Optimization method for improving stability and economical efficiency of wind-solar-thermal bundled power system by applying optimal dynamic convergence method

Abstract

The invention provides an optimization method for improving the stability and economy of a wind-light-fire-bundled power system by using an optimal dynamic convergence method, and the method comprises the steps: connecting various generators to a bus of a multi-machine power system through a transformer, and constructing the wind-light-fire-bundled power system; a power system stabilizer, a unified power flow controller and a static synchronous compensator of an additional power oscillation damper are connected into the wind-light-fire bundled power system; the method comprises the following steps: establishing a target function based on a damping ratio, voltage and active power according to the bus voltage of a controller and the active power of a generator in a wind-solar-thermal bundled power system; and performing coordinated optimization on the parameters of the wind, light and fire bundled power system controller by using an optimal dynamic convergence method based on an Euclidean distance weighting strategy, an inertia weight dynamic adjustment strategy, a gradient position updating strategy and an adaptive step length updating strategy to obtain optimized controller parameters. According to the method, low-frequency oscillation in the power system is effectively improved, active power loss generated by system operation is reduced, and the overall stability and economical efficiency of the power system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power system controller parameter optimization, and in particular to an optimization method for improving the stability and economy of a wind-solar-thermal bundled power system. Background Art

[0002] As the global energy transition process accelerates, the proportion of renewable energy in the power system continues to increase. Wind power and photovoltaics, as the two most representative renewable energy sources, have developed rapidly due to their clean and low-carbon characteristics. However, wind power and photovoltaic power generation are intermittent and volatile. Their large-scale grid connection is prone to cause low-frequency oscillations in the power system. If not controlled, it will greatly threaten the safe and stable operation of the power system. The wind-solar-thermal bundling system can effectively suppress the negative impact of wind power and photovoltaics. By combining with traditional thermal power, it can effectively improve the stability and economy of the power system, and is also of great significance to the sustainable development of global energy.

[0003] Installing a power system stabilizer (PSS) on a synchronous generator can effectively suppress low-frequency oscillations in the power grid. However, due to the intermittent and volatile nature of wind and photovoltaic power generation, it is difficult to effectively improve the small-disturbance stability of the power system using only a PSS. Static synchronous compensator (STATCOM), unified power flow controller (UPFC) and additional damping controller (POD) are important components of flexible AC transmission system (FACTS) devices. By connecting POD in parallel in the transmission line and measuring the active and reactive power signals of the line, the system's interval oscillations can be effectively suppressed. STATCOM and UPFC can quickly respond to dynamic changes in the system, optimize power flow distribution and voltage control, improve the stability of the system bus voltage and generator active power, and improve the stability of the system.

[0004] When there are multiple controllers in a power system, there will be interactions between the controllers. If the controller parameters are not set properly, it will have a negative impact on the system. Therefore, coordinating and optimizing the parameters of multiple controllers to improve the negative interactions between multiple controllers has become one of the hot issues in power systems. Traditional methods have some shortcomings. They are difficult to deal with nonlinear multi-extreme value optimization problems, easily fall into local optimality, and cannot adapt to low-frequency oscillation problems under various operating modes. Summary of the invention

[0005] In response to the low-frequency oscillation problem existing in wind-solar-thermal bundled power systems, existing methods face the technical problem of poor stability in the selection and adjustment of controller parameters in high-dimensional nonlinear systems. The present invention proposes an optimization method that uses the optimal dynamic convergence method to improve the stability and economy of wind-solar-thermal bundled power systems. The optimization problem is solved based on multiple adaptive strategies such as distance mean strategy, inertia weight, and learning factor, which solves the low-frequency oscillation problem in the power system and the economic difficulties caused by installing a multi-controller system.

[0006] In order to achieve the above object, the technical solution of the present invention is implemented as follows: an optimization method for improving the stability and economy of a wind-solar-thermal bundled power system using an optimal dynamic convergence method, the steps of which are as follows:

[0007] Step 1: Connect the double-fed asynchronous induction wind turbine, photovoltaic PV generator and thermal generator to the busbar of the multi-machine power system through a transformer to build a wind-solar-thermal bundled power system;

[0008] Step 2: Connect the power system stabilizer, unified power flow controller and static synchronous compensator of the additional power oscillation damper to the wind-solar-thermal bundled power system;

[0009] Step 3: According to the bus voltage and generator active power of the controller in the wind-solar-thermal bundled power system, establish an objective function based on damping ratio, voltage and active power;

[0010] Step 4: Use the optimal dynamic convergence method based on Euclidean distance weighted strategy, inertia weight dynamic adjustment strategy, gradient position update strategy and adaptive step size update strategy to coordinate and optimize the parameters of the wind-solar-thermal bundled power system controller to obtain the optimized controller parameters.

[0011] Preferably, the objective function is: J=N1A1+N2A2+N3A3; wherein N1, N2, N3 are weight coefficients of the total damping ratio A1, the bus voltage A2 and the generator active power A3, respectively, and N1+N2+N3=1.

[0012] Preferably, the total damping ratio

[0013] The bus voltage

[0014] The generator active power

[0015] Among them, ξ l is the damping ratio of the lth busbar node in the low-frequency oscillation mode, n is the number of busbar nodes; k is the number of generators, m is the number of buses; V l (t) is the voltage of the lth node at time t; V l(0) is the voltage of the lth node before the fault occurs; V lmin Indicates voltage V l (t) is the minimum voltage value; O1, O2, L1, L2 and L3 are proportional coefficients, t0, t a are the end time of the fault and simulation respectively; P s (t) represents the active power of the sth generator at time t, P s (0) represents the active power at the initial moment; P smin represents the minimum active power of the sth generator; R z is the resistance on the zth busbar, θ z is the power factor on the zth bus, P z is the power on the zth bus.

[0016] Preferably, the proportionality coefficient O1=1 / [n×(t a -t0)], O2=1, L1=1 / k, L2=1 and L3=1 / n.

[0017] Preferably, the optimal dynamic convergence method is implemented as follows:

[0018] (1) Set the initial parameters of the algorithm and initialize the search individuals by generating random positions;

[0019] (2) The state matrix of the linear equation of the wind-solar-thermal bundled power system connected to the controller is used to calculate the damping ratio ξ of each node under low-frequency oscillation l , calculate the total damping ratio A1;

[0020] (3) Calculate the voltage V of the lth node through time domain simulation l , and find its minimum value V lmin , calculate the active power P of the sth generator s , and extract the zth resistor and controller parameters to calculate the generator active power deviation and line loss;

[0021] (4) Calculate the proportional coefficients O1, O2, L1, L2, and L3, and calculate the bus voltage A2 and the generator active power A3 respectively based on the proportional coefficients.

[0022] (5) Taking the objective function J as the fitness function, the Euclidean distance weighted strategy, the inertia weight dynamic adjustment strategy, the gradient position update strategy and the adaptive step size update strategy are used to update the position of the search individual and obtain the current optimal search individual position;

[0023] (6) Update the objective function J according to the current optimal search individual position and update the number of iterations;

[0024] (7) Determine whether the iteration termination condition is met. If the iteration termination condition is met, output the optimized controller parameters of the wind-solar-thermal bundled power system; otherwise, return to step (5).

[0025] Preferably, in step (1), the search individuals are initialized by random allocation: X b = l b +rand(u b -l b );where u b and l b are the upper and lower bounds of the searched individual, rand is a random number between 0 and 1; X b To search for individuals.

[0026] Preferably, the method for updating the location of the search individual comprises:

[0027] The position update formula in the food search phase is:

[0028]

[0029] Where: X i (t) is the position of the i-th search individual at iteration number t; X i (t+1) is the position of the i-th search individual after iteration; is the uth search individual X at iteration number t u (t) and the position X of the R1th search individual R1 (t), represents the Euclidean distance function; w(t) is the inertia weight and T is the maximum number of iterations; X u (t) represents the position of the u-th search individual selected randomly, and R1≠u; is the adaptive coefficient; ε is the decision factor; rand is a random number between 0 and 1;

[0030] The attack prey position update formula is:

[0031]

[0032] Where α(t) is the adaptive step size; is the adaptive factor; X gbest is the optimal fitness position; X gbest Search for the current optimal individual position; is a gradient operator used to calculate the optimal search individual position X gbest and the current position X of the i-th search individual i (t) between the gradient information.

[0033] Preferably, the inertia weight w max 、w min are the maximum and minimum values ​​of the inertia weight respectively;

[0034] The adaptive coefficient c h is a constant.

[0035] Preferably, the adaptive step size α max is the maximum step length, T is the maximum number of iterations;

[0036] The adaptive factor

[0037] Preferably, the nonlinear function in the differential-algebraic equation of the wind-solar-thermal bundled power system connected to the controller is Taylor expanded near the equilibrium point, and the deviation of the algebraic vector at the equilibrium point is inversely solved using the state matrix F to obtain the linear equation at the equilibrium point, and the state matrix F is calculated by the first-order partial derivative constant matrix of the nonlinear function;

[0038] The method for calculating the damping ratio under each low-frequency oscillation is: the eigenvalue of the state matrix F is λ F =α+jβ, where α and β are the real and imaginary parts of the eigenvalue with the largest real part, respectively, and the oscillation frequency is The damping ratio is:

[0039] The iteration termination condition is that the number of iterations t reaches the preset maximum number of iterations T or the search individual overall error function Ek(t)≤η; and the threshold

[0040] The objective function J is used to evaluate the fitness of the search individual position of each iteration, and the optimal search individual position is recorded. Through multiple iterations, the optimal solution is gradually reached. At this time, the objective function J is the smallest, and the optimal damping ratio is obtained. The information of the optimal search individual position represents the parameters of the optimized controller.

[0041] The present invention is achieved by:

[0042] 1. The introduction of the Euclidean distance weighted strategy can dynamically adjust the weights according to the relative distances between solutions, so that the optimal dynamic convergence method can better adapt to different search spaces and improve search efficiency.

[0043] 2. The inertia weight dynamic adjustment strategy can dynamically adjust the inertia weight according to the current search status. In the early stage of the search, a larger inertia weight is used to expand the search range and improve the global search capability; in the later stage of the search, a smaller inertia weight is used to narrow the search range and improve the local search accuracy.

[0044] 3. Using the gradient position update strategy, the update direction of the remaining search individuals is guided according to the position information of the optimal solution. This update method can better guide the algorithm to perform fine search in local areas, thereby improving local search capabilities.

[0045] 4. Apply an adaptive step size update strategy to dynamically adjust the step size according to the current number of iterations, so that a larger step size is used for fast search in the early stage and a smaller step size is used for fine search in the later stage. This dynamic adjustment can improve the convergence speed of the algorithm.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows: in view of the problems of the existing methods in the selection and adjustment of controller parameters under high-dimensional nonlinear systems, an advanced swarm intelligent optimization algorithm, namely the optimal dynamic convergence method, is introduced to realize the coordinated optimization control of wind, solar and thermal bundled power systems. The optimal dynamic convergence method proposed in the present invention has shown many advantages in dealing with the optimization problem of power system controllers: the Euclidean distance weighting strategy effectively improves the search efficiency, the inertia weight dynamic adjustment strategy fully improves the local search accuracy of the algorithm, the gradient position update strategy significantly improves the local development capability, and the adaptive step size update strategy ensures the convergence speed of the algorithm. The optimal dynamic convergence method of the present invention combines the Euclidean distance weighting strategy, the inertia weight dynamic adjustment strategy, the gradient position update strategy and the adaptive step size update strategy to update the individual optimal position, which not only enhances the local search capability, but also improves the global search efficiency, effectively improves the low-frequency oscillation in the power system, and reduces the active network loss generated by the system operation, thereby significantly improving the overall stability and economy of the power system.

[0047] The optimal dynamic convergence method proposed in the present invention optimizes the parameters of the power system controller and is applicable to problems such as the negative interactions between multiple control parameters and the economy of system operation in high-dimensional nonlinear power systems. It can significantly improve the negative interactions between multiple controllers, give full play to the performance of the controllers, and ensure the safety and economy of power system operation. The new algorithm idea of ​​the present invention achieves optimal coordination between multiple controllers, thereby eliminating the negative interactions between multiple controllers, which can effectively improve the overall damping characteristics of the system, enhance the system's ability to suppress low-frequency oscillations, improve the economy of system operation, and reduce the economic cost of system operation. Through this collaborative optimization control method, the present invention is expected to provide a more efficient and reliable improvement method for wind, solar and thermal bundled power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0049] Figure 1 It is a flow chart of the present invention.

[0050] Figure 2 It is a schematic diagram of a doubly-fed asynchronous wind generator (DFIG) model of the present invention.

[0051] Figure 3 It is a schematic diagram of a photovoltaic power station model of the present invention.

[0052] Figure 4 It is a schematic diagram of the wind-solar-thermal bundled power system model of the present invention.

[0053] Figure 5 Schematic diagram of a parameter model of a conventional power system stabilizer (PSS) of the present invention.

[0054] Figure 6 Schematic diagram of the static synchronous compensator (STATCOM) model of the present invention, wherein (a) is the equivalent model and (b) is the transfer function.

[0055] Figure 7 Schematic diagram of the unified power flow controller (UPFC) model of the present invention.

[0056] Figure 8 It is a schematic diagram of the transfer function of the unified power flow controller (UPFC) of the present invention.

[0057] Fig. 9 Schematic diagram of the transfer function of the static synchronous compensator (STATCOM-PODC) of the additional power oscillation damper of the present invention.

[0058] Fig.10 is the response curve of the system when there is a three-phase short circuit fault, where (a) is the relative power angle δ G1,G5 , (b) is the relative power angle δ G2,G5 , (c) is the bus 1 voltage V Bus1 , (d) is the bus 2 voltage V Bus2 , (e) is the active power P of generator 1 G1 , (f) is the active power P of generator 2 G2 .

[0059] Fig.11The three-phase short-circuit fault response curve when the tie line power is reduced by 25%, where (a) is the relative power angle δ G1,G5 , (b) is the relative power angle δ G2,G5 , (c) is the bus 1 voltage V Bus1 , (d) is the bus 2 voltage V Bus2 , (e) is the active power P of generator 1 G1 , (f) is the active power P of generator 2 G2 .

[0060] Fig.12 To change the three-phase short-circuit fault response curve when the wind, solar and thermal power output is changed, (a) is the relative power angle δ G1,G5 , (b) is the relative power angle δ G2,G5 , (c) is the bus 1 voltage V Bus1 , (d) is the bus 2 voltage V Bus2 , (e) is the active power P of generator 1 G1 , (f) is the active power P of generator 2 G2 . DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0062] like Figure 1 As shown in FIG. 1 , an optimization method for improving the stability and economy of a wind-solar-thermal bundled power system using an optimal dynamic convergence method is shown in FIG. 1 . The specific steps are as follows:

[0063] Step 1: Based on the IEEE four-machine two-area power system model, the double-fed asynchronous induction wind turbine DFIG, photovoltaic PV generator and thermal generator (for the convenience of analysis, the fourth-order synchronous generator G5 is used instead of the thermal generator) are connected to bus 6 through a transformer to build a dynamic model of the wind-solar-thermal bundled power system.

[0064] The mathematical model of the multi-machine power system can be expressed by the following differential-algebraic equation:

[0065]

[0066] Where x is the state vector, is the derivative of the state vector x, y is an algebraic variable. g(x,y) and z(x,y) are nonlinear functions of the state variable x.

[0067] Taylor expansion of the nonlinear functions g(x,y) and z(x,y) in the above equation near the equilibrium point (x0,y0) and ignoring high-order infinitesimals yields:

[0068]

[0069] Where Δx and Δy are the deviations of the state vector x and the algebraic vector y relative to the equilibrium point (x0, y0), respectively; F1, F2, F3, F4 are the first-order partial derivative constant matrices of the nonlinear functions g(x, y) and z(x, y), and F4 is reversible. The inverse solution of Δy gives the following formula:

[0070]

[0071] Among them, formula (3) is called the linearized system of the original system formula (1); F is the system matrix. According to the Lyapunov stability law, if the eigenvalues ​​of the state matrix F all have negative real parts, the multi-machine power system is asymptotically stable.

[0072] DFIG is widely used in various types of wind farms due to its low cost, fast response time, small required power converter capacity, and decoupled control of active and reactive power. Its operating principle is that the wind turbine converts wind energy into mechanical energy and then converts mechanical energy into electrical energy. Figure 2 As shown in the figure, in this system, the wind turbine is connected to the DFIG through a transmission system, which consists of a low-speed shaft, a high-speed shaft and a gearbox in the middle. The DFIG system is an induction generator whose stator winding is directly connected to the three-phase grid, and the rotor winding is powered by the rotor side converter (RSC), the grid side converter (GSC) and the DC link capacitor placed between the two converters. The controller also consists of three parts: rotor side controller (RSC Controller), grid side controller (GSC Controller) and wind turbine controller (WT Controller).

[0073] like Figure 3 As shown, photovoltaic PV generators mainly use photovoltaic array power generation mode. Photovoltaic arrays are used to convert solar radiation into electrical energy, and the smallest component is photovoltaic cells. According to the characteristics of photovoltaic cells, the model describing the characteristics of photovoltaic arrays is:

[0074]

[0075] Where V and I are the voltage and current output by the photovoltaic array respectively; I ph and I d are the photocurrent and reverse saturation current of the diode respectively; R s and R share the equivalent series impedance and the equivalent parallel impedance respectively; q, k, T' are the electron charge, Boltzmann constant and photovoltaic cell temperature respectively, A' is a constant that changes with voltage V and 1≤A'≤2.

[0076] On bus 6 of the IEEE four-machine two-area power system, a double-fed asynchronous induction wind turbine DFIG, a photovoltaic PV generator and a thermal generator are connected through a transformer to construct a wind-solar-thermal bundled (WPTB) power system dynamic model. Figure 4 As shown. At this time, the dimensions of the state vector x and the algebraic variable y in the system state matrix will be expanded. In the WPTB system, generators G1, G2 and G3, G4 are respectively located in two similar areas, which are connected by a 220km transmission line. Since the system capacity of area 1 and area 2 is large, it is easy to become unstable when the power system is disturbed. There are two coupled units in each area, namely generators G1, G2 and G3, G4. The system load model is a constant impedance load L7 and L9, which are connected to bus No. 07 and bus No. 09 respectively; C7 and C9 are reactive compensation capacitors, which provide reactive power for the system; bus 06 and bus 10 are symmetrically connected to generators G1, G2 and G3, G4 respectively through transformers, and connected to bus No. 7 and bus No. 9 through resistors to balance the output power of different generators.

[0077] Step 2: Connect the power system stabilizer PSS, unified power flow controller UPFC and static synchronous compensator STATCOM-PODC with additional power oscillation damper to the wind-solar-thermal bundled power system.

[0078] like Figure 4 As shown in the figure, the power system stabilizer PSS is set in front of the generator set. The power system stabilizer PSS receives feedback signals such as grid frequency and generator speed. After signal processing, it provides additional damping to the wind-solar-thermal bundled power system to suppress system oscillation. Figure 5 As shown, the overall function of the PSS of the present invention is expressed as follows:

[0079]

[0080] Where V(s) is the output signal of the PSS connected to the generator; T ω is the time constant of the DC isolation link of PSS; T1 and T3 are the leading time constants of the phase compensation link of PSS; T2 and T4 are the lagging time constants of the phase compensation link of PSS; K T is the gain of PSS. The transfer function block diagram of lead-lag PSS is as follows: Figure 5 As shown, and They are the output upper and lower limits of the limiting link respectively.

[0081] The equivalent diagram of static synchronous compensator STATCOM is as follows Figure 6 As shown, STATCOM is connected in parallel in the bus 08 of the WPTB system, which is equivalent to a variable reactive current source, and can flexibly control its reactive current and automatically compensate for the reactive power required by the system. In the present invention, a STATCOM current injection model is adopted, such as Figure 6 As shown, STATCOM is equivalent to a time constant regulator, whose nodes inject reactive power Q and injection current i SH It can be expressed as:

[0082]

[0083] Where V a is the reference voltage; V1 is the node measurement voltage; V POD POD output signal; K r , T 1S are the gain and time constant of the STATCOM voltage regulator respectively; i SH is the injected current of STATCOM node; is the injected current i SH The first-order differential of .

[0084] The structure diagram of the unified power flow controller UPFC is as follows Figure 7 As shown in the figure, UPFC consists of two voltage source converters T1 and T2 that share a DC side capacitor. One converter T1 is connected to the system in parallel through a transformer, and the other converter T2 is connected in series in the transmission line through a transformer. The unified power flow controller UPFC can flexibly and dynamically adjust and optimize the power flow distribution to ensure that the transmission capacity of the power grid is maximized, while significantly improving the stability and reliability of the power grid.

[0085] In the picture: U a -U d is the node voltage; U e ,I e are the voltage and current on the parallel converter side respectively; U f ,I f is the voltage and current on the series conversion side; C a and δ a are the amplitude regulation ratio and phase angle of the parallel transformer VSC1; C b and δ b They are the amplitude regulation ratio and phase angle of the series transformer VSC2 respectively. By adjusting the amplitude regulation ratio and the phase angle, the power flow distribution and the line voltage drop of the line can be effectively controlled.

[0086] The mathematical model of UPFC series-parallel converter can be expressed as:

[0087]

[0088] Where, L f , L e are the equivalent inductances of the series and parallel lines respectively; R f , R e are the equivalent internal resistances of the series and parallel converters, respectively; the subscripts i = a, b, c, respectively, representing the three phases a, b, and c, U bi is the abc three-phase voltage of node 2; U fi , U ei Respectively represent the three-phase voltages on the series side and the parallel side converter side; i fi 、i ei Represent the three-phase currents on the series and parallel converter sides respectively.

[0089] As a highly integrated composite control device, STATCOM-PODC controller integrates the core functions of STATCOM and PODC systems, and can effectively improve the stability and damping ratio of the system. Fig. 9 In the design of the present invention, P is selected to be used i As the input signal of STATCOM-PODC. The STATCOM-PODC control block diagram designed by the present invention is as follows Fig. 9 As shown, the transfer function of the additional power oscillation damper PODC is as follows:

[0090]

[0091] In the formula, K PODC is the proportional gain of PODC; T a is the time constant of the DC isolation link; T 1P , T 3P is the time constant of the leading link; T 2P and T 4P K is the time constant of the lag link; G(S) is the output signal of POD. r ,T 1S are the gain and time constant of the STATCOM voltage regulator respectively; i max ,i min They represent the upper and lower amplitudes of the current injected into the STATCOM node respectively.

[0092] Connect the POD in parallel in the transmission line and measure the active power P of the line. iSignals can effectively suppress the system's interval oscillation. The integrated use of STATCOM-POD and UPFC can quickly respond to the dynamic changes of the system, optimize power flow distribution and voltage control, ensure the stability of the system bus voltage and generator active power, and improve the stability of the system.

[0093] Step 3: According to the bus voltage and generator active power of the wind-solar-thermal bundled power system controller, establish an objective function based on damping ratio, voltage and active power.

[0094] Assume that the eigenvalue of the state matrix F of the mathematical model of the multi-machine power system is λ F =α+jβ, where α and β are the real and imaginary parts of the eigenvalue with the largest real part, and the oscillation frequency is Then the damping ratio ξ can be obtained F :

[0095]

[0096] Damping ratio ξ F It determines the speed of oscillation amplitude attenuation. As can be seen from the above formula, the higher the damping ratio and the lower the oscillation frequency, the better the stability of the power system. At the same time, the damping adjustment is used to suppress power oscillation while minimizing the impact on voltage control and active power. By controlling the active power of the generator, the active network loss of the system can be effectively reduced and the economy of the system can be improved. Based on this, the objective function of the present invention is as follows:

[0097]

[0098] In the formula, ξ l is the damping ratio in the lth low-frequency oscillation mode; n is the number of busbar nodes; k is the number of generators; m is the number of buses; V l (t) is the per unit value of the voltage at the lth node at time t; V l (0) is the per unit value of the voltage of the lth node before the fault occurs; V lmin Indicates voltage V l (t) minimum voltage value; O1 = 1 / [n×(t a -t0)], O2 = 1, L1 = 1 / k, L2 = 1 and L3 = 1 / n are proportional coefficients. t0, t a are the fault and simulation end time respectively; N1=0.4, N2=0.3, N3=0.3 are the weight coefficients of the total damping ratio A1, bus voltage A2 and generator active power A3. s (t) represents the active power per unit value of the sth generator at time t, P s (0) represents the active power per unit value at the initial moment; P smin Represents the minimum active power of the sth generator.z is the resistance on the zth busbar; θ z is the power factor on the zth bus, P z is the power on the zth bus.

[0099] Step 4: Use the optimal dynamic convergence method based on Euclidean distance weighted strategy, inertia weight dynamic adjustment strategy, gradient position update strategy and adaptive step size update strategy to coordinate and optimize the parameters of the wind-solar-thermal bundled power system controller to obtain the optimized controller parameters.

[0100] The optimal dynamic convergence method of the present invention has the following steps:

[0101] (1) Set the initial parameters of the optimal dynamic convergence optimization (ODCO) algorithm and initialize the search individuals by generating random positions. Random initialization ensures that different areas in the search space have the possibility of being explored.

[0102] The population and parameters are initialized by random allocation using the following formula.

[0103] X b = l b +rand(u b -l b ) (10)

[0104] Where: u b and l b are the upper and lower bounds of the searched individual, rand is a random number between 0 and 1; X b To search for individuals.

[0105] (2) Linearize the WPTEB system at the equilibrium point (x0, y0) to obtain the state matrix F of the system, calculate the eigenvalues ​​of the state matrix F, and calculate the damping ratio ξ under each low-frequency oscillation l .

[0106] (3) Calculate the voltage V of the first node through time domain simulation l , and find its minimum value V lmin .

[0107] (4) Calculate the active power P of the sth generator s , and extract the zth resistor and controller parameters to calculate the generator active power deviation and line loss.

[0108] (5) Calculate the weight coefficients O1, O2, L1, L2, and L3.

[0109] (6) Calculate the objective functions A1, A2 and A3 respectively.

[0110] (7) Update the position and objective function value of the search individual, and update the number of iterations.

[0111] According to the population information, the Euclidean distance weighted strategy, the inertia weight dynamic adjustment strategy, the gradient position update strategy and the adaptive step size update strategy are used to determine the optimal dynamic convergence method, and position update strategies for searching for food and attacking prey are formulated respectively.

[0112] (I) Calculate the objective function value J of each individual through formula (11) and formula (12), and update the optimal individual position.

[0113] (II) The position update formula in the food search phase is:

[0114]

[0115] Where: X i (t) is the current position of the i-th search individual; X i (t+1) is the position of the i-th search individual after iteration; For the u-th search element individual X u (t) and the position X of the R1th search individual R1 (t), represents the Euclidean distance function; w(t) is the inertia weight, and w max 、w min are the maximum and minimum values ​​of the inertia weight respectively. T is the maximum number of iterations; X u (t) represents the position of the randomly selected u-th individual, represents the position of the R1th individual randomly selected, and R1≠u; is the adaptive coefficient, c h is a constant; ε is a decision factor, ε=0.5; rand is a random number between 0 and 1; the adaptive coefficient is used It can weigh the average position of the search individuals according to the current iteration information, reduce the algorithm's dependence on the initial value, and improve the robustness of the algorithm.

[0116] (III) The attack prey position update formula is:

[0117]

[0118] Where: α max is the maximum step length, step length is the adaptive factor, X gbest is the optimal fitness position; X gbestSearch for the current optimal individual position; is a gradient operator used to calculate the optimal search individual position X gbest and the current search individual position X i The gradient information between (t) is used to guide the remaining search individuals to move towards the optimal search individual.

[0119] The optimal solution is found through formula (11) and formula (12). The position of each search individual represents a candidate solution in the solution space. The quality of the candidate solution is evaluated using the objective function J as the fitness function. The smaller the objective function J, the closer the candidate solution is to the optimal solution. Through multiple iterations, the position of the search individual is continuously updated, gradually approaching the optimal solution of the objective function.

[0120] (8) Determine whether the iteration termination condition is met. The iteration termination condition is that the number of iterations t reaches the preset maximum number of iterations or the search individual overall error function Ek(t) ≤ η. And the threshold Search individual overall error function X i (t) is the position of the current search individual.

[0121] (9) End and output the optimized controller parameters (K T , T1, T3, K UP , T u , K PODC , T a , K 1ST , T 1ST , T 1P , T 3P ).

[0122] The method of outputting the controller parameters in the WPTB system corresponding to the optimal damping ratio is as follows: in the optimization process, the objective function J is used to evaluate the position information value of each search individual at each iteration, and the optimal position information of the search individual after each iteration is recorded. The objective function J includes the total damping ratio A1. The convergence of the algorithm to the optimal value indicates that the objective function has reached the minimum at this time. At this time, the damping ratios ξ in the total damping ratio A1 l Maximum, in the power system, we want to ensure the stability of the system, that is, the real part of the eigenvalue of the state matrix F is all negative, the larger the damping ratio, the better. Use the optimal dynamic convergence method, through formula (11) and formula (12), update the position of the search individual, use the objective function J to evaluate the value of the search individual position of each iteration, and record the optimal solution. Through multiple iterations, the optimal solution is gradually reached. When the algorithm ends, it means that it has converged to the optimal solution. At this time, the objective function J is the smallest, and the optimal damping ratio can be obtained. The position information of the optimal search individual represents the parameters of the optimized controller.

[0123] The optimal dynamic convergence method proposed in the present invention adopts a variety of adaptive update strategies for position update, wherein the effects of adopting different update strategies are as follows:

[0124] (1) Euclidean distance weighted strategy: by calculating the current search individual position X i (t) and the position X of the R1th individual R1 The Euclidean distance between (t) and the weighted average of these distances is used to update the position, which can effectively explore the local area, increase the diversity of the search, avoid falling into the local optimum too early, and thus find a better local area in the search space, thereby improving the exploration ability and optimization performance of the algorithm.

[0125] (2) Dynamic adjustment strategy of inertia weight: The inertia weight w(t) is used to dynamically adjust the position of the search individual using the information of each iteration to balance the strategies of global search and local search. When the number of iterations is small, a larger inertia weight helps the search individual to conduct extensive exploration in the search space and find potential high-quality areas. When the number of iterations is large, a smaller inertia weight helps the search individual to conduct a detailed search in known high-quality areas, thereby improving the convergence speed and accuracy.

[0126] (3) Gradient position update strategy: The gradient position update strategy can accurately guide the search direction, accelerate convergence and avoid oscillation by utilizing the gradient information at the optimal search cell position. It is particularly suitable for fine-tuning when approaching the optimal solution, thereby improving the convergence speed and accuracy of the algorithm and enhancing its adaptability and robustness.

[0127] (4) Adaptive step-size update strategy: The step-size is dynamically adjusted by the number of iterations (when updating the position of the attacking prey). Large step-size exploration can be performed in the early stage of the search, and small step-size fine adjustment can be performed in the later stage, thereby balancing the global search and local search, improving the convergence speed and accuracy of the algorithm, enhancing its adaptability and robustness, and avoiding falling into the local optimum too early.

[0128] In summary, the optimal dynamic convergence method can improve the local search ability and global search ability, accelerate the convergence speed, increase the stability of the algorithm, and thus improve the performance and effect of the algorithm. When solving the optimal value of the objective function, it is not easy to fall into the trap of local minima and the local optimal phenomenon occurs, which makes up for the shortcomings of the traditional iterative method. Therefore, the application of the optimal dynamic convergence method to the optimization controller parameters has certain theoretical and practical significance.

[0129] Simulation verification: The present invention chooses to use the IEEE four-machine two-zone system for simulation. The system consists of a sending-end system and a receiving-end system connected by an AC transmission line. In addition, a wind-solar-thermal bundled transmission system consisting of a thermal power unit, a wind power unit and a photovoltaic power station is added to the original sending-end system. In order to facilitate analysis, the present invention chooses to use wind turbine and photovoltaic single-machine models to replace the lumped models of wind farms and photovoltaic power plants in the modeling process. The benchmark capacity of the generator set in this model is 100MVA, the frequency is 50Hz, the transmission power of the interconnection line is set to 400MW, the wind turbine output is 30MW, the photovoltaic output is 40MW, and the thermal power unit output is 300MW. Since the four-machine two-zone system is a weakly damped system, the present invention introduces a type II speed regulator and a type III automatic voltage regulator into the original system architecture to enhance system stability.

[0130] In order to verify the performance of the controller after the proposed optimization scheme optimizes the parameters, a detailed eigenvalue analysis and time domain simulation of the system are carried out. The eigenvalue calculation results of the system are shown in Table 5, where the system exhibits five low-frequency oscillation modes. Specifically, mode 1 corresponds to the intra-regional oscillation mode of region 2, and mode 2 reflects the intra-regional oscillation mode of region 1. Modes 3 and 4 are inter-regional oscillation modes between regions 1 and 2, while mode 5 is a global oscillation mode.

[0131] Case 1: Assume that at 1.00s, a three-phase short circuit fault occurs in one of the double-circuit interconnection lines 8 and 9, and the fault duration is 50ms. The electromechanical oscillation modes of the system in Case 1 are shown in Table 1.

[0132] Table 1 Set the electromechanical oscillation modes of the power system after three-phase fault

[0133]

[0134] ALL in the table means G1-G5, DFIG.

[0135] Case 2: In order to verify the performance of the optimal dynamic convergence method to optimize the controller in terms of small signal stability of the power system, the output of the synchronous generator sets in the adjustment area 1 is adjusted to reduce the power of the interconnection line by 25%, and the impact of the change in the transmission power of the interconnection line on the five oscillation modes of the system is studied.

[0136] Table 2 Electromechanical oscillation modes of the power system after changing the transmission power of the tie line

[0137]

[0138] ALL in the table represents G1-G5, DFIG.

[0139] Case 3: In order to more comprehensively study the effect of the controller designed by the present invention on improving system stability, the present invention will change the output of the wind, solar and thermal bundled transmission system, in which the photovoltaic output is increased by 40MW, the wind turbine output is 40MW, the thermal power output is 255MW, the benchmark capacity is 100MW, and the interconnection line power is 400MW.

[0140] Table 3 Adding low-frequency oscillation mode of wind-solar-thermal bundled power transmission system

[0141]

[0142] ALL in the table represents G1-G5, DFIG.

[0143] Case 1 studies the time domain simulation results of the power system after the coordinated optimization of controller parameters and positions in the case of a three-phase short circuit fault. Fig.10 shown. Fig.10 The figure shows the relative power angle curves of generators G1, G5 and G2, G5, the active power curves of generators 1 and 2, and the voltage curves of busbars 1 and 2 in the case of a three-phase short circuit fault. Fig.10 It can be seen that when the system is equipped with a controller with optimized parameters, the relative power angle curve and active power oscillation curve of the generator set are significantly improved, the decay speed is faster, the oscillation amplitude is smaller, and the voltage oscillation amplitude of bus 1 and bus 2 is effectively improved, and the system can be quickly stabilized. The problem of power oscillation in the system after receiving interference has been significantly improved.

[0144] Case 2 studied a three-phase short circuit fault when the tie line power was reduced by 25%. The time domain simulation results of the power system after the controller parameters and position were coordinated and optimized are shown in the attached figure. Fig.11 As shown. Fig.11 It can be clearly seen that after optimizing the controller parameters, the overall robustness of the system has been enhanced, and the system's anti-interference ability has been significantly improved.

[0145] Case 3 studies the three-phase short circuit fault under the condition of changing wind, solar and thermal output. The time domain simulation results of the power system after the coordinated optimization of controller parameters and positions are shown in the attached figure. Fig.12 As shown. Fig.12 It can be seen that after optimizing the controller parameters, the overall system performance is improved, the role of the controller is fully played, and the negative interactions between multiple controllers are significantly reduced.

[0146] Tables 1 to 3 are the electromechanical oscillation modes of the system, which can reflect the damping ratio and low-frequency oscillation characteristics of the system. It can be seen from Table 1 that after optimizing the controller parameters, the overall damping characteristics of the system have been significantly improved. For example, the damping ratio of the optimized intra-region oscillation mode of Mode 2 is improved by about 56.82% compared with the original system, and by about 42.82% compared with before optimization; the global oscillation mode is improved by about 3.07% and 2.07% respectively; the overall loss of the system operation is reduced by about 16.74% and 5.64% respectively. Therefore, it can be shown that the designed optimization algorithm can significantly improve the damping ratio of the system, thereby significantly enhancing the stability of the system and reducing the economic cost of the system operation.

[0147] It can be seen from Table 2 that after optimizing the controller parameters, the minimum damping of the intra-region oscillation mode in region 2 is about 40.96% higher than the maximum intra-region oscillation damping ratio in the original system; the global oscillation mode is improved by about 54.8%; and the overall loss of the system operation is reduced by about 20.13%. This shows that the damping characteristics of the optimized system have been greatly improved, thereby improving the stability of the system and the economy of the system operation.

[0148] It can be seen from Table 3 that after optimizing the controller parameters, the damping of the oscillation mode in the lowest area is improved by about 13.4% compared with the original system; the global oscillation mode is improved by about 22.35% compared with before optimization; and the overall loss of the system operation is reduced by about 29.11%. It can be seen that after the optimization, the damping ratio of the system has been significantly improved, thereby improving the stability of the system and reducing the network loss of the system operation.

[0149] In order to suppress low-frequency oscillations and economic operation in wind-solar-thermal bundled power systems, the present invention discusses the simultaneous optimization and setting of PSS, UPFC and STATCOM-PODC controller parameters in multi-machine power systems. The optimal dynamic convergence method is used to optimize PSS and multi-controller parameters, eliminate the negative interactions between multiple controllers, give full play to the advantages of the controller, and move the weakly damped electromechanical oscillation mode of the multi-machine system to the strong damping area specified in the complex plane. Simulation tests of various working conditions are carried out in a four-machine two-area power system to verify the effectiveness of the coordinated optimization of PSS and multi-controller parameter optimization design scheme of the present invention. The eigenvalue results show that after optimizing the controller parameters, the damping characteristics of the system as a whole can be significantly improved, and the network loss of the system operation can be reduced, ensuring the stability of the system while improving the economy of the system. A three-phase short-circuit fault is set in the wind-solar-thermal bundled power system. By observing the response curves of the system under different working conditions, it is further verified that the present invention can significantly improve the damping characteristics of the system and reduce the network loss during system operation when solving the eigenvalue-based multi-machine power system PSS and multi-controller parameter optimization and operation economy problems. At the same time, the optimal dynamic convergence method is insensitive to initial values, has strong adaptive ability, fast convergence speed, and robust and reliable optimization results when performing parameter optimization, and can be further applied to other high-dimensional optimization problems. The present invention provides a new optimization method, which provides new ideas and technical means for solving controller parameter optimization and operation economy in power systems.

[0150] The present invention adopts the optimal dynamic convergence method based on Euclidean distance weighting strategy, inertia weight dynamic adjustment strategy, gradient position update strategy and adaptive step size update strategy to coordinate and optimize the parameters of multiple controllers in the wind-solar-thermal bundled power system. This method has the advantages of strong global optimization ability, fast convergence speed and strong adaptive ability, and can quickly find the optimal controller parameters. It has high theoretical and practical application value.

[0151] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An optimization method for improving the stability and economy of a wind-solar-thermal bundled power system using an optimal dynamic convergence method, characterized in that: The steps are as follows: Step 1: Connect the double-fed asynchronous induction wind turbine, photovoltaic PV generator and thermal generator to the busbar of the multi-machine power system through a transformer to build a wind-solar-thermal bundled power system; Step 2: Connect the power system stabilizer, unified power flow controller and static synchronous compensator of the additional power oscillation damper to the wind-solar-thermal bundled power system; Step 3: According to the bus voltage and generator active power of the controller in the wind-solar-thermal bundled power system, establish an objective function based on damping ratio, voltage and active power; Step 4: Use the optimal dynamic convergence method based on Euclidean distance weighted strategy, inertia weight dynamic adjustment strategy, gradient position update strategy and adaptive step size update strategy to coordinate and optimize the parameters of the wind-solar-thermal bundled power system controller to obtain the optimized controller parameters.

2. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to claim 1 is characterized in that: The objective function is: J=N1A1+N2A2+N3A3; wherein N1, N2, and N3 are weight coefficients of the total damping ratio A1, the bus voltage A2, and the generator active power A3, respectively, and N1+N2+N3=1.

3. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to claim 2 is characterized in that: The total damping ratio The bus voltage The generator active power Among them, ξ l is the damping ratio of the lth busbar node in the low-frequency oscillation mode, n is the number of busbar nodes; k is the number of generators, m is the number of buses; V l (t) is the voltage of the lth node at time t; V l (0) is the voltage of the lth node before the fault occurs; V lmin Indicates voltage V l (t) is the minimum voltage value; O1, O2, L1, L2 and L3 are proportional coefficients, t0, t a are the end time of the fault and simulation respectively; P s (t) represents the active power of the sth generator at time t, P s (0) represents the active power at the initial moment; P smin represents the minimum active power of the sth generator; R z is the resistance on the zth busbar, θ z is the power factor on the zth bus, P z is the power on the zth bus.

4. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to claim 3 is characterized in that: The proportionality coefficient O1=1 / [n×(t a -t0)], O2=1, L1=1 / k, L2=1 and L3=1 / n.

5. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to any one of claims 2 to 4, characterized in that: The implementation method of the optimal dynamic convergence method is: (1) Set the initial parameters of the algorithm and initialize the search individuals by generating random positions; (2) The state matrix of the linear equation of the wind-solar-thermal bundled power system connected to the controller is used to calculate the damping ratio ξ of each bus node under low-frequency oscillation l , calculate the total damping ratio A1; (3) Calculate the voltage V of the lth node through time domain simulation l , and find its minimum value V lmin , calculate the active power P of the sth generator s , and extract the zth resistor and controller parameters to calculate the generator active power deviation and line loss; (4) Calculate the proportional coefficients O1, O2, L1, L2, and L3, and calculate the bus voltage A2 and the generator active power A3 respectively based on the proportional coefficients. (5) Taking the objective function J as the fitness function, the Euclidean distance weighted strategy, the inertia weight dynamic adjustment strategy, the gradient position update strategy and the adaptive step size update strategy are used to update the position of the search individual and obtain the current optimal search individual position; (6) Update the objective function J according to the current optimal search individual position and update the number of iterations; (7) determining whether an iteration termination condition is satisfied. If the iteration termination condition is satisfied, outputting the optimized controller parameters of the wind-solar-thermal bundled power system; Otherwise, return to step (5).

6. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to claim 5 is characterized in that: In step (1), the search individuals are initialized by random assignment: X b = l b +rand(u b -l b );where u b and l b are the upper and lower bounds of the searched individual, rand is a random number between 0 and 1; X b To search for individuals.

7. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to claim 5 is characterized in that: The method for updating the location of a search individual comprises: The position update formula in the food search phase is: Where: X i (t) is the position of the i-th search individual at iteration number t; X i (t+1) is the position of the i-th search individual after iteration; is the uth search individual X at iteration number t u (t) and the position X of the R1th search individual R1 (t), represents the Euclidean distance function; w(t) is the inertia weight and T is the maximum number of iterations; X u (t) represents the position of the u-th search individual selected randomly, and R1≠u; is the adaptive coefficient; ε is the decision factor; rand is a random number between 0 and 1; The attack prey position update formula is: Where α(t) is the adaptive step size; is the adaptive factor; X gbest is the optimal fitness position; X gbest Search for the current optimal individual position; is a gradient operator used to calculate the optimal search individual position X gbest and the current position X of the i-th search individual i (t) between the gradient information.

8. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to claim 7 is characterized in that: The inertia weight w(t)∈[0.4,0.9],w max 、w min are the maximum and minimum values ​​of the inertia weight respectively; The adaptive coefficient c h is a constant.

9. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to claim 7 is characterized in that: The adaptive step size α max is the maximum step length, T is the maximum number of iterations; The adaptive factor 10. The optimization method for improving the stability and economy of wind-solar-thermal bundled power system by using the optimal dynamic convergence method according to any one of claims 6 to 9, characterized in that: The nonlinear function in the differential-algebraic equation of the wind-solar-thermal bundled power system connected to the controller is Taylor expanded near the equilibrium point, and the deviation of the algebraic vector at the equilibrium point is inversely solved using the state matrix F to obtain the linear equation at the equilibrium point. The state matrix F is calculated by the first-order partial derivative constant matrix of the nonlinear function; The method for calculating the damping ratio under each low-frequency oscillation is: the eigenvalue of the state matrix F is λ F =α+jβ, where α and β are the real and imaginary parts of the eigenvalue with the largest real part, respectively, and the oscillation frequency is The damping ratio is: The iteration termination condition is that the number of iterations t reaches the preset maximum number of iterations T or the search individual overall error function Ek(t)≤η; and the threshold The objective function J is used to evaluate the fitness of the search individual position of each iteration, and the optimal search individual position is recorded. Through multiple iterations, the optimal solution is gradually reached. At this time, the objective function J is the smallest, and the optimal damping ratio is obtained. The information of the optimal search individual position represents the parameters of the optimized controller.

Images