A Method for Suppressing Overcurrent in the Rotor of DFIG Based on Trajectory Sensitivity Analysis

Abstract

The present invention discloses a method for suppressing the overcurrent of a DFIG rotor based on the analysis of trajectory sensitivity, including: 1. Solving the reactive power setting value of the DFIG grid-side converter and the value range of the DFIG reactive power setting value; 2. Defining the reactive power distribution coefficient λ; 3. Establishing the trajectory sensitivity analytical formula of the DFIG reactive power setting value and the distribution coefficient λ; 4. Defining the sensitivity α of the distribution coefficient λ to the DFIG reactive power setting value; 5. Constructing a reactive power optimization model for suppressing the rotor overcurrent; 6. Obtaining the optimal value of the DFIG reactive power setting value and the optimal value of the distribution coefficient λ through the interior point method optimization algorithm. The present invention can effectively suppress the rotor overcurrent and improve the fault ride-through ability of the DFIG by constructing a reactive power optimization model for suppressing the rotor overcurrent to obtain the optimized parameters.

Description

Technical Field

[0001] The invention relates to the field of DFIG reactive power adjustment and distribution, and in particular to a reactive power optimization method for suppressing DFIG rotor overcurrent based on analytical trajectory sensitivity. Background Art

[0002] Doubly fed induction generator (DFIG) is one of the mainstream models currently used in wind power generation. It has the advantages of small converter capacity, active and reactive power decoupling control, and wide speed regulation range. However, since the DFIG stator is directly connected to the power grid, it is easily affected by power grid faults. In addition, the converter capacity is small, only about 30% of the rated capacity, so the rotor current is prone to generate impact current under fault ride-through, causing the unit to be disconnected from the grid, further deteriorating the safe operation of the power system.

[0003] At present, the methods for suppressing rotor overcurrent at home and abroad can be mainly divided into protection and control. The protection type generally adopts the addition of auxiliary equipment to release excess transient energy. In engineering, a crowbar protection circuit is usually connected in parallel on the rotor side. When the grid fault excites the rotor current to reach the crowbar protection setting value, the crowbar protection is activated and the rotor side converter control pulse is blocked at the same time. After the fault is over, the unit returns to normal working state. However, when the rotor side converter is blocked, the DFIG operates as an induction motor and needs to absorb a large amount of reactive power from the grid for excitation, which is not conducive to the recovery of the grid voltage, and the additional protection circuit will also increase the system cost. The control type usually adds feedforward compensation and adopts demagnetization control and its derivative solutions. Feedforward compensation basically does not change the traditional vector control structure, and only adds voltage or current compensation items in the forward channel, such as adding voltage dynamic compensation in the calculation of the rotor voltage reference instruction. However, this method fails to achieve the optimal suppression of the rotor current peak, and the control method is complex.

[0004] After a fault occurs, the rotor overcurrent can be suppressed by adjusting the reactive power output of the DFIG, but the response speed is slow and the suppression effect is poor. For parameter optimization, the traditional intelligent algorithm has a large amount of calculation and is prone to local optimization. The optimization based on trajectory sensitivity often uses the perturbation method, but each optimization perturbation method requires two time domain simulations, which is computationally intensive and can only obtain the trajectory sensitivity of one parameter each time. Therefore, in order to suppress the rotor overcurrent, it is necessary to design a simple reactive optimization method based on anticipated faults to suppress the DFIG rotor overcurrent. Summary of the invention

[0005] The present invention aims to solve the deficiencies of the above-mentioned prior art and provide a DFIG rotor overcurrent suppression method based on trajectory sensitivity analysis, so as to suppress the DFIG rotor overcurrent under fault ride-through, thereby improving the DFIG fault ride-through capability.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A method for suppressing overcurrent in the rotor of a DFIG based on trajectory sensitivity analysis according to the present invention is characterized by including the following steps:

[0008] S1. Solve the reactive power setpoint of the grid-side converter of the DFIG and the value range of the reactive power setpoint of the DFIG ;

[0009] S2. Define the distribution coefficient λ of the reactive power setpoint of the grid-side converter of the DFIG and the reactive power setpoint of the DFIG , solve the upper and lower limits of the distribution coefficient λ, and obtain the initial value of the steady-state trajectory sensitivity for the reactive power setpoint of the DFIG and the distribution coefficient λ;

[0010] S3. Establish an analytical formula for the trajectory sensitivity of the peak rotor current under fault ride-through to the reactive power setpoint of the DFIG and the distribution coefficient λ;

[0011] S4. Define the sensitivity α of the distribution coefficient λ to the reactive power setpoint of the DFIG at the peak rotor current;

[0012] S5. Taking the peak rotor current as the target and considering the reactive power output constraint, thus constructing a reactive power optimization model for suppressing rotor overcurrent;

[0013] S6. Solve the reactive power optimization model for suppressing rotor overcurrent by the interior point method, and adjust the distribution coefficient λ through the reactive power setpoint of the DFIG and the sensitivity α, so as to obtain the optimal value of the reactive power setpoint of the DFIG and the optimal value of the distribution coefficient λ.

[0014] The method for suppressing overcurrent in the rotor of a DFIG based on trajectory sensitivity analysis according to the present invention is also characterized in that the S1 includes:

[0015] S1.1. According to the capacity limit of the grid-side converter of the DFIG, use Equation (1) to calculate the output range of the reactive power setpoint of the grid-side converter :

[0016]

[0017] In Equation (1), S GSC is the capacity of the grid-side converter of the DFIG, P DFIGis the active power output of the DFIG, s is the rotor slip, and Q g is the reactive power flowing from the grid-side converter to the stator node. When the DFIG operates in a steady state, let

[0018] S1.2. According to the stator and rotor current constraints under steady state, use Equations (2) and (3) to obtain the range of the reactive power output Q sm of the stator node:

[0019]

[0020]

[0021] In Equations (2) and (3), r s is the radius of the reactive power output range on the rotor side, and r r is the radius of the reactive power output range on the stator side. R s and X s are the stator resistance and reactance respectively, and X m is the excitation reactance, V s is the stator voltage, I smax and I rmax are the maximum values of the stator and rotor currents under steady state respectively.

[0022] S1.3. According to the capacity limit of the DFIG grid-side converter and the stator and rotor current constraints, use Equation (4) to obtain the range of the DFIG reactive power setting value :

[0023]

[0024] In Equation (4), represents the upper limit of the DFIG reactive power setting value operating range, and represents the lower limit of the DFIG reactive power operating range.

[0025] The said S2 is carried out according to the following steps:

[0026] S2.1. Define the distribution coefficient λ of the reactive power setting value of the DFIG grid-side converter and the reactive power setting value of the DFIG using Equation (5), and determine the upper and lower limits of the distribution coefficient λ using Equation (6):

[0027]

[0028]

[0029] In Equation (6), λ min is the lower limit of the distribution coefficient λ, and λ maxis the upper limit of the distribution coefficient λ;

[0030] S2.2. Set the DFIG reactive power setpoint and the distribution coefficient λ as parameter a, and use Equation (7) to describe the DFIG dynamic model:

[0031]

[0032] In Equation (7), f is the differential equation, g is the algebraic equation, x is the state variable, y is the algebraic variable, represents the derivative of the state variable x with respect to time;

[0033] S2.3. Differentiate Equation (7) with respect to parameter a, and thus use Equation (8) to obtain the partial derivative equation of the DFIG dynamic model with respect to parameter a:

[0034]

[0035] In Equation (8), x a , y a are the partial derivatives of the state variable x and the algebraic variable y with respect to parameter a, respectively, represents the partial derivative of the derivative of the state variable x with respect to time with respect to parameter a, f x , f y , f a are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a, respectively, g x , g y , g a are the partial derivatives of the algebraic equation g with respect to the state variable x, the algebraic variable y, and the parameter a, respectively;

[0036] S2.4. Set the DFIG reactive power setpoint and the distribution coefficient λ as parameter a, and use Equation (9) to obtain the initial values x a0 and y a0 of the steady-state trajectory sensitivity of the state variable x and the algebraic variable y with respect to parameter a:

[0037]

[0038] The above S3 is carried out according to the following steps:

[0039] S3.1. Use Equation (10) to obtain the trajectory sensitivities of the state variables x and y at the (n + 1)-th moment with respect to parameter a

[0040]

[0041] In Equation (10), I is the identity matrix, The partial derivatives of the state variable x and the algebraic variable y with respect to the parameter a at time n, respectively. When n = 0, let The partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at time n, respectively; The partial derivatives of the state variable x and the algebraic variable y with respect to the parameter a at time n + 1, respectively, The partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at time n, respectively, The partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at time n + 1, respectively; Is the remainder matrix;

[0042] Using Equation (11) and Equation (12), the reactive power setpoint of the DFIG is obtained respectively And the non-zero elements of the distribution coefficient λ in the remainder matrix are:

[0043]

[0044]

[0045] In Equation (11) and Equation (12), e D1 And e D2 Are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the outer loop of the reactive power control of the rotor side converter of the DFIG with respect to the reactive power setpoint of the DFIG Of, e D3 And e D4 Are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the outer loop of the reactive power control of the grid side converter of the DFIG with respect to the reactive power setpoint of the DFIG Of, e λ1 And e λ2 Are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the outer loop of the reactive power control of the rotor side converter of the DFIG with respect to the distribution coefficient λ, e λ3 And e λ4 Are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the outer loop of the reactive power control of the grid side converter of the DFIG with respect to the distribution coefficient λ, k pr2 And k ir2 Are the proportional coefficient and the integral coefficient of the outer loop of the reactive power control of the rotor side converter of the DFIG, k pg2 And k ig2 Are the proportional coefficient and the integral coefficient of the outer loop of the reactive power control of the grid side converter of the DFIG respectively;

[0046] S3.2. If there is a peak rotor current I r,peak At time m, then use Equation (13) to calculate the trajectory sensitivity of the peak rotor current I r,peak To the parameter a:

[0047]

[0048] In Equation (13), represents the d-axis component of the rotor current at time m, represents the q-axis component of the rotor current at time m.

[0049] In step S4, the sensitivity α of the DFIG reactive power setpoint is defined by using Equation (14) for the distribution coefficient λ:

[0050]

[0051] The reactive power optimization model for suppressing rotor overcurrent in step S5 is constructed with Equation (15) as the objective function J and Equations (4) and (6) as the constraints;

[0052] minJ = |I r,peak | (15).

[0053] Step S6 is carried out as follows:

[0054] S6.1. Use Equations (16) and (17) to obtain the distribution coefficient λ (k+1) and its adjustment value Δλ (k+1) at the (k + 1)-th optimization iteration:

[0055]

[0056]

[0057] In Equations (16) and (17), k represents the current optimization iteration number, and α (k+1) is the sensitivity of the distribution coefficient λ obtained at the (k + 1)-th optimization iteration to the DFIG reactive power setpoint , is the DFIG reactive power setpoint obtained at the (k + 1)-th optimization iteration, is the correction amount of the DFIG reactive power setpoint at the (k + 1)-th optimization iteration, λ (k) is the distribution coefficient obtained at the k-th optimization iteration, λ (k+1) is the distribution coefficient obtained at the (k + 1)-th optimization iteration, and Δλ (k+1) is the correction amount of the distribution coefficient λ at the (k + 1)-th optimization iteration; when k = 0, let λ (k) = λ, and let

[0058] S6.2. Calculate the dual gap Gap at the (k + 1)-th optimization iteration from the slack variable and the Lagrange multiplier (k +1); Determine whether the formula (18) holds. If it holds, use the formula (19) to output the optimal value of the DFIG reactive power setting value and the optimal value of the distribution coefficient λ opt , otherwise, assign k + 1 to k, and then use the formula (20) to re-initialize the DFIG reactive power setting value and the distribution coefficient λ, and then jump to S2;

[0059] Gap (k+1) <ε (18)

[0060]

[0061]

[0062] In the formula (18), ε represents the set convergence accuracy.

[0063] An electronic device of the present invention includes a memory and a processor, characterized in that the memory is used to store a program for supporting the processor to execute the DFIG rotor overcurrent suppression method, and the processor is configured to execute the program stored in the memory.

[0064] A computer-readable storage medium of the present invention stores a computer program on the computer-readable storage medium, characterized in that the computer program executes the steps of the DFIG rotor overcurrent suppression method when run by a processor.

[0065] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0066] 1. The method proposed by the present invention considers the output range of the DFIG grid-side reactive power setting value and the DFIG reactive power setting value , and satisfies the stator and rotor current and converter capacity constraints under steady state during the calculation of the optimal parameters, so that the obtained optimal parameters are within the steady-state safe operation range.

[0067] 2. The present invention takes the absolute value of the rotor current peak I during fault ride-through as the objective function, considers the DFIG reactive power output constraint, and constructs a reactive power optimization model for suppressing rotor overcurrent, which can quickly obtain the optimal value of the DFIG reactive power setting value r,peak , is applicable to various fault conditions, and thus can suppress rotor overcurrent under different fault conditions.

[0068] 3. The present invention obtains the optimal values of the DFIG reactive power setting value and the distribution coefficient λ for suppressing rotor overcurrent during fault ride-through, does not affect the operation of the DFIG under steady state, and does not add additional equipment, thus reducing the economic cost.

[0069] 4. The present invention calculates the initial value of the steady-state trajectory sensitivity for the DFIG reactive power set value and the distribution coefficient λ, and proposes an analytical formula for the trajectory sensitivity of the rotor current to the DFIG reactive power set value and the distribution coefficient λ under fault ride-through, and calculates the peak value I of the rotor current r,peak for the DFIG reactive power set value and the trajectory sensitivity of the distribution coefficient λ, which greatly reduces the sensitivity calculation amount compared with the perturbation method. And at the peak value I of the rotor current r,peak define the sensitivity α of the distribution coefficient λ to the DFIG reactive power set value During the interior point method iteration process, adjust the distribution coefficient λ by the DFIG reactive power set value and the sensitivity α, thereby reducing the optimization calculation amount, greatly reducing the time required for optimization, and improving the calculation efficiency. Description of the Drawings

[0070] Figure 1 is the structure diagram of the DFIG in the prior art;

[0071] Figure 2 is the double-loop PI control diagram of the DFIG converter in the prior art.

[0072] Figure 3 is the flow chart of the reactive power optimization for suppressing the DFIG rotor overcurrent in the present invention. Detailed Embodiment

[0073] In this embodiment, as Figure 1 shown, the DFIG is composed of four parts: a wind turbine system, a drive system, an induction motor, and a converter. The wind turbine system converts wind energy into mechanical energy through the rotation of the wind turbine blades, and the captured power of the wind turbine can be obtained. The drive system corresponds to the mechanical transient equation and adopts a two-mass model, where the wind turbine and the low-speed shaft are taken as one mass, and the gearbox and the high-speed shaft are taken as another mass. The induction motor adopts the motor case, with the direction of the DFIG stator current flowing into the motor as positive, and the stator and rotor flux linkages, voltages, and currents are decomposed into d-q components. The converter consists of three parts: a rotor side converter (RSC), a DC capacitor, and a grid side converter (GSC). The converter adopts Figure 2 the double-loop PI control of the power outer loop and the current inner loop shown.

[0074] For Figure 1The DFIG shown is connected to a single-machine infinite bus system. A method for suppressing the overcurrent of the DFIG rotor based on the analysis of trajectory sensitivity is proposed in this example for the three-phase short-circuit problem at the point of common coupling. When a three-phase short-circuit occurs at the point of common coupling, the reactive power optimization process for suppressing the overcurrent of the rotor is as follows Figure 3 shown, which specifically includes the following steps:

[0075] S1. Solve the reactive power setting value of the grid-side converter of the DFIG and the range of the reactive power setting value of the DFIG ;

[0076] S1.1. According to the capacity limit of the grid-side converter of the DFIG, use Equation (1) to calculate the output range of the reactive power setting value of the grid-side converter:

[0077]

[0078] In Equation (1), S GSC is the capacity of the grid-side converter of the DFIG, P DFIG is the active power output of the DFIG, s is the rotor slip, Q g is the reactive power flowing from the grid-side converter to the stator node. When the DFIG is operating in a steady state, let

[0079] S1.2. According to the stator and rotor current constraints under steady state, use Equation (2) and Equation (3) to obtain the range of the reactive power output Q sm of the stator node:

[0080]

[0081]

[0082] In Equation (2) and Equation (3), r s is the radius of the reactive power output range on the rotor side, r r is the radius of the reactive power output range on the stator side, R s and X s are the stator resistance and reactance respectively, X m is the magnetizing reactance, V s is the stator voltage, I smax and I rmax are the maximum values of the stator and rotor currents under steady state respectively.

[0083] S1.3. According to the capacity limit of the grid-side converter of the DFIG and the stator and rotor current constraints, use Equation (4) to obtain the range of the reactive power setting value of the DFIG:

[0084]

[0085] In Equation (4), represents the upper limit of the operating range of the DFIG reactive power setpoint, represents the lower limit of the operating range of the DFIG reactive power.

[0086] S2. Define the reactive power setpoint of the DFIG grid-side converter and the distribution coefficient λ of the DFIG reactive power setpoint . Solve the upper and lower limits of the distribution coefficient λ, and obtain the initial value of the steady-state trajectory sensitivity for the DFIG reactive power setpoint and the distribution coefficient λ;

[0087] S2.1. To reflect the proportion of grid-side reactive power in the DFIG reactive power output, use Equation (5) to define the reactive power setpoint of the DFIG grid-side converter and the distribution coefficient λ of the DFIG reactive power setpoint . At a certain determined DFIG reactive power output, determine the upper and lower limits of the distribution coefficient λ through the reactive power setpoint of the DFIG grid-side converter using Equation (6):

[0088]

[0089]

[0090] In Equation (6), λ min is the lower limit of the distribution coefficient λ, and λ max is the upper limit of the distribution coefficient λ.

[0091] S2.2. Take the DFIG reactive power setpoint and the distribution coefficient λ as parameters a, and use Equation (7) to describe the DFIG dynamic model:

[0092]

[0093] In Equation (7), f is a differential equation, g is an algebraic equation, x is a state variable, y is an algebraic variable, represents the derivative of the state variable x with respect to time.

[0094] S2.3. Differentiate Equation (7) with respect to parameter a, and use Equation (8) to obtain the partial derivative equation of the DFIG dynamic model with respect to parameter a:

[0095]

[0096] In Equation (8), x a , y a are the partial derivatives of the state variable x and the algebraic variable y with respect to parameter a, respectively, Denotes the partial derivative of the derivative of the state variable x with respect to time with respect to the parameter a, f x 、f y 、f a Are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a respectively, g x 、g y 、g a Are the partial derivatives of the algebraic equation g with respect to the state variable x, the algebraic variable y, and the parameter a respectively.

[0097] S2.4. Take the DFIG reactive power set value and the distribution coefficient λ as the parameter a, and use Equation (9) to obtain the initial values x a0 and y a0 of the steady-state trajectory sensitivity of the state variable x and the algebraic variable y with respect to the parameter a:

[0098]

[0099] S3. Establish an analytical formula for the trajectory sensitivity of the peak value of the rotor current during fault ride-through with respect to the DFIG reactive power set value and the distribution coefficient λ;

[0100] S3.1. According to Equation (8), use Equation (10) to obtain the trajectory sensitivities of the state variables x and y with respect to the parameter a at the n + 1th moment

[0101]

[0102] In Equation (10), I is the identity matrix, Are the partial derivatives of the state variable x and the algebraic variable y with respect to the parameter a at the nth moment respectively. When n = 0, let Are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at the nth moment respectively. Are the partial derivatives of the state variable x and the algebraic variable y with respect to the parameter a at the n + 1th moment respectively, Are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at the nth moment respectively, Are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at the n + 1th moment respectively; Is the remainder matrix;

[0103] Use Equations (11) and (12) to obtain the non-zero elements of the DFIG reactive power set value and the distribution coefficient λ in the remainder matrix as:

[0104]

[0105]

[0106] In equations (11) and (12), e D1 and e D2 are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the outer reactive power control loop of the DFIG rotor-side converter with respect to the DFIG reactive power set value respectively, e D3 and e D4 are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the outer reactive power control loop of the DFIG grid-side converter with respect to the DFIG reactive power set value respectively, e λ1 and e λ2 are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the outer reactive power control loop of the DFIG rotor-side converter with respect to the distribution coefficient λ respectively, e λ3 and e λ4 are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the outer reactive power control loop of the DFIG grid-side converter with respect to the distribution coefficient λ respectively, k pr2 and k ir2 are the proportional coefficient and the integral coefficient of the outer reactive power control loop of the DFIG rotor-side converter respectively, k pg2 and k ig2 are the proportional coefficient and the integral coefficient of the outer reactive power control loop of the DFIG grid-side converter respectively.

[0107] S3.2. If there is a peak rotor current I r,peak at time m, then use equation (13) to calculate the trajectory sensitivity of the peak rotor current I r,peak to the parameter a:

[0108]

[0109] In equation (13), represents the d-axis component of the rotor current at time m, represents the q-axis component of the rotor current at time m.

[0110] S4. Define the sensitivity α of the distribution coefficient λ to the DFIG reactive power set value using equation (14):

[0111]

[0112] S5. Considering the DFIG reactive power set value and the distribution coefficient λ constraint, the reactive power optimization model for suppressing rotor overcurrent is constructed with equation (15) as the objective function J and equations (4) and (6) as the constraints;

[0113] minJ = |I r,peak | (15)

[0114] S6. Solve the reactive power optimization model for suppressing rotor overcurrent by the interior point method, and adjust the distribution coefficient λ through the reactive power setpoint of the DFIG and the sensitivity α, so as to obtain the optimal value of the reactive power setpoint of the DFIG and the optimal value of the distribution coefficient λ.

[0115] S6.1. Let k represent the optimization iteration number, set the lower and upper Lagrange multipliers z and w to 1, and initialize the lower slack variable l and the upper slack variable u using Equation (16) when k = 0:

[0116]

[0117] In Equation (16), l (k) , u (k) are the lower and upper slack variables in the k-th optimization iteration respectively, is the reactive power setpoint of the DFIG obtained in the k-th optimization iteration.

[0118] S6.2. Set the value of the central parameter σ to 0.1, and calculate the duality gap Gap (k) and the perturbation factor μ (k) in the k-th optimization iteration using Equation (17):

[0119]

[0120] In Equation (17), z (k) , w (k) are the lower and upper Lagrange multipliers in the k-th optimization iteration respectively.

[0121] S6.3. The second-order trajectory sensitivity is more computationally intensive than the first-order trajectory sensitivity. Therefore, the BFGS algorithm is used to approximately calculate the second-order trajectory sensitivity G (k) in the k-th optimization iteration using Equation (18):

[0122]

[0123] In Equation (18), G (k-1) is the second-order trajectory sensitivity in the (k - 1)-th optimization iteration, is the reactive power setpoint of the DFIG obtained in the (k - 1)-th optimization iteration, b (k) , c (k) , d (k) are the first, second, and third types of intermediate variables in the k-th optimization iteration respectively, and b (k-1) is the second type of intermediate variable in the (k - 1)-th optimization iteration.

[0124] S6.4. Use Equation (19) to approximately calculate the second-order approximate unbalanced equation:

[0125]

[0126] S6.5. Use Equation (20) to calculate the correction amount of the DFIG reactive power setting value during the k-th optimization iteration:

[0127]

[0128] S6.6. Use Equation (21) to calculate the new values of each parameter during the k-th optimization iteration:

[0129]

[0130] In Equation (21), l (k+1) and u (k+1) are the lower relaxation variable and the upper relaxation variable during the (k + 1)-th optimization iteration respectively, z (k +1) and w (k+1) are the lower Lagrange multiplier and the upper Lagrange multiplier during the (k + 1)-th optimization iteration respectively, is the DFIG reactive power setting value obtained during the (k + 1)-th optimization iteration.

[0131] S6.7. To reduce the optimization calculation amount, during the interior point method iteration process, use Equations (16) and (17) to adjust λ from the DFIG reactive power setting value and α, and use Equations (22) and (23) to obtain the distribution coefficient λ (k+1) and its adjustment value Δλ (k+1) during the (k + 1)-th optimization iteration respectively:

[0132]

[0133]

[0134] In Equations (22) and (23), α (k+1) is the sensitivity of the distribution coefficient λ obtained during the (k + 1)-th optimization iteration to the DFIG reactive power setting value , is the DFIG reactive power setting value obtained during the (k + 1)-th optimization iteration, is the correction amount of the DFIG reactive power setting value during the (k + 1)-th optimization iteration, λ (k) is the distribution coefficient obtained during the k-th optimization iteration, λ (k+1) is the distribution coefficient obtained during the (k + 1)-th optimization iteration, Δλ (k+1) is the correction amount of the distribution coefficient λ during the (k + 1)-th optimization iteration. When k = 0, λ (k) = λ,

[0135] S6.8. Determine whether to converge according to the set convergence accuracy ε. If it converges, use Equation (24) to output the optimal value of the DFIG reactive power setting value and the optimal value λ of the distribution coefficient opt , otherwise, assign k + 1 to k, and then use Equation (26) to re-initialize the DFIG reactive power setting value and the distribution coefficient λ, and jump to S2.

[0136] Gap (k+1) <ε (24)

[0137]

[0138]

[0139] In this embodiment, an electronic device includes a memory and a processor. The memory is used to store a program that supports the processor to execute the above DFIG rotor overcurrent suppression method, and the processor is configured to execute the program stored in the memory.

[0140] In this embodiment, a computer-readable storage medium stores a computer program, and when the computer program is run by a processor, it executes the steps of the above DFIG rotor overcurrent suppression method.

Claims

1. A method for suppressing overcurrent in the rotor of a DFIG based on trajectory sensitivity analysis, characterized in that It includes the following steps: S1. Solve the reactive power set value of the DFIG grid-side converter and the value range of the DFIG reactive power set value ; S2. Define the reactive power setpoint of the DFIG grid-side converter and the distribution coefficient λ of the DFIG reactive power setpoint . Solve the upper and lower limits of the distribution coefficient λ, and obtain the initial value of the steady-state trajectory sensitivity for the DFIG reactive power setpoint and the distribution coefficient λ S3. Establish the analytical formula of the trajectory sensitivity of the peak value of the rotor current during fault ride-through to the reactive power setting value of the DFIG and the distribution coefficient λ; S4. Define the sensitivity α of the DFIG reactive power set value using the distribution coefficient λ at the peak rotor current according to Equation (14). ; (14) S5. Taking the peak value of the rotor current as the target and considering the reactive power output constraint, a reactive power optimization model for suppressing rotor overcurrent with Equation (15) as the objective function is constructed. The reactive power optimization model for suppressing rotor overcurrent; (15) In Equation (15), represents the peak value of the rotor current at time m; S6. Solve the reactive power optimization model for suppressing rotor overcurrent by the interior point method, and adjust the distribution coefficient λ through the DFIG reactive power set value and the sensitivity α, so as to obtain the optimal value of the DFIG reactive power set value and the optimal value of the distribution coefficient λ.

2. The method for suppressing DFIG rotor overcurrent based on trajectory sensitivity analysis according to claim 1, wherein The S1 includes: S1.

1. Calculate the set value of the reactive power of the grid-side converter according to the capacity limit of the DFIG grid-side converter using Equation (1). Output range: (1) In formula (1), S GSC is the capacity of the DFIG grid-side converter, P DFIG is the active power output of the DFIG, s is the rotor slip, Q g is the reactive power flowing from the grid-side converter to the stator node. When the DFIG operates in a steady state, let ; S1.

2. Obtain the range of the reactive power output of the stator nodes according to the stator and rotor current constraints under steady state by using Equations (2) and (3): ​ (2) (3) In Formula (2) and Formula (3), is the radius of the reactive power output range on the rotor side, is the radius of the reactive power output range on the stator side, R s and X s are the stator resistance and reactance respectively, X m is the magnetizing reactance, V s is the stator voltage, I smax and I rmax are the maximum values of the stator and rotor currents under steady state respectively; S1.

3. Obtain the range of the DFIG reactive power setting value using Equation (4) according to the capacity limit of the DFIG grid-side converter and the stator and rotor current constraints as follows: (4) In formula (4), represents the upper limit of the operating range of the DFIG reactive power setpoint, represents the lower limit of the operating range of the DFIG reactive power.

3. The DFIG rotor overcurrent suppression method based on trajectory sensitivity analysis according to claim 2, characterized in that The S2 is carried out according to the following steps: S2.

1. Define the reactive power setpoint of the DFIG grid-side converter using Equation (5). And the distribution coefficient λ of the DFIG reactive power setpoint And determine the upper and lower limits of the distribution coefficient λ using Equation (6): (5) (6) In formula (6), is the lower limit of the distribution coefficient λ, is the upper limit of the distribution coefficient λ; S2.

2. Set the DFIG reactive power setpoint and the distribution coefficient λ as parameter a, and use Equation (7) to describe the DFIG dynamic model: (7) In Equation (7), f is a differential equation, g is an algebraic equation, x is a state variable, and y is an algebraic variable, represents the derivative of the state variable x with respect to time; S2.

3. Derive the parameter a according to Equation (7), so as to obtain the partial derivative equation of the DFIG dynamic model with respect to the parameter a by using Equation (8): (8) In Equation (8), and are the partial derivatives of the state variable x and the algebraic variable y with respect to the parameter a, respectively. represents the partial derivative of the derivative of the state variable x with respect to time with respect to the parameter a. and and are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a, respectively. and and are the partial derivatives of the algebraic equation g with respect to the state variable x, the algebraic variable y, and the parameter a, respectively. S2.

4. Set the DFIG reactive power setpoint and the distribution coefficient λ as parameter a, and use Equation (9) to obtain the initial values of the steady-state trajectory sensitivities of the state variables x and the algebraic variables y with respect to parameter a and : (9)。 4. The DFIG rotor overcurrent suppression method based on trajectory sensitivity analysis according to claim 3, characterized in that The S3 is carried out according to the following steps: S3.

1. Obtain the trajectory sensitivities of the state variables \(x\) and \(y\) at the \((n + 1)\)-th moment with respect to the parameter \(a\) using Equation (10). and : (10) In Equation (10), I is the identity matrix, and are the partial derivatives of the state variable x and the algebraic variable y with respect to the parameter a at time n. When n = 0, let = and = ; and and are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at time n respectively; and are the partial derivatives of the state variable x and the algebraic variable y with respect to the parameter a at time n + 1 respectively, and and are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at time n respectively, and and are the partial derivatives of the differential equation f with respect to the state variable x, the algebraic variable y, and the parameter a at time n + 1 respectively; is the remainder matrix; The DFIG reactive power setpoint is obtained by using Equation (11) and Equation (12) respectively and the non-zero elements of the distribution coefficient λ in the remainder matrix are as follows: (11) (12) In equations (11) and (12), and are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the reactive power control outer loop of the DFIG rotor-side converter with respect to the DFIG reactive power set value , and are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the reactive power control outer loop of the DFIG grid-side converter with respect to the DFIG reactive power set value , and are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the reactive power control outer loop of the DFIG rotor-side converter with respect to the distribution coefficient λ, and are the partial derivatives of the proportional coefficient equation and the integral coefficient equation of the reactive power control outer loop of the DFIG grid-side converter with respect to the distribution coefficient λ, and are the proportional coefficient and the integral coefficient of the reactive power control outer loop of the DFIG rotor-side converter respectively, and are the proportional coefficient and the integral coefficient of the reactive power control outer loop of the DFIG grid-side converter respectively; S3.

2. If there is a peak rotor current at time m , then calculate the trajectory sensitivity of the peak rotor current to parameter a using Equation (13): ​ (13) In formula (13), represents the d-axis component of the rotor current at time m, represents the q-axis component of the rotor current at time m.

5. The method for suppressing the overcurrent of the DFIG rotor based on the analysis of the trajectory sensitivity according to claim 4, wherein S6.

1. Obtain the distribution coefficient and its adjustment value at the (k + 1)-th optimization iteration using equations (16) and (17) respectively: ​​ (22) (23) In Equations (16) and (17), k represents the current optimization iteration number, is the sensitivity of the distribution coefficient λ obtained in the (k + 1)-th optimization iteration to the DFIG reactive power set value ; is the DFIG reactive power set value obtained in the (k + 1)-th optimization iteration, is the correction amount of the DFIG reactive power set value in the (k + 1)-th optimization iteration, is the distribution coefficient obtained in the k-th optimization iteration, is the distribution coefficient obtained in the (k + 1)-th optimization iteration, is the correction amount of the distribution coefficient λ in the (k + 1)-th optimization iteration; When k = 0, let , and let ; S6.

2. Calculate the dual gap at the (k + 1)-th optimization iteration from the slack variables and Lagrange multipliers ; Use equation (18) to determine whether it holds. If it holds, use equation (19) to output the optimal value of the DFIG reactive power setpoint and the optimal value of the distribution coefficient , otherwise, assign k + 1 to k, and then use equation (20) to re-initialize the DFIG reactive power setpoint and the distribution coefficient λ, and then jump to S2; (18) (19) (20) In Equation (18), ε represents the set convergence accuracy.

6. An electronic device, comprising a memory and a processor, characterized in that, The memory is used to store a program for supporting the processor to execute any one of the DFIG rotor overcurrent suppression methods described in claims 1-5, and the processor is configured to execute the program stored in the memory.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it executes the steps of any one of the DFIG rotor overcurrent suppression methods described in claims 1-5.

Images