Method of Calculating Power Flow Solution of a Power Grid that Includes Generalized Power Flow Controllers

Abstract

A method to incorporate the steady-state model of the generalized power flow controller into a Newton-Raphson power flow algorithm is disclosed. The disclosed method adopts a flexible steady-state model of the generalized power flow controller, which can be applied to calculate the power flow solution of a power grid embedded with STATCOM, UPFC, GUPFC and the generalized power flow controller in a single framework. The disclosed method only incorporates the control variables of the shunt voltage sourced converter into the state vector of Newton-Raphson power flow algorithm. The increment of state variables due to incorporating the generalized power flow controller is less than the prior art. Further, the method can preserve the quadratic convergence characteristic of the Newton-Raphson power flow algorithm after embedding the generalized power flow controller into a power grid.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

The present invention relates generally to the power flow analysis of a power grid embedded with a generalized power flow controller in the technical field, and more particularly to a method to incorporate the steady-state model of a generalized power flow controller into a conventional Newton-Raphson power flow algorithm, which can be applied to calculate the power flow solution of a power grid embedded with the generalized power flow controller.

2. Description of Related Art

In the last decade, the power industry has extensively employed an innovative Flexible Alternative Current Transmission System (FACTS) technology to improve the utilization of existing transmission facilities. Connecting several Voltage Sourced Converters (VSCs) together forms various multiple functional FACTS controllers, such as: Static Synchronous Compensator (STATCOM), Unified Power Flow Controller (UPFC), and Generalized Unified Power Flow Controller.

The STATCOM, as shown in FIG. 1, has one shunt VSC. The AC side of the shunt VSC connects to a bus of a power grid through a coupling transformer, and the DC side connects to a capacitor. The STATCOM provides reactive power compensation to regulate the voltage magnitude of the connected bus at a fix level.

The UPFC, as shown in FIG. 2, has one shunt VSC and one series VSC. The AC side of the shunt VSC connects to a bus through a shunt coupling transformer, whereas the AC side of the series VSC is in series with a transmission line through a series coupling transformer. The DC sides of the shunt and series VSCs sharing the same capacitor. The shunt VSC can regulate the voltage magnitude of the connected bus at a fixed level, and the series VSC can control the active and reactive power of the connected transmission line.

The GUPFC, as shown in FIG. 3, comprises one shunt VSC and a plurality of series VSCs. The connections and functions of the VSCs of GUPFC like those in the UPFC. These VSCs enable the GUPFC to control the voltage magnitude of the bus connected with the shunt VSC, and to control the active and reactive power of each transmission line which is in series with the series VSC.

The disclosed generalized power flow controller has a more flexible structure than the GUPFC. Comparing FIG. 3 and FIG. 4, both the GUPFC and the generalized power flow controller has one shunt branch and a plurality of series branches. In GUPFC, the shunt branch and the receiving-end of each series branch connect to a common bus, bus s₁, however, in the disclosed generalized power flow controller, the shunt branch and the receiving-end of each series branch are allowed to connect to different buses, bus s₁, bus s₂˜s_n, respectively.

The versatility of the generalized power flow controller can be applied to equalize both the active and reactive power in the transmission lines, relieve the overloaded transmission lines from the burden of reactive power flow, and restore for declines in resistive as well as reactive voltage drops.

Developing a steady-state model of the generalized power flow controller is fundamentally important for a power flow analysis of a power grid embedded with the generalized power flow controller. The power flow analysis provides the information of impacts on a power system after installing the generalized power flow controller. Many steady-state models of STATCOM, UPFC and GUPFC applied to power flow analysis have been set forth. In 2000, a STATCOM steady-state model accounting for the high-frequency effects and power electronic losses is proposed in an article “An improved StatCom model for power flow analysis”, by Zhiping Yang; Chen Shen; Crow, M. L.; Lingli Zhang; in IEEE Power Engineering Society Summer Meeting, 2000, Volume 2, Page(s):1121-1126.

A conventional approach to calculate the power flow solution of a power grid that includes a unified power flow controller is disclosed in an article “Unified power flow controller: a critical comparison of Newton-Raphson UPFC algorithms in power flow studies” by C. R. Fuerte-Esquivel and E. Acha in IEE Proc. Generation, Transmission & Distribution, 1997, and in an article “A comprehensive Newton-Raphson UPFC model for the quadratic power flow solution of practical power network” by C. R. Fuerte-Esquivel, E. Acha and H. Ambriz-Perez in IEEE Trans. Power System, 2000. In 2003, X.-P. Zhang developed a method to incorporate a voltage sourced based model of GUPFC into a Newton-Raphson power flow algorithm in an article “Modeling of the interline power flow controller and the generalized unified power flow controller in Newton power flow”, IEE Proceedings. Generation, Transmission & Distribution, Vol. 150, No. 3, May. 2003, pp. 268-274. The method included the voltage magnitude and phase angle of the equivalent voltage source into the state vector of Newton-Raphson iteration formula. The number of appended state variables is twice the number of VSCs. Thus, the length of state vector is varied depending on the number of VSCs. Therefore, the prior art can only be applied to the case with fixed number of VSCs. It can not be extended to the applications of STATCOM and UPFC. Besides, the speed of convergence is sensitive to the initial values of control variables of GUPFC. The initial values of control variables need a careful selection. Improper selection of the control variables may cause the solutions oscillating or even divergent.

Although the steady-state models of STATCOM, UPFC and GUPFC have been widely discussed individually, a method to incorporate steady-state models of STATCOM, UPFC, GUPFC and the generalized power flow controller into a Newton-Raphson power flow algorithm in a single framework have not been disclosed.

SUMMARY OF THE INVENTION

It is, therefore, an object of the present invention to provide a method to incorporate the steady-state model of a generalized power flow controller into a Newton-Raphson power flow algorithm with a robustness and rapid convergence characteristic, wherein the convergence speed is not sensitive to the selection of initial values of control variables of the generalized power flow controller.

It is another object of the present invention to provide a method to incorporate the steady-state model of the generalized power flow controller into a Newton-Raphson power flow algorithm, wherein the steady-state model has a flexible structure which can be applied to calculate the power flow solution of a power grid embedded with STATCOM, UPFC, GUPFC and the generalized power flow controller in a single framework.

To carry out previously mentioned objects, an innovative steady-state model of the generalized power flow controller is disclosed. The steady-state model has a flexible structure, wherein the sending-end of the each series VSC doesn't confine to connect to the same bus as the shunt converter connected. The feature of the steady-state model is expressing the variables of the steady-state model in a rectangular coordinate. Transforming the phasor from a conventional polar coordinate into d-q components reduces the appended state variables in the Newton-Raphson iteration. As a result, the increased iterations introducing by the generalized power flow controller is fewer than the prior art. The power flow calculation can preserve a rapid convergence characteristic.

In addition, a method to incorporate the steady-state model of the generalized power flow controller is disclosed. The method only incorporates the control variables of the shunt VSC into the state vector of Newton-Raphson power flow algorithm. The equivalent voltages of the series VSCs are calculated directly from the power flow control objectives and the bus voltages. Thus, the length of the state vector is the same regardless the number of series VSCs. As a result, the present invention can be utilized to calculate the power flow solution of a power grid embedded with STATCOM, UPFC, GUPFC and the generalized power flow controller.

The above and other objects and efficacy of the present invention will become more apparent after the description takes from the preferred embodiments with reference to the accompanying drawings is read.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is the interconnection of a STATCOM and a power grid according to the present invention;

FIG. 2 is the interconnection of a UPFC and a power grid according to the present invention;

FIG. 3 is the interconnection of a GUPFC and a power grid according to the present invention;

FIG. 4 is the interconnection of a generalized power flow controller and a power grid according to the present invention;

FIG. 5 is an equivalent circuit of a generalized power flow controller according to the present invention;

FIG. 6 is a flow chart for finding the power flow solution of a power grid embedded with the generalized power flow controller according to the present invention;

FIG. 7 shows the progress of control variables of the generalized power flow controller at each iteration according to the present invention.

FIG. 8 shows power mismatches of buses of the generalized power flow controller at each iteration according to the present invention; and

FIG. 9 shows a quadratic convergence pattern of the solution process according to the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

The generalized power flow controller 100 is a multi-functional FACTS controller. As depicted in FIG. 4, the generalized power flow controller comprises a plurality of voltage sourced converters (VSCs) 111, 121, 131 and a plurality of coupling transformers 112, 122, 132. These VSCs are connected back-to-back to share a common DC bus. The AC sides of VSCs couple to a power grid through coupling transformers, and the DC sides of VSCs link together to a DC coupling capacitor 110.

One of these VSCs, VSC₁ 111, connects to an AC bus 113 in parallel, and the other VSCs 121, 131 coupled to transmission lines 125, 135 in series. These VSCs exchange active power via the common DC bus. The shunt VSC, VSC₁ 111, can provide the reactive power compensation to regulate the voltage magnitude at its connected bus s₁ 113, whereas each of the series VSCs, VSC₂-VSC_n 121, 131, can provide both the active and reactive power compensation to concurrently control the active and reactive power of the connected transmission line 125, 135.

The main function of VSC₁ is to keep a fixed DC voltage at DC bus by balancing the active power transfer among VSCs. The remaining capacity of VSC₁ is utilized to regulate the voltage magnitude at bus s₁. In other words, the active power generated/absorbed by VSC₁ is restricted by the operation of other VSCs. Thus, the VSC₁ 111 has only one control degree of freedom. It can provide the reactive power compensation to regulate the voltage magnitude of bus s₁. On the other hand, each of series VSCs, VSC₂-VSC_n 121, 131, has two control degrees of freedom. It can simultaneously provide the active and reactive power compensation to control the active and reactive power in transmission line.

The equivalent circuit of the generalized power flow controller according to the present invention is derived next. As shown in FIG. 5. The equivalent circuit includes one shunt branch and a plurality of series branches. Each branch comprises an equivalent voltage source in series with an impedance, wherein the equivalent voltage source models the VSC, and the impedance models the coupling transformer. The operation of these equivalent voltage source is dependent on each other. An active power balance equation, which will be derived later, must be satisfied to conform the energy conservation law.

The distinct feature of the present invention is expressing the control variables of the equivalent circuit in a rectangular coordinate. These variables are decomposed into d-q components by an orthogonal projection technique. For each generalized power flow controller, the voltage of the bus connecting the shunt branch is chosen as a reference phasor. The d component is in phase with the reference phasor, whereas the q component leads the reference phasor by 90 degree. For examples, the d-q decomposition on a voltage phasor, V_xk=|V_xk|∠θ_xk, is expressed as:

V _xk ^D =|V _rk| cos(θ_rk−θ_s1); V_xk^Q=|V_rk| sin(θ_rk−θ_s1)  eq. (1)

where θ_s1, is the phase angle of the voltage at bus s1. The superscripts “D” and “Q” symbolize the d-q components of the corresponding variables, subscript “k” is the index of the VSC. The subscript “x” can be replaced with “s”, “r”, “sh” or “ser” to represent variables related to the sending-end, receiving-end, shunt branch and series branch, respectively. The d-q decomposition of a current phasor can be performed in a similar way.

The steady-state model of the generalized power flow controller can be incorporated into a Newton-Raphson power flow algorithm by replacing the generalized power flow controller with equivalent loads at the ends connected with the power grid. By the definition of the complex power, the equivalent load of the shunt branch is:

$\begin{array}{cc}\left[\begin{array}{c}{P}_{s1}\\ Q_{s1}\end{array}\right]=\left[\begin{array}{cc}V_{s1}& 0\\ 0& -V_{s1}\end{array}\right]\left[\begin{array}{c}{I}_{\mathrm{sh}}^D\\ I_{\mathrm{sh}}^Q\end{array}\right],.& \mathrm{eq}.\left(2\right)\end{array}$

where I_sh^D and I_sh^Q are the d-q current components of the shunt branch. The equivalent load at the receiving-end of each series branch is set to achieve a power flow control objective as,

$\begin{array}{cc}\left[\begin{array}{c}{P}_{\mathrm{rk}}\\ Q_{\mathrm{rk}}\end{array}\right]=-\left[\begin{array}{c}{P}_{\mathrm{linek}}^{\mathrm{ref}}\\ Q_{\mathrm{linek}}^{\mathrm{ref}}\end{array}\right],k=2,\Lambda ,n,& \mathrm{eq}.\left(3\right)\end{array}$

where n is the total number of VSCs, P_linek^ref and Q_linek^ref are the reference commands of the active and reactive power from the receiving-end of the kth series branch toward the connected transmission line. The equivalent load at the sending-end of the kth series branch is,

$\begin{array}{cc}\left[\begin{array}{c}{P}_{\mathrm{sk}}\\ Q_{\mathrm{sk}}\end{array}\right]=-\left[\begin{array}{cc}{V}_{\mathrm{sk}}^D& V_{\mathrm{sk}}^Q\\ V_{\mathrm{sk}}^Q& -V_{\mathrm{sk}}^D\end{array}\right]\left[\begin{array}{c}{I}_{\mathrm{serk}}^D\\ I_{\mathrm{serk}}^Q\end{array}\right],k=2,\Lambda ,n,& \mathrm{eq}.\left(4\right)\end{array}$

where I_serk^D and I_serk^Q are the d-q current components of the kth series branch, which can be obtained explicitly as:

$\left[\begin{array}{c}{I}_{\mathrm{serk}}^D\\ I_{\mathrm{serk}}^Q\end{array}\right]=-\frac{1}{V_{\mathrm{rk}}^2}\left[\begin{array}{cc}{V}_{\mathrm{rk}}^D& V_{\mathrm{rk}}^Q\\ V_{\mathrm{rk}}^Q& -V_{\mathrm{rk}}^D\end{array}\right]\left[\begin{array}{c}{P}_{\mathrm{rk}}\\ Q_{\mathrm{rk}}\end{array}\right].$

Balancing the active power transfer among VSCs is a main function of VSC₁. The remaining capacity of VSC₁ can provide the reactive power compensation to regulate the voltage magnitude of the connected bus at a fixed level. Therefore, the voltage magnitude at the bus connecting the shunt branch can be set to achieve a voltage magnitude control objective,

|V_s1|=V_s1^ref,  eq. (5)

where V_s1^ref is the desired voltage magnitude at the bus s₁.Under the lossless assumption of the VSCs, the sum of the active power generated by the VSCs must equal to zero. Therefore, the active power generated by the VSCs must be constrained by an active power balance equation,

$\begin{array}{cc}{P}_{\mathrm{dc}}=P_{\mathrm{sh}}+\sum _{k=2}^nP_{\mathrm{serk}}=0,& \mathrm{eq}.\left(6\right)\end{array}$

where P_sh is the active power generated from the equivalent voltage source of the shunt branch, and P_serk is the active power generated from the equivalent voltage source of the kth series branch,After simple algebra manipulations, P_sh and P_serk, can be expressed as:

P _sh =I _sh ^D V _s1 ^D+(I_sh^D2+I_sh^Q2)R_sh

P _serk =I _serk ^D(V_rk^D−V_sk^D)+I_serk^Q(V_rk^Q−V_sk^Q)+(I_serk^D2+I_serk^Q2)R_serk

Each of STATCOM, UPFC, GUPFC and the generalized power flow controller has different numbers of series VSCs. However, they are in common by having one shunt VSC. Consequently, UPFC, STATCOM and GUPFC can be regarded as a subdevice of the generalized power flow controller. For example, if the shunt branch and series branches share the same sending-end bus, ie. bus s₁, s₂ and s_n connect together, the foregoing derivations can be applied to the GUPFC. Similarly, UPFC has only one series branch, set n=2 in Eq (6) in a UPFC application. Furthermore, because the STATCOM has no series branch, the summation part of Eq (6) is omitted, ie. Eq. (6) becomes P_dc=P_sh, in a STATCOM application.

In the present invention, regardless of the number of series VSCs, and whether the shunt branch and the series branch share the same sending-end or not, the power flow solution can be found under the same procedures. That is, the present invention can be utilized to calculate the power flow solution of a power grid embedded with STATCOM, UPFC, GUPFC and the generalized power flow controller.

The power flow solution can be obtained by solving power flow equations, which is a set of nonlinear equations describing the power balance at each bus of a power grid. The Newton-Raphson power flow algorithm is an iterative procedure to solve power flow equations. The iterative formula of the algorithm is expressed as,

x ⁱ⁺¹ =x ⁽ⁱ⁾ −[J(x⁽ⁱ⁾)]⁻¹f(x⁽ⁱ⁾),  eq. (7)

where x is a state vector, f(x) is a mismatch vector, and i means the ith iteration. The elements of the state vector are called state variables which include the voltage magnitudes and the phase angles of buses of a power grid. The elements of the mismatch vector include the net active and reactive power flowing into each bus, and the other constraints of the power system. J is a corresponding Jacobian matrix which is formed by the first-order partial derivatives of the mismatch vector. After considering the equivalent loads of the generalized power flow controller, the mismatch vector is modified as:

f′=f+Δf _GUPFC,  eq. (8)

where Δf_GUPFC=[Δf_{bus |Δfcontrol}]^T=[P_s1 Q_s1 P_sk Q_sk P_rk Q_rk | P_dc]^T,The first part of Δf_GUPFC, Δf_bus, relates to the equivalent loads at the ends of the generalized power flow controller. The elements of Δf_bus are added to the corresponding position of f. The second part of Δf_GUPFC, Δf_control, is the added constraints introduced by the generalized power flow controller. The element of Δf_control augments the size of the mismatch vector. Therefore, the length of f′ is longer than that of f by 1. The elements of Δf_GUPFC have been derived in eq. (2), (3), (4) and (6).

With regard to the state vector of the iteration formula, Instead of selecting the voltage magnitudes and phase angles as state variables, the d-q current components of the shunt branch have been chosen as state variables. Hence, elements of the state vector associated with the generalized power flow controller are expressed as:

x _GUPFC =[x _bus |x _control]^T=[θ_s1θ_sk|V_sk|θ_rk|V_rk||I_sh^DI_sh^Q]^T,  eq. (9)

where x_bus consists of the original state variables relevant to the generalized power flow controller, and x_control consists of the added state variables introduced by the generalized power flow controller. Because |V_s1| is regulated by I_sh^Q at a fixed voltage level, |V_s1| has been omitted from x_bus. The elements of x_control augments the size of the original state vector. Thus, one element is omitted and two new elements are appended to the state vector. The length of the state vector is increased by one after embedding the generalized power flow controller. The Jacobian matrix is also modified according to the first-order partial derivatives of f′ as:

$\begin{array}{cc}{J}^{\prime }=J+\Delta J_{\mathrm{GUPFC}},& \mathrm{Eq}.\left(10\right)\end{array}$ $\mathrm{where}$ $\Delta J_{\mathrm{GUPFC}}=\left[\begin{array}{cccccccc}0& 0& 0& 0& 0& \mid & \frac{\partial P_{s1}}{\partial I_{\mathrm{sh}}^D}& \frac{\partial P_{s1}}{\partial I_{\mathrm{sh}}^Q}\\ 0& 0& 0& 0& 0& \mid & \frac{\partial Q_{s1}}{\partial I_{\mathrm{sh}}^D}& \frac{\partial Q_{s1}}{\partial I_{\mathrm{sh}}^Q}\\ \frac{\partial P_{\mathrm{sk}}}{\partial {\theta }_{s1}}& \frac{\partial P_{\mathrm{sk}}}{\partial {\theta }_{\mathrm{sk}}}& \frac{\partial P_{\mathrm{sk}}}{\partial V_{\mathrm{sk}}}& \frac{\partial P_{\mathrm{sk}}}{\partial {\theta }_{\mathrm{rk}}}& \frac{\partial P_{\mathrm{sk}}}{\partial V_{\mathrm{rk}}}& \mid & 0& 0\\ \frac{\partial Q_{\mathrm{sk}}}{\partial {\theta }_{s1}}& \frac{\partial Q_{\mathrm{sk}}}{\partial {\theta }_{\mathrm{sk}}}& \frac{\partial Q_{\mathrm{sk}}}{\partial V_{\mathrm{sk}}}& \frac{\partial Q_{\mathrm{sk}}}{\partial {\theta }_{\mathrm{rk}}}& \frac{\partial Q_{\mathrm{sk}}}{\partial V_{\mathrm{rk}}}& \mid & 0& 0\\ 0& 0& 0& 0& 0& \mid & 0& 0\\ 0& 0& 0& 0& 0& \mid & 0& 0\\ -& -& -& -& -& ⊣& -& -\\ \frac{\partial P_{dc}}{\partial {\theta }_{s1}}& \frac{\partial P_{dc}}{\partial {\theta }_{\mathrm{sk}}}& \frac{\partial P_{dc}}{\partial V_{\mathrm{sk}}}& \frac{\partial P_{dc}}{\partial {\theta }_{\mathrm{rk}}}& \frac{\partial P_{dc}}{\partial V_{\mathrm{rk}}}& \mid & \frac{\partial P_{dc}}{\partial I_{\mathrm{sh}}^D}& \frac{\partial P_{dc}}{\partial I_{\mathrm{sh}}^Q}\end{array}\right]$

The upper left part of ΔJ_GUPFC adds to the corresponding position of the original Jacobian matrix J. The other parts of ΔJ_GUPFC augment the size of J. Since P_r2□Q_r2□P_rk and Q_rk are constants, the elements of ΔJ_GUPFC in the fifth and sixth rows are all zeros. Because the length of the mismatch vector and the state vector are both increased by one, the size of J′ is bigger than J by one row and one column.

After modifying the mismatch vector and Jacobian matrix, the iterative formula for updating the state vector becomes

x ⁽ⁱ⁺¹⁾ =x ⁽ⁱ⁾ −[J′(x⁽ⁱ⁾)]⁻¹f′(x⁽ⁱ⁾)  eq. (11)

When the state vector converges within a specified tolerance, the equivalent voltage of shunt VSC can be recovered from I_sh^D and I_sh^Q. Simple manipulations yield the d-q components of the equivalent voltage of the shunt VSC,

$\begin{array}{cc}\left[\begin{array}{c}{V}_{\mathrm{sh}}^D\\ V_{\mathrm{sh}}^Q\end{array}\right]=\left[\begin{array}{cc}{R}_{\mathrm{sh}}& -X_{\mathrm{sh}}\\ X_{\mathrm{sh}}& R_{\mathrm{sh}}\end{array}\right]\left[\begin{array}{c}{I}_{\mathrm{sh}}^D\\ I_{\mathrm{sh}}^Q\end{array}\right]+\left[\begin{array}{c}V_{s1}\\ 0\end{array}\right].& \mathrm{Eq}.\left(12\right)\end{array}$

The equivalent voltages of the series VSCs can be calculated explicitly by:

$\begin{array}{cc}\left[\begin{array}{c}{V}_{\mathrm{serk}}^D\\ V_{\mathrm{serk}}^Q\end{array}\right]=\left[\begin{array}{cc}{R}_{\mathrm{serk}}& -X_{\mathrm{serk}}\\ X_{\mathrm{serk}}& R_{\mathrm{serk}}\end{array}\right]\left[\begin{array}{c}{I}_{\mathrm{serk}}^D\\ I_{\mathrm{serk}}^Q\end{array}\right]+\left[\begin{array}{cc}{V}_{\mathrm{rk}}^D& -V_{\mathrm{sk}}^D\\ V_{\mathrm{rk}}^Q& -V_{\mathrm{sk}}^Q\end{array}\right],k=2,\Lambda ,n.& \mathrm{Eq}.\left(13\right)\end{array}$

Finally, the polar form of the equivalent voltage of the shunt VSC and the series VSC can be obtained by:

$\begin{array}{cc}V_{\mathrm{sh}}\angle {\theta }_{\mathrm{sh}}=\sqrt{V_{\mathrm{sh}}^{D^2}+V_{\mathrm{sh}}^{Q^2}}\angle \left({\tan }^{-1}\frac{V_{\mathrm{sh}}^Q}{V_{\mathrm{sh}}^D}+{\theta }_{s1}\right)& \mathrm{Eq}.\left(14\right)\\ V_{\mathrm{serk}}\angle {\theta }_{\mathrm{serk}}=\sqrt{V_{\mathrm{serk}}^{D^2}+V_{\mathrm{serk}}^{Q^2}}\angle \left({\tan }^{-1}\frac{V_{\mathrm{serk}}^Q}{V_{\mathrm{serk}}^D}+{\theta }_{s1}\right)& \mathrm{Eq}.\left(15\right)\end{array}$

Under the assumption of known generations and loads, the basic power flow solutions, including voltages of all buses in the power grid and the equivalent voltages of the shunt and series VSCs of the generalized power flow controller, can be find by using the disclosed method, and the detail power flow solutions, including the active and reactive power flows into each transmission line, reactive power output of each generator, can be determined by using the basic power flow solution together with the fundamental circuit theory. A summary of procedures to calculate the power flow solution of a power grid embedded with the generalized power flow controller is depicted in FIG. 6.

• • Step 301: set the initial value of the state vector, wherein the elements of the state vector, called state variables, comprising the voltage magnitudes of all buses excluding the bus connected to the shunt branch of the generalized power flow controller, the phase angles of all buses, the d-q components of the shunt branch current of the generalized power flow controller; the voltage magnitude of each bus initially sets to 1.0 p.u., the phase angle of each bus, the d and q components of the shunt branch current of the generalized power flow controller all initially set to 0.
  • Step 302: construct the mismatch vector, f of a power grid ignoring the generalized power flow controller.
  • Step 303: establish the corresponding Jacobian matrix J using the first order derivatives of the mismatch vector f obtained in step 302.
  • Step 304: perform a d-q decomposition on the voltage of the receiving-end of each series branch by eq. (1). The d-q decomposition uses the voltage of the bus connected to shunt branch of the generalized power flow control as a reference phasor. The d component, V_rk^D, is in phase with the reference phasor, whereas the q component, V_rk^Q, leads the reference phasor by 90 degree.
  • Step 305: use eq. (2) to calculate the equivalent load of the shunt branch of the generalized power flow controller.
  • Step 306: judge whether there exists a series VSC, if it exists, go to step 307, otherwise, go to step 308.
  • Step 307: calculate the equivalent loads at the sending-end and receiving-end of each series branch by using eq. (3) and eq. (4), from the 2^nd to n^th series branch, wherein n is the number of VSCs.
  • Step 308: use eq. (6) to calculate the total active power generated from VSCs, P_dc.
  • Step 309: modify the mismatch vector by using eq. (8).
  • Step 310: modify the Jacobian matrix by using eq. (10).
  • Step 311: substitute the modified mismatch vector and modified Jacobian matrix into eq. (11) to update the state vector.
  • Step 312: judge whether the state vector converges within specified tolerance. If it does not, go back to step 302. Otherwise, proceed to step 313.
  • Step 313: calculate the equivalent voltages of the VSCs by using use eq. (14) and eq. (15).
  • Step 314: calculate the power flow solution, which includes the voltage of each bus, the active and reactive power flow of each transmission line, the reactive power generated from each generator.

Simulating several test systems embedded with STATCOM, UPFC, GUPFC and the generalized power flow controller has been performed to validate the present invention. The descriptions of the test systems are as follows:

• • Case 1: IEEE 300-bus test system without installing FACTS controller, referred to as a base case. This case provides a comparison basis with other cases.
  • Case 2: Similar to Case 1, except that it has been installed with one additional GUPFC. The GUPFC has one shunt branch and three series branches. The shunt branch is in parallel with bus 37 to control its voltage magnitude. The series branches are in series with line 37-49 (the transmission line linking bus 37 and bus 49), line 37-89 and line 37-40, respectively, to control their active and reactive power flow.
  • Case 3: Similar to Case 2 except that it has been installed with one additional generalized power flow controller, it is referred to as GPFC. GPFC has one shunt branch and two series branches. The shunt branch is in parallel with bus 102 to control its voltage magnitude. The series branches are in series with line 102-104 and line 103-105, respectively, to control their active and reactive power flow.
  • Case 4: Similar to Case 3 except that it has been installed with one additional UPEC. The UPFC has one shunt branch and one series branch. The shunt branch is in parallel with bus 7 to control its voltage magnitude. The series branch is in series with line 7-131 to control its active and reactive power.
  • Case 5: Similar to Case 4 except that it has been installed with one additional STATCOM. The STATCOM has one shunt branch, and it is in parallel with bus 81 to control its voltage magnitude.

In the above test cases, assuming the coupling transformers have the same impedances as 0.01+j0.05 p.u. The allowable tolerance of Newton-Raphson algorithm is set to 10⁻¹². The initial values for the state variables are 1∠0° for the bus voltages, and 0 for I_sh^D and I_sh^Q. Table 1 shows the iteration numbers required for obtaining power flow solution in the different test systems. The simulation results showed that incorporating the steady-state model of the generalized power flow controller will not increase the iteration number for obtaining the power flow solution within the same allowable tolerance.

TABLE 1
Iteration numbers required for obtaining power
flow solution in the different test systems
	Case
	1	2	3	4	5
	Iteration	6	6	6	6	6
	numbers

FIG. 7 shows the convergence pattern of state variables of the GUPFC. As shown, the current components I_sh^Q and I_sh^D converge to their target values after six iterations, respectively.

Case 3 is designed to demonstrate a distinguishing feature of the present invention. Even through the sending-ends of shunt branch and series branches of the GPFC connect to different buses, the power flow solution converges as rapid as the base case does. FIG. 8 shows the power mismatches at the sending-end and receiving-end buses of the GPFC. After six iterations, the power mismatches are within a tight tolerance, and thus a precise power flow solution is obtained. Therefore, it is well demonstrated that the power flow calculation achieves a rapid convergence characteristic using the steady state model of the generalized power flow controller according to the present invention.

FIG. 9 shows a quadratic convergence pattern of the solution process of Case 4, in which the dotted line is a typical quadratic convergence pattern and the solid line is the convergence curve of Case 4. It reveals that the quadratic convergence characteristic is preserved after embedding one STATCOM, one UPFC, one GUPFC and one generalized power controller into a power grid.

According to the simulation results, the power flow solution of the test cases, installed with STATCOM, UPFC, GUPFC the generalized power flow controller, can converge as rapidly as the base case does. It concludes that incorporating the steady-state model of the generalized power flow controller will not degrade the convergence speed of Newton-Raphson algorithm.

Many changes and modifications in the above described embodiment of the invention can, of course, be carried out without departing from the scope thereof. Accordingly, to promote the progress in science and the useful arts, the invention is disclosed and is intended to be limited only by the scope of the appended claims.

Claims

1. A method of calculating power flow solution of a power grid that includes generalized power flow controllers, the power grid comprising a plurality of buses, a plurality of transmission lines, a plurality of generators, a plurality of loads, and at least one generalized power flow controller; the generalized power flow controller having one shunt branch, and a plurality of series branch;the shunt branch having a sending-end connecting in parallel with a bus of the power grid, and comprising a shunt voltage sourced converter and a shunt coupling transformer, being modeled by an equivalent shunt voltage source in series with an impedance; each of the series branches having a sending-end and a receiving end, connecting in series with a transmission line of the power grid, and comprising a series voltage sourced converter and a series coupling transformer, being modeled by an equivalent series voltage source in series with an impedance;the AC side of the shunt voltage sourced converter connecting to the power grid through the shunt coupling transformer, and AC side of each of the series voltage converter connecting to the power grid through the series coupling transformer; the DC sides of the shunt voltage sourced converter and the series voltage sourced converters jointly connecting to a DC capacitor to share the same DC bus;the power flow solution including the voltage of each bus, the active and reactive power flow of each transmission line, the reactive power generated from each generator and the equivalent voltages of the voltage sourced converters of each generalized power flow controller;the method characterized in that the variable related to the generalized power flow controller being expressed in d-q components, wherein the d component is in phase with a reference phasor, and the q component leads the reference phasor by 90 degree;

2. A method in accordance with claim 1, wherein the voltage magnitude of each bus initially sets to 1.0 p.u., the phase angle of each bus, the d and q components of the shunt branch current of the generalized power flow controller all initially set to 0.

3. A method in accordance with claim 1, wherein the d-q decomposition uses the voltage of the bus connected to the shunt branch of the generalized power flow control as a reference phasor.

4. A method in accordance with claim 1, wherein the sending-ends of the shunt branch and series branches connect to a common bus.

5. A method in accordance with claim 4, wherein the number of the voltage sourced converters is 2.

6. A method in accordance with claim 1, wherein the sending-ends of the shunt branch and series branches connect to different buses.