{oqi}}=0\end{array}& \mathrm{formula}\left(3\right)\end{array}$ in formula (3), {dot over (V)}.sub.odi denotes the change rate of the d-axis component of the output voltage of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: V/S; V.sub.ni denotes the output voltage reference value of the i.sup.th distributed generation, and u.sub.i denotes the secondary voltage control quantity: V; a dynamic equation of the output current of the distributed generation is as shown in formula (4): $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{i}}_{\mathrm{odi}}=-\frac{R_{\mathrm{ci}}}{L_{\mathrm{ci}}}{i}_{\mathrm{odi}}+{}_i{i}_{\mathrm{oqi}}+\frac{1}{L_{\mathrm{ci}}}\left(V_{\mathrm{odi}}-V_{\mathrm{busdi}}\right)\\ {\stackrel{.}{i}}_{\mathrm{oqi}}=-\frac{R_{\mathrm{ci}}}{L_{\mathrm{ci}}}{i}_{\mathrm{oqi}}-{}_i{i}_{\mathrm{odi}}+\frac{1}{L_{\mathrm{ci}}}\left(V_{\mathrm{oqi}}-V_{\mathrm{busqi}}\right)\end{array}& \mathrm{formula}\left(4\right)\end{array}$ in formula (4), {dot over (i)}.sub.odi denotes the change rate of the d-axis component of the output current of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: A/S; R.sub.ci denotes the connection resistance from the i.sup.th distributed generation to the bus i: ; L.sub.ci denotes the connection inductance from the i.sup.th distributed generation to the bus i: H; V.sub.busdi denotes the d-axis component of the voltage of the bus i in the dq reference coordinate of the i.sup.th distributed generation; {dot over (i)}.sub.oqi denotes the change rate of the q-axis component of the output current of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: A/S; V.sub.busqi denotes the q-axis component of the voltage of the bus i in the dq reference coordinate of the i.sup.th distributed generation: V; according to formulas (1) to (4), a dynamic equation of the i.sup.th distributed generation is obtained, as shown in formula (5): $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{x}}_{\mathrm{invi}}=f_{\mathrm{invi}}\left(x_{\mathrm{invi}}\right)+k_{\mathrm{invi}}\left(x_{\mathrm{invi}}\right)V_{\mathrm{busDQi}}+h_{\mathrm{invi}}{}_{\mathrm{com}}+g_{\mathrm{invi}}{u}_i\\ i_{\mathrm{oDQi}}=C_{\mathrm{invci}}{x}_{\mathrm{invi}}\end{array}& \mathrm{formula}\left(5\right)\end{array}$ in the formula, {dot over (x)}.sub.invi denotes the change rate of the state variable of the i.sup.th distributed generation, {dot over (x)}.sub.invi=[{dot over ()}.sub.i, {dot over (P)}.sub.i, {dot over (Q)}.sub.i, {dot over (V)}.sub.odi, {dot over (i)}.sub.odi, {dot over (i)}.sub.oqi].sup.T; {dot over ()} denotes the change rate of .sub.i; x.sub.invi denotes the state variable of the i.sup.th distributed generation, x.sub.invi=[.sub.i, P.sub.i, Q.sub.i, V.sub.odi, i.sub.odi, i.sub.oqi].sup.T; wherein .sub.i denotes the phase angle difference between the dq axis of the dq reference coordinate of the i.sup.th distributed generation and the DQ axis in the common reference coordinate DQ of the microgrid: rad; f.sub.invi(x.sub.invi) denotes a state function of the i.sup.th distributed generation, and k.sub.invi(x.sub.invi) denotes a voltage disturbance function of the bus i; V.sub.busDQi=[V.sub.busDi, V.sub.busQi].sup.T, and V.sub.busDi denotes the D-axis component of the bus i in the common reference coordinate DQ: V; V.sub.busQi denotes the Q-axis component of the bus i in the common reference coordinate DQ: V; .sub.com denotes the angular frequency of the common reference coordinate: rad/s; h.sub.invi denotes a connection matrix of the angular frequency of the common reference coordinate; g.sub.invi denotes an input matrix of the i.sup.th distributed generation; i.sub.oDQi=[i.sub.oDi, i.sub.oQi].sup.T, i.sub.oDi denotes the D-axis component of the output current of the i.sup.th distributed generation in the common reference coordinate DQ, and i.sub.oQi denotes the Q-axis component of the output current of the i.sup.th distributed generation in the common reference coordinate DQ: A; C.sub.invci denotes an output matrix of the i.sup.th distributed generation; a current dynamic equation of the i.sup.th line between the bus i and the bus j is as shown in formula (6): $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{i}}_{\mathrm{lineDi}}=-\frac{r_{\mathrm{linei}}}{L_{\mathrm{linei}}}{i}_{\mathrm{lineDi}}+{}_{\mathrm{com}}{i}_{\mathrm{lineQi}}+\frac{1}{L_{\mathrm{linei}}}\left(V_{\mathrm{busDi}}-V_{\mathrm{busDj}}\right)\\ {\stackrel{.}{i}}_{\mathrm{lineQi}}=-\frac{r_{\mathrm{linei}}}{L_{\mathrm{linei}}}{i}_{\mathrm{lineQi}}-{}_{\mathrm{com}}{i}_{\mathrm{lineDi}}+\frac{1}{L_{\mathrm{linei}}}\left(V_{\mathrm{busQi}}-V_{\mathrm{busQj}}\right)\end{array}& \mathrm{formula}\left(6\right)\end{array}$ in the formula, i.sub.lineDi denotes the change rate of the D-axis component of the current of the i.sup.th line in the common reference coordinate DQ: A/S; r.sub.linei denotes the line resistance of the i.sup.th line: Q; L.sub.linei denotes the line inductance of the i.sup.th line: H; i.sub.lineDi denotes the D-axis component of the current of the i.sup.th line in the common reference coordinate DQ, and i.sub.lineQi denotes the Q-axis component of the current of the i.sup.th line in the common reference coordinate DQ: A; V.sub.busDi denotes the D-axis component of the bus i in the common reference coordinate DQ, V.sub.busDj denotes the D-axis component of the bus j in the common reference coordinate DQ, and i.sub.lineQi denotes the change rate of the Q-axis component of the current of the i.sup.th line in the common reference coordinate DQ: A/S; V.sub.busQi denotes the Q-axis component of the bus i in the common reference coordinate DQ: V; V.sub.busQj denotes the Q-axis component of the bus j in the common reference coordinate DQ: V; a current dynamic equation of the j.sup.th load connected to the bus j is as shown in formula (7): $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{i}}_{\mathrm{loadDj}}=-\frac{R_{\mathrm{loadj}}}{L_{\mathrm{loadj}}}{i}_{\mathrm{loadDj}}+{}_{\mathrm{com}}{i}_{\mathrm{loadQj}}+\frac{1}{L_{\mathrm{loadj}}}{V}_{\mathrm{busDj}}\\ {\stackrel{.}{i}}_{\mathrm{loadQj}}=-\frac{R_{\mathrm{loadj}}}{L_{\mathrm{loadj}}}{i}_{\mathrm{loadQj}}-{}_{\mathrm{com}}{i}_{\mathrm{loadDj}}+\frac{1}{L_{\mathrm{loadj}}}{V}_{\mathrm{busQj}}\end{array}& \mathrm{formula}\left(7\right)\end{array}$ in the formula, i.sub.loadDj denotes the change rate of the D-axis component of the current of the j.sup.th load in the common reference coordinate DQ: A/S; R.sub.loadj denotes the load resistance of the j.sup.th load: QS; L.sub.loadj denotes the load inductance of the j.sup.th load: H; i.sub.loadDj denotes the D-axis component of the current of the j.sup.th load in the common reference coordinate DQ, and i.sub.loadQj denotes the Q-axis component of the current of the j.sup.th load in the common reference coordinate DQ: A; and i.sub.loadQj denotes the change rate of the Q-axis component of the current of the j.sup.th load in the common reference coordinate DQ: A/S; according to formulas (5)-(7), the large-signal model of microgrid including n distributed generations, s buses and p loads is as shown in formula (8): $\begin{array}{cc}\{\begin{array}{c}\stackrel{.}{x}=f\left(x\right)+gu\\ y_i=h_i\left(x\right)\end{array}& \mathrm{formula}\left(8\right)\end{array}$ wherein, x=[x.sub.inv1, . . . , x.sub.invn, i.sub.lineDQ1, . . . , i.sub.lineDQs, i.sub.loadDQ1, . . . , i.sub.loadDQ p].sup.T, x.sub.inv1 denotes the state variable of the 1.sup.st distributed generation, x.sub.invn denotes the state variable of the n.sup.th distributed generation, i.sub.lineDQ1=[i.sub.lineD1, i.sub.lineQ1].sup.T, i.sub.lineD1 denotes the D-axis component of the current of the 1.sup.st line in the common reference coordinate DQ, i.sub.lineQ1 denotes the Q-axis component of the current of the 1.sup.st line in the common reference coordinate DQ, i.sub.lineDQs=[i.sub.lineDs, i.sub.lineQs].sup.T, i.sub.lineDs denotes the D-axis component of the current of the s.sup.th line in the common reference coordinate DQ, i.sub.lineQs denotes the Q-axis component of the current of the s.sup.th line in the common reference coordinate DQ, i.sub.loadDQ1=[i.sub.loadD1, i.sub.loadQ1].sup.T, i.sub.loadD1 denotes the D-axis component of the current of the 1.sup.st load in the common reference coordinate DQ, i.sub.loadQ1 denotes the Q-axis component of the current of the 1.sup.st load in the common reference coordinate DQ, i.sub.loadDQp=[i.sub.loadDp, i.sub.loadQp].sup.T, i.sub.loadDp denotes the D-axis component of the current of the p.sup.th load in the common reference coordinate DQ, and i.sub.loadQp denotes the Q-axis component of the current of the p.sup.th load in the common reference coordinate DQ; u=[u.sub.1 . . . u.sub.n]T, u.sub.1 denotes the secondary control quantity of the 1.sup.st distributed generation, and u.sub.n denotes the secondary control quantity of the n.sup.th distributed generation; f(x) denotes a state function of the microgrid; g denotes an input matrix, and y.sub.i denotes an output value of the i.sup.th distributed generation; h.sub.i (x) denotes an output function of the i.sup.th distributed generation; the step 20) further comprising: the nonlinear state observer is a local Luenberger nonlinear state observer, as shown in formula (9):
{dot over ({circumflex over (x)})}=f({circumflex over (x)})+g u+L(h.sub.i({circumflex over (x)})y.sub.i)formula (9) in the formula, {circumflex over (x)} denotes the estimated state values of the microgrid in formula (8), {circumflex over (x)} denotes the change rate of the estimated state values in the microgrid; f({circumflex over (x)}) denotes a connection matrix of the microgrid under the action of the estimated value; L denotes a Luenberger state observer matrix; h.sub.i({circumflex over (x)}) denotes an output function of the i.sup.th distributed generation corresponding to the estimated state value {circumflex over (x)}; the step 30) further comprising: estimating the output voltages and output active and reactive powers of other distributed generations in real time according to local measured values of each distributed generation; the step 40) further comprising: implementing requirements of accurate reactive power sharing and voltage restoration based on the local measured values of each distributed generation and an estimated measurement of other distributed generations, a decentralized secondary voltage control, yielding a secondary voltage compensation, transmitting secondary voltage compensation to a PWM (Pulse Width Modulation) module of a local controller; sending generated PWM pulse signal to a drive and power amplifier unit to trigger a power electronic switching transistor; the reactive power sharing indicates that the output reactive power of each distributed generation is allocated according to the power capacity, and the implementation process is as shown in formula (10): $\begin{array}{cc}\{\begin{array}{c}{Q}_i=k_{\mathrm{PQ}}\left(Q_i^{*}-Q_i\right)+k_{\mathrm{iQ}}\left(Q_i^{*}-Q_i\right)\mathrm{dt}\\ Q_i^{*}=\frac{1\text{/}{n}_{\mathrm{Qi}}}{\underset{j=1}{\overset{n}{.Math.}}1\text{/}{n}_{\mathrm{Qj}}}\left(Q_i+\underset{j=1,ji}{\overset{n}{.Math.}}{\hat{Q}}_j\right)\end{array}& \mathrm{formula}\left(10\right)\end{array}$ in the formula, Q.sub.i denotes a reactive power control signal for each calculation cycle; k.sub.PQ denotes a proportional coefficient in the reactive power proportional integral controller; Q*.sub.i denotes the reactive power reference of the i.sup.th distributed generation; k.sub.iQ denotes an integral coefficient in the reactive power proportional integral controller; n.sub.Qj denotes a voltage droop coefficient of the j.sup.th distributed generation; {circumflex over (Q)}.sub.j denotes a reactive power estimation value of the j.sup.th distributed generation; the voltage restoration indicates that the average value of the output voltages of the distributed generations through the microgrid is restored to the rated value, and the implementation process is as shown in formula (11): $\begin{array}{cc}\{\begin{array}{c}{V}_i=k_{\mathrm{PE}}\left(V^{*}-\stackrel{\_}{V_i}\right)+k_{\mathrm{iE}}\left(V^{*}-\stackrel{\_}{V_i}\right)\mathrm{dt}\\ \stackrel{\_}{V_i}=\left(V_{\mathrm{odi}}+\underset{j=1,ji}{\overset{n}{.Math.}}{\hat{V}}_{\mathrm{odj}}\right)\text{/}n\end{array}& \mathrm{formula}\left(11\right)\end{array}$ in the formula, V.sub.i denotes an average voltage restoration control signal for each calculation cycle; k.sub.PE denotes a proportional coefficient in the average voltage controller; V* denotes a voltage rated value; V.sub.i denotes an average value of output voltages of all the distributed generations as estimated from the i.sup.th distributed generation; k.sub.iE denotes an integral coefficient in the average voltage controller; {circumflex over (V)}.sub.odj denotes an estimated value of the d-axis component of the output current of the j.sup.th distributed generation in the dq reference coordinate of the j.sup.th distributed generation; by combining formulas (10) and (11), a secondary voltage control input for the i.sup.th distributed generation is as shown in formula (12):
u.sub.i=Q.sub.i+V.sub.iformula (12).

2. The method according to claim 1, is characterized in that the common reference coordinate DQ refers to the dq reference coordinate of the 1.sup.st distributed generation, and the state variables of the remaining distributed generations, buses and loads are converted to the common reference coordinate DQ through coordinate transformation.

3. The method according to claim 1, is characterized in that in step 10), the loads are impedance-type loads.

4. The method according to claim 1, is characterized in that in step 10), the frequency droop characteristic coefficient m.sub.Pi is determined according to an active power capacity of each distributed generation; and the voltage droop characteristic coefficient n.sub.Qi is determined according to a reactive power capacity of each distributed generation.

5. The method according to claim 1, is characterized in that in step 40), the decentralized voltage control of reactive power sharing and average voltage restoration is implemented based on the estimated value of each distributed generation on real-time states of other distributed generations, which does not rely on communication.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a structural diagram of parallel inverter microgrid of the present invention;

(2) FIG. 2 is a detailed hardware implementation diagram of the present invention; Rf, Cf and Lf represent the resistance, inductance and capacitance of the filter of DG, respectively;

(3) FIG. 3 is a flow diagram according to an embodiment of the present invention;

(4) FIG. 4 is a block diagram of microgrid distributed generation control according to an embodiment of the present invention;

(5) FIG. 5 is a structure diagram of the decentralized voltage control according to an embodiment of the present invention;

(6) FIG. 6 is a diagram of a microgrid test system adopted in an embodiment of the present invention;

(7) FIG. 7(a) is a control effect diagram of output active power throughout a microgrid adopting a conventional latency-free centralized control method;

(8) FIG. 7(b) is a control effect diagram of output reactive power throughout the microgrid adopting the conventional latency-free centralized control method;

(9) FIG. 7(c) is a control effect diagram of the output voltage of the microgrid adopting the conventional latency-free centralized control method;

(10) FIG. 8(a) is a control effect diagram of output active power of a microgrid adopting a control method according to an embodiment of the present invention;

(11) FIG. 8(b) is a control effect diagram of output reactive power of the microgrid adopting the control method according to an embodiment of the present invention;

(12) FIG. 8(c) is a control effect diagram of the output voltage of the microgrid adopting the control method according to an embodiment of the present invention;

(13) FIG. 9(a) is a control effect diagram of output reactive power of a microgrid adopting a conventional latent centralized control method; and

(14) FIG. 9(b) is a control effect diagram of the output voltage of the microgrid adopting the conventional latent centralized control method.

DETAILED DESCRIPTION OF THE INVENTION

(15) For the purpose of making the objects, the technical schemes and the advantages of the present invention clearer, the following contents describe the present invention in details in combination with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used for interpreting the present invention, rather than limiting the present invention.

(16) The control method of the present invention can be applied to a parallel inverter microgrid. As shown in FIG. 1, establish a large-signal microgrid model of distributed generations, a connection network and impedance-type loads after scanning the connection/disconnection statuses of all DG units and loads, and querying the corresponding parameters. The microgrid voltage reference instruction is entered through a human-machine interface and sent to each distributed generation via the 485 communication mode. The distributed generation performs the decentralized voltage control and the diagram of detailed hardware implementation is as shown in FIG. 2. The data acquisition unit of each distribution generation collects the output voltage and current from the local sensor, which are to be sent to the nonlinear state observer of respective DSP to estimate the output voltages and reactive powers of other distributed generations. Based on the local measurement of each distributed generation and the estimated measurement of other distributed generations, the decentralized voltage control is implemented and the voltage compensation instruction is generated and transmitted to the PWM module of the local controller to trigger the inverter operation. As shown in FIG. 3, a decentralized voltage control method for a microgrid based on nonlinear state observers according to an embodiment of the present invention includes the following steps:

(17) Step 10), Suppose that there are N distributed generations in a microgrid. After scanning the connection/disconnection statuses of all DG units and loads, and querying the corresponding parameters, establish, a large-signal model of microgrid, with distributed generations, a connection network and impedance-type loads. Step 10) specifically includes:

(18) each distributed generation employs the droop control to generate the output voltage and frequency reference for individual inverters, as shown in formula (1):

(19) $\begin{array}{cc}\{\begin{array}{c}{}_i={}_n-m_{\mathrm{Pi}}{P}_i\\ k_{\mathrm{Vi}}{\stackrel{.}{V}}_{o,\mathrm{magi}}=V_n-V_{o,\mathrm{magi}}-n_{\mathrm{Qi}}{Q}_i\end{array}& \mathrm{formula}\left(1\right)\end{array}$

(20) In formula (1), .sub.i denotes the local angular frequency of the i.sup.th distributed generation; .sub.n denotes the local angular frequency reference value of the distributed generation: rad/s; m.sub.Pi denotes the frequency droop characteristic coefficient of the i.sup.th distributed generation: rad/s; P.sub.i denotes the actual output active power of the i.sup.th distributed generation: W; k.sub.Vi denotes the droop control gain of the i.sup.th distributed generation; {dot over (V)}.sub.o,magi denotes the change rate of the output voltage of the i.sup.th distributed generation: V/S; V.sub.n denotes the output voltage reference value of the distributed generation: V; V.sub.o,magi denotes the output voltage of the i.sup.th distributed generation: V; n.sub.Qi denotes the voltage droop characteristic coefficient of the i.sup.th distributed generation: V/Var; and Q.sub.i denotes the actual output reactive power of the it distributed generation: Var.

(21) The actual output active power P.sub.i and reactive power Q.sub.i of the i.sup.th distributed generation are obtained through a low-pass filter, as shown in formula (2):

(22) $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{P}}_i=-{}_{\mathrm{ci}}{P}_i+{}_{\mathrm{ci}}(V_{\mathrm{odi}}{i}_{\mathrm{odi}}+V_{\mathrm{oqi}}{i}_{\mathrm{oqi}}\\ {\stackrel{.}{Q}}_i=-{}_{\mathrm{ci}}{Q}_i+{}_{\mathrm{ci}}(V_{\mathrm{oqi}}{i}_{\mathrm{odi}}-V_{\mathrm{odi}}{i}_{\mathrm{oqi}}\end{array}& \mathrm{formula}\left(2\right)\end{array}$

(23) In formula (2), {dot over (P)}.sub.i denotes the change rate of active power of the i.sup.th distributed generation: W/S; .sub.ci denotes the low-pass filter cutoff frequency of the i.sup.th distributed generation: rad/s; V.sub.odi denotes the d-axis component of the output voltage of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: V; V.sub.oqi denotes the q-axis component of the output voltage of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: V; i.sub.odi denotes the d-axis component of the output current of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: A; i.sub.oqi denotes the q-axis component of the output current of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: A; and {dot over (Q)}.sub.i denotes the change rate of the reactive power of the i.sup.th distributed generation: Var/S.

(24) As the primary voltage control for each distributed generation enables the voltage magnitude on the q axis as zero, the secondary voltage control can be further obtained as formula (3):

(25) $\begin{array}{cc}\{\begin{array}{c}{k}_{\mathrm{Vi}}{\stackrel{.}{V}}_{\mathrm{odi}}=V_{\mathrm{ni}}-V_{\mathrm{odi}}-n_{\mathrm{Qi}}{Q}_i+u_i\\ V_{\mathrm{oqi}}=0\end{array}& \mathrm{formula}\left(3\right)\end{array}$

(26) In formula (3), {dot over (V)}.sub.odi denotes the change rate of the d-axis component of the output voltage of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: V/S; V.sub.ni denotes the output voltage reference value of the i.sup.th distributed generation, and u.sub.i denotes the secondary voltage control input: V.

(27) The dynamic equation of the output current of the distributed generation is as shown in formula (4):

(28) $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{i}}_{\mathrm{odi}}=-\frac{R_{ci}}{L_{ci}}{i}_{odi}+{}_i{i}_{oqi}+\frac{1}{L_{ci}}\left(V_{odi}-V_{busdi}\right)\\ {\stackrel{.}{i}}_{oqi}=-\frac{R_{ci}}{L_{ci}}{i}_{oqi}-{}_i{i}_{odi}+\frac{1}{L_{ci}}\left(V_{oqi}-V_{busqi}\right)\end{array}& \mathrm{formula}\left(4\right)\end{array}$

(29) In formula (4), i.sub.odi denotes the change rate of the d-axis component of the output current of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: A/S; R.sub.ci denotes the connection resistance from the i.sup.th distributed generation to the bus i: ; L.sub.ci denotes the connection inductance from the i.sup.th distributed generation to the bus i: H; V.sub.busdi denotes the d-axis component of the voltage of the bus i in the dq reference coordinate of the i.sup.th distributed generation; {dot over (i)}.sub.oqi denotes the change rate of the q-axis component of the output current of the i.sup.th distributed generation in the dq reference coordinate of the i.sup.th distributed generation: A/S; and V.sub.busqi denotes the q-axis component of the voltage of the bus i in the dq reference coordinate of the i.sup.th distributed generation: V.

(30) According to formulas (1) to (4), a dynamic equation of the it distributed generation is obtained, as shown in formula (5):

(31) $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{x}}_{invi}=f_{invi}\left(x_{invi}\right)+k_{invi}\left(x_{invi}\right)V_{busDQi}+h_{invi}{}_{com}+g_{invi}{u}_i\\ i_{oDQi}=C_{invci}{x}_{invi}\end{array}& \mathrm{formula}\left(5\right)\end{array}$

(32) In the formula, {dot over (x)}.sub.invi denotes the change rate of the state variable of the i.sup.th distributed generation, {dot over (x)}.sub.invi=[{dot over ()}.sub.i, {dot over (P)}.sub.i, {dot over (Q)}.sub.i, {dot over (V)}.sub.odi, {dot over (i)}.sub.odi, {dot over (i)}.sub.oqi].sup.T; {dot over ()}.sub.i denotes the change rate of .sub.i; x.sub.invi denotes the state variable of the i.sup.th distributed generation, x.sub.invi=[.sub.i, P.sub.i, Q.sub.i, V.sub.odi, i.sub.odi, i.sub.oqi].sup.T; wherein .sub.i denotes the phase angle difference between the dq axis of the dq reference coordinate of the i.sup.th distributed generation and the DQ axis in the common reference coordinate DQ of the microgrid: rad; f.sub.invi(x.sub.invi) denotes a state function of the i.sup.th distributed generation, and k.sub.invi(x.sub.invi) denotes a voltage disturbance function of the bus i; V.sub.busDQi=[V.sub.busDi, V.sub.busQi].sup.T, and V.sub.busDi denotes the D-axis component of the bus i in the common reference coordinate DQ: V; V.sub.busQi denotes the Q-axis component of the bus i in the common reference coordinate DQ: V; .sub.com denotes the angular frequency of the common reference coordinate: rad/s; h.sub.invi denotes a connection matrix of the angular frequency of the common reference coordinate; g.sub.invi denotes an input matrix of the it distributed generation; i.sub.oDQi=[i.sub.oDi, i.sub.oQi].sup.T, i.sub.oDi denotes the D-axis component of the output current of the i.sup.th distributed generation in the common reference coordinate DQ, and i.sub.oQi denotes the Q-axis component of the output current of the i.sup.th distributed generation in the common reference coordinate DQ: A; and C.sub.invi denotes an output matrix of the i.sup.th distributed generation.

(33) A current dynamic equation of the i.sup.th line between the bus i and the bus j is as shown in formula (6):

(34) $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{i}}_{\mathrm{lineDi}}=-\frac{r_{\mathrm{linei}}}{L_{\mathrm{linei}}}{i}_{\mathrm{lineDi}}+{}_{\mathrm{com}}{i}_{\mathrm{lineQi}}+\frac{1}{L_{\mathrm{linei}}}\left(V_{\mathrm{busDi}}-V_{\mathrm{busDj}}\right)\\ {\stackrel{.}{i}}_{\mathrm{lineQi}}=-\frac{r_{\mathrm{linei}}}{L_{\mathrm{linei}}}{i}_{\mathrm{lineQi}}-{}_{\mathrm{com}}{i}_{\mathrm{lineDi}}+\frac{1}{L_{\mathrm{linei}}}\left(V_{\mathrm{busQi}}-V_{\mathrm{busQj}}\right)\end{array}& \mathrm{formula}\left(6\right)\end{array}$

(35) In the formula, i.sub.lineDi denotes the change rate of the D-axis component of the current of the it line in the common reference coordinate DQ: A/S; r.sub.linei denotes the line resistance of the it line: ; L.sub.linei denotes the line inductance of the it line: H; i.sub.lineDi denotes the D-axis component of the current of the i.sup.th line in the common reference coordinate DQ, and i.sub.lineQi denotes the Q-axis component of the current of the i.sup.th line in the common reference coordinate DQ: A; V.sub.busDi denotes the D-axis component of the bus i in the common reference coordinate DQ, V.sub.busDj denotes the D-axis component of the bus j in the common reference coordinate DQ, and {dot over (i)}.sub.lineQi denotes the change rate of the Q-axis component of the current of the it line in the common reference coordinate DQ: A/S; V.sub.busQi denotes the Q-axis component of the bus i in the common reference coordinate DQ: V; and V.sub.busQj denotes the Q-axis component of the bus j in the common reference coordinate DQ: V.

(36) A current dynamic equation of the j.sup.th load connected to the bus j is as shown in formula (7):

(37) $\begin{array}{cc}\{\begin{array}{c}{\stackrel{.}{i}}_{\mathrm{loadDj}}=-\frac{R_{\mathrm{loadj}}}{L_{\mathrm{loadj}}}{i}_{\mathrm{loadDj}}+{}_{\mathrm{com}}{i}_{\mathrm{loadQj}}+\frac{1}{L_{\mathrm{loadj}}}{V}_{\mathrm{busDj}}\\ {\stackrel{.}{i}}_{\mathrm{loadQj}}=-\frac{R_{\mathrm{loadj}}}{L_{\mathrm{loadj}}}{i}_{\mathrm{loadQj}}-{}_{\mathrm{com}}{i}_{\mathrm{loadDj}}+\frac{1}{L_{\mathrm{loadj}}}{V}_{\mathrm{busQj}}\end{array}& \mathrm{formula}\left(7\right)\end{array}$

(38) In the formula, {dot over (i)}.sub.loadDj denotes the change rate of the D-axis component of the current of the j.sup.th load in the common reference coordinate DQ: A/S; R.sub.loadj denotes the load resistance of the j.sup.th load: ; L.sub.loadj denotes the load inductance of the j.sup.th load: H; i.sub.loadDj denotes the D-axis component of the current of the j.sup.th load in the common reference coordinate DQ, and i.sub.loadQj denotes the Q-axis component of the current of the j.sup.th load in the common reference coordinate DQ: A; and i.sub.loadQj denotes the change rate of the Q-axis component of the current of the j.sup.th load in the common reference coordinate DQ: A/S.

(39) According to formulas (5)-(7), the large-signal model of microgrid including n distributed generations, s buses and p loads is as shown in formula (8):

(40) $\begin{array}{cc}\{\begin{array}{c}\stackrel{.}{x}=f\left(x\right)+gu\\ y_i=h_i\left(x\right)\end{array}& \mathrm{formula}\left(8\right)\end{array}$

(41) Wherein, x=[x.sub.inv1 . . . x.sub.invn, i.sub.lineDQ1 . . . i.sub.lineDQs, i.sub.loadDQ1 . . . i.sub.loadDQ p].sup.T, x.sub.inv1 denotes the state variable of the 1.sup.st distributed generation, x.sub.invn denotes the state variable of the n.sup.th distributed generation, i.sub.lineDQ1=[i.sub.lineD1, i.sub.lineQ1].sup.T, i.sub.lineD1 denotes the D-axis component of the current of the 1.sup.st line in the common reference coordinate DQ, i.sub.lineQ1 denotes the Q-axis component of the current of the 1.sup.st line in the common reference coordinate DQ, i.sub.lineDQs=[i.sub.lineDs, i.sub.lineQs].sup.T, i.sub.lineDs denotes the D-axis component of the current of the st line in the common reference coordinate DQ, i.sub.lineQs denotes the Q-axis component of the current of the s.sup.th line in the common reference coordinate DQ, i.sub.loadDQ1=[i.sub.loadD1, i.sub.loadQ1].sup.T, i.sub.loadD1 denotes the D-axis component of the current of the 1.sup.st load in the common reference coordinate DQ, i.sub.loadQ1 denotes the Q-axis component of the current of the 1.sup.st load in the common reference coordinate DQ, i.sub.loadDQp=[i.sub.loadDp, i.sub.loadQp].sup.T, i.sub.loadDp denotes the D-axis component of the current of the pt load in the common reference coordinate DQ, and i.sub.loadQp denotes the Q-axis component of the current of the p.sup.th load in the common reference coordinate DQ; u=[u.sub.1 . . . u.sub.n].sup.T, u.sub.1 denotes the secondary control quantity of the 1.sup.st distributed generation, and u.sub.n denotes the secondary control quantity of the n.sup.th distributed generation; f(x) denotes a state function of the microgrid; g denotes an input matrix, and y.sub.i denotes an output value of the i.sup.th distributed generation; and h.sub.i(x) denotes an output function of the i.sup.th distributed generation.

(42) Step 20), establish a Luenberger-like nonlinear state observer for each distributed generation. Step 20) specifically includes:

(43) according to the microgrid large-signal model established in step 10), establish a local Luenberger nonlinear state observer for each distributed generation in the DSP, as shown in formula (9):
{dot over ({circumflex over (x)})}=f({circumflex over (x)})+g u+L(h.sub.i({circumflex over (x)})y.sub.i)formula (9)

(44) In the formula, {circumflex over (x)} denotes the estimated state values of the microgrid in formula (8), {dot over ({circumflex over (x)})} denotes the change rate of the estimated state values in the microgrid; f({circumflex over (x)}) denotes a connection matrix of the microgrid under the action of the estimated value; L denotes a Luenberger state observer matrix; and h.sub.i({circumflex over (x)}) denotes an output function of the i.sup.th distributed generation corresponding to the estimated state value {circumflex over (x)}.

(45) Step 30), The data acquisition module of each distributed generation collects the output voltage V.sub.odi and output current i.sub.odi and i.sub.oqi from the local sensor, which are to be sent to the nonlinear state observer of respective DSP in step 20). Estimate the output voltages and output powers other distributed generations in real time according to a local measured value of each distributed generation

(46) Step 40), The microgrid voltage reference instruction is entered through a human-machine interface and sent out to each distributed generation via the 485 communication mode. Based on the local measured value of each distributed generation and the estimated measurement of other distributed generations, the decentralized secondary voltage control is implemented to satisfy the requirements of accurate reactive power sharing and voltage restoration. The resultant secondary voltage compensation term is transmitted to the PWM module of the local controller; the generated PWM pulse signal is sent to the drive and power amplifier unit to trigger the power electronic switching transistor. Step 40) specifically includes:

(47) the reactive power sharing indicates that the output reactive power of each distributed generation is allocated according to the power capacity, and the implementation process is as shown in formula (10):

(48) $\begin{array}{cc}\{\begin{array}{c}{Q}_i=k_{\mathrm{PQ}}\left(Q_i^{*}-Q_i\right)+k_{\mathrm{iQ}}\left(Q_i^{*}-Q_i\right)\mathrm{dt}\\ Q_i^{*}=\frac{1/n_{\mathrm{Qi}}}{\underset{j=1}{\overset{n}{.Math.}}1/n_{\mathrm{Qj}}}\left(Q_i+\underset{j=1,ji}{\overset{n}{.Math.}}{\hat{Q}}_j\right)\end{array}& \mathrm{formula}\left(10\right)\end{array}$

(49) In the formula, Q.sub.i denotes a reactive power control signal for each calculation cycle; k.sub.PQ denotes a proportional coefficient in the reactive power proportional integral controller; Q*.sub.i denotes the reactive power reference of the i.sup.th distributed generation; k.sub.iQ denotes an integral coefficient in the reactive power proportional integral controller; n.sub.Qj denotes a voltage droop coefficient of the j.sup.th distributed generation; and {circumflex over (Q)}.sub.j denotes a reactive power estimation value of the j.sup.th distributed generation.

(50) The voltage restoration indicates that the average value of the output voltages of the distributed generations through the microgrid is restored to the rated value, and the implementation process is as shown in formula (11):

(51) 0$\begin{array}{cc}\{\begin{array}{c}{V}_i=k_{\mathrm{PE}}\left(V^{*}-{\stackrel{\_}{V}}_i\right)+k_{\mathrm{iE}}\left(V^{*}-{\stackrel{\_}{V}}_i\right)\mathrm{dt}\\ {\stackrel{\_}{V}}_i=\left(V_{\mathrm{odi}}+\underset{j=1,ji}{\overset{n}{.Math.}}{\hat{V}}_{\mathrm{odj}}\right)/n\end{array}& \mathrm{formula}\left(11\right)\end{array}$

(52) In the formula, V.sub.i denotes an average voltage restoration control signal for each calculation cycle; k.sub.PE denotes a proportional coefficient in the average voltage controller; V* denotes a voltage rated value; V.sub.i denotes an average value of output voltages of all the distributed generations as estimated from the i.sup.th distributed generation; k.sub.iE denotes an integral coefficient in the average voltage controller; and {circumflex over (V)}.sub.odj denotes an estimated value of the d-axis component of the output current of the j.sup.th distributed generation in the dq reference coordinate of the j.sup.th distributed generation.

(53) By combining formulas (10) and (11), the secondary voltage control input for the i.sup.th distributed generation is as shown in formula (12):
u.sub.i=Q.sub.i+V.sub.iformula (12).

(54) In the above embodiment, the common reference coordinate DQ refers to the dq reference coordinate of the 1.sup.st distributed generation, and the state variables of the remaining distributed generations, buses and loads are converted to the common reference coordinate DQ through coordinate transformation. In step 10), the frequency droop coefficient m.sub.Pi is determined according to the active power capacity of each distributed generation; and the voltage droop characteristic coefficient n.sub.Qi is determined according to the reactive power capacity of each distributed generation. In step 40), the decentralized voltage control of reactive power sharing and average voltage restoration is implemented based on the estimated value of each distributed generation on real-time states of other distributed generations, which does not rely on communication.

(55) In the present embodiment, the state of each distributed generation in the microgrid is estimated in real time through a nonlinear state observer based on a microgrid large-signal model, thereby realizing the decentralized secondary control in the microgrid. Compared with the conventional centralized or distributed secondary control method, the control method provided in the present embodiment avoids the problems of communication latency and information drop-out due to the communication independence. According to the microgrid voltage control method based on the nonlinear state observer in the present embodiment, the output voltage and output reactive power of other distributed generations are estimated in real time through operation data of local output voltage, output current and the like, so that the reactive power sharing and average voltage restoration in microgrid independent from a communication technology are realized, the power quality is effectively improved and the system stability and the dynamic performance are improved.

(56) The block diagram of distributed generation control in the embodiment of the present invention is as shown in FIG. 4. The control block diagram mainly includes three parts, wherein the first part is power calculation, the second part is droop control, and the third part is voltage and current double-loop control. The power calculation includes acquiring the local output voltage and output current, and calculating the output active power and reactive power; the power calculated by the power calculation is obtained by a low-pass filter, and the droop control includes calculating the output voltage reference value and frequency reference value of each distributed generation; and the output voltage reference is transmitted to the voltage and current double-loop controller, and the frequency reference value acts on inverter operation.

(57) The structure block diagram of the decentralized voltage control in the embodiment of the present invention is as shown in FIG. 5.N distributed generations form a microgrid through a connection network, and each distributed generation acquires local operation data, estimates output voltages and output reactive power of other distributed generations in real time according to the local nonlinear state observer, and generates a secondary voltage control input in the local secondary controller, which is eventually transmitted to the local primary controller. Through the real-time estimation of the nonlinear state observers, decentralized secondary control which does not rely on high-performance sensors and high-pass bandwidth is implemented, thereby avoiding the problems of communication latency and data drop-out in the communication process, effectively improving the power quality.

(58) An embodiment is given below.

(59) A simulation system is as shown in FIG. 6. The microgrid consists of three distributed generations, two connection lines and two loads. The load 1 is connected to bus 1, and the load 2 is connected to bus 2. The loads in the system are impedance-type loads. Assuming that the capacity ratio of the distributed generation 1, the distributed generation 2 and the distributed generation 3 is 1:1.5:2, the frequency droop coefficients and the voltage droop coefficients for corresponding ratios are designed so that the desired output active power and reactive power ratio of the distributed generations is 1:1.5:2. A simulated microgrid model is built based on the MATLAB/Simulink platform. The microgrid voltage control effect is simulated to compare the secondary voltage control method of the present invention with the conventional secondary voltage control method. The conventional secondary voltage control method adopts the centralized control format.

(60) FIG. 7 shows simulation results when the microgrid adopts a conventional centralized control method without communication latency. At the beginning of operation, each distributed generation is operated in a droop control mode, and a centralized secondary control method is adopted at 0.5 s. A 10 kW/5 kVar load is added to bus 3 at 1.5 s, and the load is removed at 2.5 s, with the simulation results shown in FIG. 7. FIG. 7(a) shows an output curve diagram of active power of each distributed generation in the microgrid, where the abscissa represents time in second, and the ordinate represents active power in watt. As shown in FIG. 7(a), at the beginning, under the action of droop control, the output active power of each distributed generation is allocated according to the ratio of the droop coefficients, that is, P1:P2:P3=1:1.5:2. After 1.5 s, the power is increased proportionally for each distributed generation to meet the power supply for the newly added load. After 2.5 s, the power is reduced proportionally for each distributed generation. From FIG. 7(a), it can be known that active power is supplied to each distributed generation according to the capacity regardless of droop control or secondary control. FIG. 7(b) is the curve diagram of reactive power of individual distributed generations in the microgrid, where the abscissa represents time in second, and the ordinate represents reactive power in var. From FIG. 7(b), it can be seen that the reactive power sharing effect is not ideal under the droop action at first; and after 0.5 s, no matter whether the load increases or decreases, the reactive power is allocated according to the ratio of the droop coefficient, that is, Q1:Q2:Q3=1:1.5:2. According to FIG. 7(b), under the action of secondary control, the reactive power sharing effect of the microgrid is significantly improved. FIG. 7 (c) is the curve diagram of output voltage of individual distributed generations in the microgrid, where the abscissa represents time in second, and the ordinate represents voltage in volt. From FIG. 7(c), it can be seen that the output voltage of each distributed generation under the droop action has a steady-state deviation at first; and after 0.5 s, no matter whether the load increases or decreases, the voltage magnitude is improved and the average voltage is restored to a rated value. According to FIG. 7(c), under the action of secondary control, the dynamics of the voltage throughout the microgrid is improved.

(61) FIG. 8 shows simulation results when the microgrid adopts a decentralized control method according to an embodiment of the present invention. At the beginning of operation, each distributed generation operates in a droop control mode, and a centralized secondary control method is adopted at 0.5 s. A 10 kW/5 kVar load is added for the bus 3 at 1.5 s, and the load is removed at 2.5 s. The simulation results are as shown in FIG. 8. FIG. 8(a) shows the output curve diagram of active power of each distributed generation in the microgrid, where the abscissa represents time in second, and the ordinate represents active power in watt. As shown in FIG. 8(a), at the beginning, under the action of droop control, the output active power of each distributed generation is allocated according to the ratio of the droop coefficient, that is, P1:P2:P3=1:1.5:2. After 1.5 s, the power is increased proportionally for each distributed generation to meet the power supply for the newly added load. After 2.5 s, the power is reduced proportionally for each distributed generation. FIG. 8(b) is the curve diagram of reactive power of individual distributed generations in the microgrid, where the abscissa represents time in second, and the ordinate represents reactive power in var. From FIG. 8(b), it can be seen that the reactive power sharing effect is not ideal under the droop action at first; and after 0.5 s, no matter whether the load increases or decreases, the reactive power is allocated according to the ratio of the droop coefficient. According to FIG. 8(b), under the action of secondary control, the reactive power sharing effect of the microgrid is significantly improved. FIG. 8 (c) is the curve diagram of output voltage of individual distributed generations in the microgrid, where the abscissa represents time in second, and the ordinate represents voltage in volt. From FIG. 8(c), it can be seen that the output voltage of each distributed generation under the droop action has a steady-state deviation at first; and after 0.5 s, no matter whether the load increases or decreases, the voltage magnitude is improved and the average voltage is restored to a rated value. From FIGS. 7(a)-(c) and FIGS. 8(a)-(c), compared with a latency-free centralized control method, the control effect of the decentralized control method according to the embodiment of the present invention slightly declines in the absence of communication latency, but it still can achieve the control objectives of reactive power sharing and voltage restoration.

(62) In order to show the advantages of the decentralized control method according to the embodiment of the present invention regarding the communication-free characteristics, FIG. 9 shows simulation results of the microgrid adopting a communication latency 200 ms. At the beginning of operation, each distributed generation is operated in a droop control mode, and a centralized secondary control method is adopted at 0.5 s. A 10 kW/5 kVar load is added for the bus 3 at 1.5 s, and the load is removed at 2.5 s. The simulation results are as shown in FIG. 9. FIG. 9(a) shows the curve diagram of output reactive power of each distributed generation in the microgrid, where the abscissa represents time in second, and the ordinate represents active power in var. FIG. 9(b) is the curve diagram of output voltage of each distributed generation in the microgrid, where the abscissa represents time in second, and the ordinate represents the voltage magnitude in volt. As shown in FIGS. 9(a) and 9(b), under the communication latency 200 ms, both the reactive power and the output voltage experience violent oscillation and eventually the system is instable. Therefore, the communication latency has an important influence on the dynamic performance of the system. However, the decentralized control method in the embodiment of the present invention, as shown in FIGS. 8(a)-8(c), can avoid the influence of the communication process.

(63) The control method according to the embodiment of the present invention is a decentralized voltage control method based on nonlinear state observers in a microgrid. The state observer of the local distributed generation estimates the output voltage and reactive power of other distributed generations in real time for secondary voltage control to achieve reactive power sharing and average voltage restoration. The method of the present embodiment is a decentralized control method, which does not rely on remote measurement and communication network, thereby avoiding the problems of communication latency and data drop-out in the conventional voltage control and improving the accurate reactive power sharing and voltage dynamics of the microgrid.

(64) The basic principles, main features and advantages of the present invention are shown and described above. It should be understood by the technical personnel in the field that the present invention is not limited by the specific embodiments described above, and the descriptions in the foregoing specific embodiments and specification are only for further illustrating the principle of the present invention.

(65) Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention, with all these changes and modifications falling within the scope of the present invention. The protection scope of the present invention is defined by the claims and the equivalents thereof.