Power system stabilizer parameter optimization method based on NLP system

Abstract

The invention belongs to the technical field of power system dynamic stability analysis and control, and particularly relates to a power system stabilizer parameter optimization method based on an NLP system. The method comprises the following steps: accessing a power system stabilizer (PSS) into a power system, setting different operation modes, constructing a dynamic model of the power system, and performing linearization processing to obtain a state equation of the power system; calculating the characteristic value of the state matrix of the state equation of the power system, calculating the damping ratio of the system, constructing an objective function based on the characteristic value and the damping ratio of the system, and constructing an optimization problem by using the objective function with the control parameter of the PSS as a constraint condition; and the NLP system for adjusting the strategy optimizes the optimization problem to obtain an optimal PSS control parameter value. According to the method, the optimal parameters can be found under different operation conditions, the low-frequency oscillation suppression capability of the system is improved, the overall damping characteristic and dynamic stability of the power system are improved, and the robustness of the power system under various operation modes is enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of analysis and control of dynamic stability of power systems, and in particular relates to a method for optimizing parameters of a power system stabilizer based on an NLP system, which is mainly used to suppress low-frequency oscillations of a power system. Background Art

[0002] At present, the power system is a cross-regional, large-scale, nonlinear system that is affected by various events and may cause low-frequency oscillations. Low-frequency oscillations are rotor angle oscillations generated by the excitation of local electromechanical oscillation modes or inter-regional oscillation modes. In recent years, power system stabilizers (PSSs) have received increasing attention to improve the dynamic performance of power systems. PSSs are very effective controllers that improve the damping of low-frequency oscillations by introducing additional signals in the excitation controller of the generator. Since its parameters such as structure and reliability are easy to adjust online, the implementation of PSSs with lead-lag control structures in actual power systems is still a priority. By properly selecting the parameters of conventional lead-lag controllers, low-frequency oscillations can be fully suppressed. In addition, fuzzy logic technology is also a powerful tool for power system stability. In the past few years, in order to meet different system requirements, researchers have proposed various methods to design the optimal PSS and reported many PSS design techniques. However, these methods still cannot give the global optimal selection of PSS parameters.

[0003] Bionic systems have been considered as effective tools for optimizing PSS parameters, such as improved artificial immune network methods, genetic systems, firefly systems, and particle swarm optimization systems. However, the network structure of the improved artificial immune network method is very complex and needs to be redesigned for different problems. The genetic system searches for the optimal solution by simulating the natural evolution process. It is easy to fall into the local optimum in the late stage of evolution and converges slowly. The firefly system relies on too many initial parameters. The initialization steps of these systems are very critical and will affect the final dynamic response of the controlled system. The particle swarm system is prone to premature convergence and stagnation, and depends on the choice of control parameters. Summary of the invention

[0004] In view of the shortcomings of the above-mentioned prior art, the present invention provides a method for optimizing the parameters of a power system stabilizer based on an NLP system. The purpose is to propose a new heuristic system, namely a neural P system, to optimize the PSS parameters, construct a new objective function based on eigenvalues ​​and damping ratios, and design operation rules such as forgetting rules, spiking rules, evolution rules, and transfer rules. In a single-machine infinite system, simulation verification is carried out under various operating conditions, so that the PSS has strong adaptability, effectively improves the damping ratio of the system, and enhances the invention purpose of the dynamic stability of the power system.

[0005] The technical solution adopted by the present invention to achieve the above-mentioned purpose is:

[0006] A method for optimizing parameters of a power system stabilizer based on an NLP system comprises the following steps:

[0007] Connect the power system stabilizer PSS to the power system, set different operation modes, build a dynamic model of the power system, and linearize the dynamic model to obtain the state equation of the power system;

[0008] The QR method is used to calculate the eigenvalues ​​of the state matrix of the state equation of the power system, the system damping ratio is calculated based on the eigenvalues, an objective function based on the system eigenvalues ​​and the damping ratio is constructed, and the optimization problem is constructed using the objective function with the control parameters of the power system stabilizer PSS as constraints;

[0009] The optimization problem is optimized using the NLP system based on the forgetting rule, spike rule, evolution rule and transfer rule adjustment strategy to obtain the optimal control parameter values ​​of the power system stabilizer PSS.

[0010] Furthermore, the differential-algebraic equation of the dynamic model of the power system is as follows:

[0011]

[0012] In the above formula, x(t)=[θ,ω,E′ q ,E fd ] T is the state variable, θ and ω are the rotor angle and angular velocity of the generator, respectively, E′ q and E fd They are the generator’s quadrature-axis transient voltage and the excitation system output voltage;

[0013] u(t)=[u pss ] is the input control variable, i.e., the auxiliary modulation signal of PSS; f is a set of first-order nonlinear differential equations representing the dynamics of power system and PSS; represents the time derivative of the state variable x;

[0014] The state equation of the power system is constructed by using the Lyapunov linearization method to linearize the generator, excitation system, PSS and load of the power system at the balance point using Taylor series expansion to obtain the state equation of the power system:

[0015]

[0016] In the above formula, is the inverse of the state variable, Δu is the output variable, A is the system state matrix, B is the system input matrix, C is the output matrix, D is the transfer, and the matrices Δx and Δv are the changes of the state variable and the control variable respectively;

[0017] The different operation modes include: thermal power high power output and thermal power deep peak regulation operation mode.

[0018] Furthermore, the optimization problem is as follows:

[0019]

[0020] Among them, J is the objective function, K is the gain in the amplification link, T 1 is the time constant of the phase compensation link, T 2 is the time constant of the phase compensation link, T 3 is the time constant of the phase compensation link, T 4 is the time constant of the phase compensation link, T w is the time constant of the DC isolation link;

[0021] The optimization is continuously iterated with the goal of minimizing the objective function J. During the optimization process, the control parameters of the PSS are used as input and are changed within the constraint range to change the size of the objective function J until the optimization effect is optimal, that is, when the value of the objective function J is minimum, the optimal PSS parameter setting result is output; the optimal PSS control parameter value is substituted into all operating modes to determine whether they all meet the requirements of dynamic stability and robustness; if so, the PSS control parameter optimization is terminated, otherwise, the optimization is performed again until all modes meet the requirements;

[0022] The mathematical model transfer function of PSS is:

[0023]

[0024] In the above formula, U pss is the auxiliary modulation signal provided by PSS to the excitation system, K is the gain in the amplification link, T 1 is the time constant of the phase compensation link, T 2 is the time constant of the phase compensation link, T 3 is the time constant of the phase compensation link, T 4 is the time constant of the phase compensation link, T w is the time constant of the DC isolation link, Δω is the speed deviation of the generator, and s is the differential operator;

[0025] The objective function based on the system eigenvalue and damping ratio is:

[0026]

[0027] In the above formula, Np is the number of runs during parameter optimization, N e is the number of eigenvalues ​​and damping ratios, σ 0 is the real part of the ideal eigenvalue obtained in the target expectation, ζ 0 is the ideal damping index expected in the optimization objective, σ i,j is the real part of the ith eigenvalue of the jth operating mode, ζ i,j is the damping ratio of the i-th low-frequency oscillation mode under the j-th operation mode; α and γ are learning factors and α, γ>0; f 1 (x) and f 2 (x) are two fitness functions in neurons, used to calculate the fitness of PSS parameters;

[0028] Pulse excitation function The expression is as follows:

[0029]

[0030] In the above formula, e is the number of operating points;

[0031] The damping ratio is calculated as follows: the eigenvalue of the system state matrix A is λ=σ±jωσ, the real part of the eigenvalue λ is the damping factor; the imaginary part ω of the eigenvalue λ is the oscillation frequency f=ω / 2π; each pair of eigenvalues ​​corresponds to a low-frequency oscillation mode; the participation factor method is used to identify the low-frequency oscillation mode and calculate the damping ratio as follows:

[0032]

[0033] In the above formula, σ i,j is the real part of the ith eigenvalue of the jth operating mode, ζ i,j is the damping ratio of the i-th low-frequency oscillation mode under the j-th operation mode, ω i,j is the angular frequency of the i-th low-frequency oscillation mode in the j-th operating mode.

[0034] Furthermore, the NLP system optimizes the optimization problem, including:

[0035] Step 31. Initialize the parameters of the NLP system: including the number of iterations, neurons, spikes, spikes and evolution probabilities;

[0036] Step 32. Establishing a membrane structure: Establishing a membrane structure containing Z neurons in the skin membrane, and generating an initial spike in each neuron in the membrane, each neuron has H spikes, and each spike W has a set of PSS parameters; applying the QR method to calculate the eigenvalue of the system state equation, and calculating the damping ratio of the power system according to the eigenvalue;

[0037] Step 33. Calculate the fitness of each peak according to the objective function and save the optimal parameters;

[0038] Step 34. Use the forgetting rule, the spike rule, and the evolution rule to update the parameters in each neuron;

[0039] In the spike rule, for any two spikes, if the update probability P0 < Pc, two new spikes are obtained, and the spikes are updated according to the fitness value; in the evolution rule, the worst t spikes evolve according to a certain probability Pm, and the first t spikes with the best fitness are selected;

[0040] Step 35. Use the transfer rule to transfer the best L spikes of each membrane i in step 32 from the neuron, and bring the best L spikes of the foreign membrane into the neuron i;

[0041] Step 36. Determine whether the objective function meets the termination condition, reaches the maximum number of iterations or converges to the theoretical optimal value, otherwise return to step 32;

[0042] Step 37. Output the best peak and optimal PSS parameters from the skin membrane in step 32.

[0043] Furthermore, the peak rule is as follows:

[0044] P 0 <P c

[0045]

[0046] In the above formula, P c is the probability given by the spike rule, P 0 and r is a random number in (0,1), o′ i1 is the first parameter to be optimized for the i-th new peak, o′ i2 is the second parameter to be optimized for the i-th new spike, The Nth peak to be optimized for the i-th new peak 0 Parameters, o i1 is the first parameter to be optimized for the ith peak, o j2 is the second parameter to be optimized for the j-th peak, The Nth peak to be optimized for the i-th peak 0 parameters, The Nth peak to be optimized for the jth peak 0 parameters, The Nth value to be optimized for the jth new peak 0 parameters;

[0047] Update the peak O according to the value of fitness i and O j :

[0048]

[0049] The evolution rules are:

[0050] P 1 <P m

[0051] R i,evolving :{O min1 ,O min2 ,...,O mint}

[0052] →{O′ min1 ,O′ min2 ,...,O′ maxt}

[0053]

[0054] In the above formula, P m is the probability O given by the evolutionary rule best is the best peak, P 1 is a (0,1) random number, O min1 is the first peak with the worst fitness, O min2 is the second peak with the worst fitness, O mint is the tth peak with the worst fitness, O′ min1 is the first peak with the worst fitness after evolution, O′ min2 is the second peak with the worst fitness after evolution, O′ minj is the jth peak with the worst fitness after evolution;

[0055] Choose the peak value with the best fit:

[0056]

[0057] In the above formula: O min1 is the first peak with the worst fitness, O min2 is the second peak with the worst fitness, O mint is the tth peak with the worst fitness, O′ min1 is the first peak with the worst fitness after evolution, O′ min2 is the second peak with the worst fitness after evolution, O′ minj is the jth peak with the worst fitness after evolution, O′ new1 is the first peak with the best fitness selected from all the peaks with poor fitness, O′ new2 O′ is the second best peak in fitness selected from all the peaks with poor fitness. newtis the tth peak with the best fitness selected from all the peaks with poor fitness mentioned above.

[0058] Furthermore, the transfer rules:

[0059] R i,c =R i,c1 and R i,c2

[0060] R i,cl :[O max1 ,O max2 ,...,O maxL ] i →[ ] i O max1 ,O max2 ,...,O maxL

[0061] R i,c2 :[] i O′ *max1 ,O′ *max2 ,...,O′ *maxL →[O′ max1 ,O′ max2 ,...,O′ maxL ] i

[0062] In the above formula, [] i represents membrane i, {O max1 ,O max2 ,…,O maxL} is the best L peak in film i,

[0063] {O′ max1 ,O′ max2 ,…,O′ maxL} is the best L peak outside membrane i.

[0064] A power system stabilizer parameter optimization device based on an NLP system, comprising:

[0065] The construction and processing module is used to connect the power system stabilizer PSS to the power system, set different operation modes, build the dynamic model of the power system, and linearize the dynamic model to obtain the state equation of the power system;

[0066] A construction module is used to calculate the eigenvalues ​​of the state matrix of the state equation of the power system by using the QR method, calculate the system damping ratio according to the eigenvalues, construct an objective function based on the system eigenvalues ​​and the damping ratio, and construct an optimization problem using the objective function with the control parameters of the power system stabilizer PSS as constraints;

[0067] The optimization module is used to optimize the optimization problem using the NLP system based on the forgetting rule, the spike rule, the evolution rule and the transfer rule adjustment strategy to obtain the optimal control parameter value of the power system stabilizer PSS.

[0068] Furthermore, the device for optimizing parameters of a power system stabilizer based on an NLP system is used to implement the operating steps of any one of the methods for optimizing parameters of a power system stabilizer based on an NLP system.

[0069] A computer device comprises a storage medium, a processor and a computer program stored on the storage medium and executable on the processor, wherein when the processor executes the computer program, the steps of any one of the methods for optimizing parameters of a power system stabilizer based on an NLP system are implemented.

[0070] A computer storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of any one of the methods for optimizing parameters of a power system stabilizer based on an NLP system are implemented.

[0071] The present invention has the following beneficial effects and advantages:

[0072] The present invention connects PSS to the power system, sets different operation modes, constructs a dynamic model of the power system, and linearizes the dynamic model to obtain the state equation of the power system; the QR algorithm is used to calculate the eigenvalue of the state matrix of the state equation of the power system, the system damping ratio is calculated according to the eigenvalue, a dual objective function based on the eigenvalue and the damping ratio is constructed, and the control parameters of the PSS are used as constraints to construct an optimization problem; an NLP (Neural-like P Systems) system based on operation rules such as forgetting rules, spike rules, evolutionary rules, and transfer rules is used to optimize the optimization problem to obtain the optimal value of the PSS control parameter. The present invention is based on the objective function of eigenvalue, damping ratio, and fitness, and can find the optimal parameters under different operating conditions.

[0073] The optimized PSS parameter values ​​of the present invention improve the system's ability to suppress low-frequency oscillations, and also effectively improve the overall damping characteristics and dynamic stability of the power system, and enhance the robustness of the power system under various operating modes. The present invention designs forgetting rules, peak rules, evolution rules and transfer rules, and has strong global search capabilities, can avoid falling into local optimal solutions, improve convergence speed, and help find better solutions.

[0074] The method of the present invention is used in PSS setting, and has good adjustability and adaptability overall. It can still improve the stable operation level of the system under the deep peak regulation operation condition of thermal power, and can simultaneously optimize the parameters of multiple power system stabilizers PSS in the interconnected system.

[0075] In order to select the optimal PSS parameters, the present invention proposes a new NLP system optimization algorithm, designs operation rules such as forgetting rules, peak rules, evolution rules, and transfer rules, so as to have a strong global search capability and avoid falling into the local optimal solution. Compared with the existing method, it has faster convergence speed and higher optimization accuracy, and can quickly find the optimal solution.

[0076] Since the present invention constructs a new objective function based on the characteristic value and the damping ratio, the optimized PSS parameters can not only output more than 60% of active power, but also improve the stable operation level of the system under the deep peak regulation operation condition of thermal power. Compared with the prior art, the present invention enhances the robustness of the power system under various operation modes, can find a good solution in an uncertain environment, and has strong adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0078] Figure 1 is a PSS parameter optimization flow chart of the present invention;

[0079] Figure 2 It is a schematic diagram of the mathematical model of the PSS of the present invention;

[0080] Figure 3 This is the structure of a single-machine infinite system used for simulation in the present invention. DETAILED DESCRIPTION

[0081] In order to more clearly understand the above-mentioned objectives, features and advantages of the present invention, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0082] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0083] Refer to the following Figure 1-Figure 3 The technical solutions of some embodiments of the present invention are described.

[0084] Example 1

[0085] The present invention provides an embodiment, which is a method for optimizing parameters of a power system stabilizer based on an NLP system. Figure 1 As shown, Figure 1 It is a flow chart of power system stabilizer PSS parameter optimization of the present invention.

[0086] The present invention provides an optimization method, which specifically includes the following contents:

[0087] Step 1. Connect the PSS to the power system, set different operation modes, build a dynamic model of the power system, and linearize the dynamic model to obtain the state equation of the power system;

[0088] The differential-algebraic equation of the dynamic model of the power system is:

[0089]

[0090] In the above formula, x(t)=[θ,ω,E′ q ,E fd ] T is the state variable, θ and ω are the rotor angle and angular velocity of the generator, respectively, E′ q and E fd They are the generator’s quadrature-axis transient voltage and the excitation system output voltage;

[0091] u(t)=[u pss ] is the input control variable, i.e., the auxiliary modulation signal of PSS; f is a set of first-order nonlinear differential equations representing the dynamics of power system and PSS; Represents the time derivative of the state variable x.

[0092] The state equation of the power system is constructed as follows:

[0093] The Lyapunov linearization method is used to linearize the generator, excitation system, PSS and load of the power system using Taylor series expansion at the balance point to obtain the state equation of the power system:

[0094]

[0095] In the above formula, is the inverse of the state variable, Δu is the output variable, A is the system state matrix, B is the system input matrix, C is the output matrix, D is the transfer, and matrices Δx and Δv are the changes in the state variable and the control variable, respectively.

[0096] The different operation modes include: thermal power high power output and thermal power deep peak regulation operation mode, the gain K in the amplification link has a value range of [10,50], and the time constant T of the DC isolation link isw The value range is [5, 25]. The time constant T in the compensation link 1 The value range of T is [0.1, 0.3]. 2 The value range is [0.01, 0.08], T 3 The value range is [0.09, 0.4], T 4 The value range is [0.01,0.1].

[0097] Step 2. Apply the existing QR method to calculate the eigenvalue of the state matrix of the state equation of the power system, calculate the system damping ratio according to the eigenvalue, construct an objective function based on the system eigenvalue and the damping ratio, and use the objective function to construct an optimization problem with the control parameters of the power system stabilizer PSS as constraints;

[0098] The QR is an existing technology, which is an orthogonal triangular decomposition method. The most effective and widely used method for finding all eigenvalues ​​of a general matrix is ​​to decompose the matrix into a regular orthogonal matrix Q and an upper triangular matrix R, which is called the QR decomposition method.

[0099] The optimization problem is as follows:

[0100]

[0101] Among them, J is the objective function, K is the gain in the amplification link, T 1 is the time constant of the phase compensation link, T 2 is the time constant of the phase compensation link, T 3 is the time constant of the phase compensation link, T 4 is the time constant of the phase compensation link, T w is the time constant of the DC isolation link.

[0102] The optimization with the objective function as the goal is to continuously iterate the optimization with the goal of minimizing the objective function J. During the optimization process, the control parameters of the PSS are used as input, and the size of the objective function J is changed within the constraint limit until the optimization effect is optimal, that is, when the value of the objective function J is minimum, the optimal PSS parameter setting result is output; the optimal PSS control parameter value is substituted into all operating modes to determine whether they all meet the requirements of dynamic stability and robustness. If so, the PSS control parameter optimization ends, otherwise it is optimized again until all modes meet the requirements. Figure 2 As shown, Figure 2 It is a schematic diagram of the mathematical model of the PSS of the present invention.

[0103] The mathematical model transfer function of the PSS is:

[0104]

[0105] In the above formula, Upss is the auxiliary modulation signal provided by PSS to the excitation system, K is the gain in the amplification link, T 1 is the time constant of the phase compensation link, T 2 is the time constant of the phase compensation link, T 3 is the time constant of the phase compensation link, T 4 is the time constant of the phase compensation link, T w is the time constant of the DC isolation link, Δω is the speed deviation of the generator, and s is the differential operator;

[0106] The objective function based on the system eigenvalue and damping ratio is:

[0107]

[0108] In the above formula, N p is the number of runs during parameter optimization, N e is the number of eigenvalues ​​and damping ratios, σ 0 is the real part of the ideal eigenvalue obtained in the target expectation, ζ 0 is the ideal damping index expected in the optimization objective, σ i,j is the real part of the ith eigenvalue of the jth operating mode, ζ i,j is the damping ratio of the i-th low-frequency oscillation mode under the j-th operation mode; α and γ are learning factors and α, γ>0; f 1 (x) and f 2 (x) are two fitness functions in neurons, which are used to calculate the fitness of PSS parameters.

[0109] is the pulse excitation function, which is expressed as follows:

[0110]

[0111] In the above formula, e is the number of operating points.

[0112] The damping ratio is calculated as follows: the eigenvalue of the system state matrix A is λ=σ±jω, the real part of the eigenvalue λ is the damping factor; the imaginary part ω of the eigenvalue λ is the oscillation frequency f=ω / 2π; each pair of eigenvalues ​​corresponds to a low-frequency oscillation mode; the participation factor method is used to identify the low-frequency oscillation mode and calculate the damping ratio as follows:

[0113]

[0114] In the above formula, σ i,j is the real part of the ith eigenvalue of the jth operating mode, ζ i,j is the damping ratio of the i-th low-frequency oscillation mode under the j-th operation mode, ω i,jis the angular frequency of the i-th low-frequency oscillation mode in the j-th operating mode.

[0115] Step 3. Use the NLP system based on the forgetting rule, peak rule, evolution rule and transfer rule adjustment strategy to optimize the optimization problem and obtain the optimal control parameter value of the power system stabilizer PSS.

[0116] The method for optimizing the optimization problem by the NLP system comprises the following steps:

[0117] Step 31. Initialize the parameters of the NLP system: including the number of iterations, neurons, spikes, spikes and evolution probabilities;

[0118] Step 32. Establishing a membrane structure: Establishing a membrane structure containing Z neurons in the skin membrane, and generating an initial spike in each neuron in the membrane, each neuron has H spikes, and each spike W has a set of PSS parameters; applying the QR method to calculate the eigenvalue of the system state equation, and calculating the damping ratio of the power system based on the eigenvalue;

[0119] Step 33. Calculate the fitness of each peak according to the objective function, and then save the optimal parameters;

[0120] Step 34. Use the forgetting rule, the spike rule, and the evolution rule to update the parameters in each neuron;

[0121] In the spike rule, for any two spikes, if the update probability P0<Pc, two new spikes are obtained, and then the spikes are updated according to the fitness value; in the evolution rule, the worst t spikes evolve according to a certain probability Pm, and then the first t spikes with the best fitness are selected.

[0122] Step 35. Use the transfer rule to transfer the best L spikes of each membrane i in step 32 out of the neuron, and the best L spikes of the foreign membrane are brought into neuron i;

[0123] Step 36. Determine whether the objective function meets the termination condition, that is, whether it reaches the maximum number of iterations or converges to the theoretical optimal value, otherwise return to step 32;

[0124] Step 37. Output the best peak and optimal PSS parameters from the skin membrane in step 32.

[0125] The peak rule is as follows:

[0126] P 0 <P c

[0127]

[0128] In the above formula, P cis the probability given by the spike rule, P 0 and r is a random number in (0,1), o′ i1 is the first parameter to be optimized for the i-th new peak, o′ i2 is the second parameter to be optimized for the i-th new spike, The Nth peak to be optimized for the i-th new peak 0 Parameters, o i1 is the first parameter to be optimized for the ith peak, o j2 is the second parameter to be optimized for the j-th peak, The Nth peak to be optimized for the i-th peak 0 parameters, The Nth peak to be optimized for the jth peak 0 parameters, The Nth value to be optimized for the jth new peak 0 parameters.

[0129] Then, the peak O is updated according to the value of fitness i and O j

[0130]

[0131] The evolution rules are:

[0132] P 1 <P m

[0133] R i,evolving :{O min1 ,O min2 ,...,O mint}

[0134] →{O′ min1 ,O′ min2 ,...,O′ mint}

[0135]

[0136] In the above formula, P m is the probability given by the evolutionary rule, O best is the best peak, P 1 is a (0,1) random number, O min1 is the first peak with the worst fitness, O min2 is the second peak with the worst fitness, O mint is the tth peak with the worst fitness, O′ min1 is the first peak with the worst fitness after evolution, O′ min2 is the second peak with the worst fitness after evolution, O′ minjis the jth peak with the worst fitness after evolution.

[0137] Then, select the peak with the best fit:

[0138]

[0139] In the above formula: O min1 is the first peak with the worst fitness, O min2 is the second peak with the worst fitness, O mint is the tth peak with the worst fitness, O′ min1 is the first peak with the worst fitness after evolution, O′ min2 is the second peak with the worst fitness after evolution, O′ minj is the jth peak with the worst fitness after evolution, O′ new1 is the first peak with the best fitness selected from all the peaks with poor fitness, O′ new2 O′ is the second best peak in fitness selected from all the peaks with poor fitness. newt is the tth peak with the best fitness selected from all the peaks with poor fitness mentioned above.

[0140] The transfer rules are:

[0141] R i,c =R i,c1 and R i,c2

[0142] R i,cl :[O max1 ,O max2 ,...,O maxL ] i →[ ] i O max1 ,O max2 ,...,O maxL

[0143] R i,c2 :[] i O′ *max1 ,O′ *max2 ,...,O′ *maxL →[O′ max1 ,O′ max2 ,...,O′ maxL ] i

[0144] In the above formula, [] i represents membrane i, {O max1 ,O max2 ,...,O maxL} is the best L peak in film i,

[0145] {O′ max1 ,O′ max2 ,...,O′ maxL} is the best L peak outside membrane i.

[0146] Example 2

[0147] The present invention further provides an embodiment, which is a power system stabilizer parameter optimization device based on the NLP system, comprising:

[0148] The construction and processing module is used to connect the power system stabilizer PSS to the power system, set different operation modes, build the dynamic model of the power system, and linearize the dynamic model to obtain the state equation of the power system;

[0149] A construction module is used to calculate the eigenvalues ​​of the state matrix of the state equation of the power system by using the QR method, calculate the system damping ratio according to the eigenvalues, construct an objective function based on the system eigenvalues ​​and the damping ratio, and construct an optimization problem using the objective function with the control parameters of the power system stabilizer PSS as constraints;

[0150] The optimization module is used to optimize the optimization problem using the NLP system based on the forgetting rule, the spike rule, the evolution rule and the transfer rule adjustment strategy to obtain the optimal control parameter value of the power system stabilizer PSS.

[0151] The power system stabilizer parameter optimization device based on the NLP system described in the present invention is used to implement the operating steps of the power system stabilizer parameter optimization method based on the NLP system as described in Example 1.

[0152] Example 3

[0153] Based on the same inventive concept, an embodiment of the present invention further provides a computer device, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor. When the processor executes the computer program, the steps of any one of the methods for optimizing parameters of a power system stabilizer based on an NLP system described in Embodiment 1 are implemented.

[0154] Example 4

[0155] Based on the same inventive concept, an embodiment of the present invention also provides a computer storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the methods for optimizing parameters of a power system stabilizer based on an NLP system described in Example 1 are implemented.

[0156] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0157] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0158] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0159] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0160] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for optimizing power system stabilizer parameters based on NLP system, Its characteristics are: The following steps are involved: Connect the power system stabilizer PSS to the power system, set different operation modes, build a dynamic model of the power system, and linearize the dynamic model to obtain the state equation of the power system; The QR method is used to calculate the eigenvalues ​​of the state matrix of the state equation of the power system, the system damping ratio is calculated based on the eigenvalues, an objective function based on the system eigenvalues ​​and the damping ratio is constructed, and the optimization problem is constructed using the objective function with the control parameters of the power system stabilizer PSS as constraints; The optimization problem is optimized using the NLP system based on the forgetting rule, spike rule, evolution rule and transfer rule adjustment strategy to obtain the optimal control parameter values ​​of the power system stabilizer PSS.

2. According to claim 1, a method for optimizing parameters of a power system stabilizer based on an NLP system, Its characteristics are: The dynamic model of the power system has the following differential-algebraic equations: In the above formula, is the state variable, θ and ω are the rotor angle and angular velocity of the generator, respectively, E′ q and E fd They are the generator’s quadrature-axis transient voltage and the excitation system output voltage; u(t)=[u pss ] is the input control variable, i.e., the auxiliary modulation signal of PSS; f is a set of first-order nonlinear differential equations representing the dynamics of power system and PSS; represents the time derivative of the state variable x; The state equation of the power system is constructed by using the Lyapunov linearization method to linearize the generator, excitation system, PSS and load of the power system at the balance point using Taylor series expansion to obtain the state equation of the power system: In the above formula, is the inverse of the state variable, Δu is the output variable, A is the system state matrix, B is the system input matrix, C is the output matrix, D is the transfer, and the matrices Δx and Δv are the changes of the state variable and the control variable respectively; The different operation modes include: thermal power high power output and thermal power deep peak regulation operation mode.

3. According to the method for optimizing parameters of a power system stabilizer based on an NLP system in claim 1, Its characteristics are: The optimization problem is as follows: Among them, J is the objective function, K is the gain in the amplification link, T 1 is the time constant of the phase compensation link, T 2 is the time constant of the phase compensation link, T 3 is the time constant of the phase compensation link, T 4 is the time constant of the phase compensation link, T w is the time constant of the DC isolation link; The optimization is continuously iterated with the goal of minimizing the objective function J. During the optimization process, the control parameters of the PSS are used as input and are changed within the constraint range to change the size of the objective function J until the optimization effect is optimal, that is, when the value of the objective function J is minimum, the optimal PSS parameter setting result is output; the optimal PSS control parameter value is substituted into all operating modes to determine whether they all meet the requirements of dynamic stability and robustness; if so, the PSS control parameter optimization is terminated, otherwise, the optimization is performed again until all modes meet the requirements; The mathematical model transfer function of PSS is: In the above formula, U pss is the auxiliary modulation signal provided by PSS to the excitation system, K is the gain in the amplification link, T 1 is the time constant of the phase compensation link, T 2 is the time constant of the phase compensation link, T 3 is the time constant of the phase compensation link, T 4 is the time constant of the phase compensation link, T w is the time constant of the DC isolation link, Δω is the speed deviation of the generator, and s is the differential operator; The objective function based on the system eigenvalue and damping ratio is: In the above formula, N p is the number of runs during parameter optimization, N e is the number of eigenvalues ​​and damping ratios, σ 0 is the real part of the ideal eigenvalue obtained in the target expectation, ζ 0 is the ideal damping index expected in the optimization objective, σ i,j is the real part of the ith eigenvalue of the jth operating mode, ζ i,j is the damping ratio of the i-th low-frequency oscillation mode under the j-th operation mode; α and γ are learning factors and α, γ>0; f 1 (x) and f 2 (x) are two fitness functions in neurons, used to calculate the fitness of PSS parameters; Pulse excitation function The expression is as follows: In the above formula, e is the number of operating points; The damping ratio is calculated as follows: the eigenvalue of the system state matrix A is λ=σ±jωσ, the real part of the eigenvalue λ is the damping factor; the imaginary part ω of the eigenvalue λ is the oscillation frequency f=ω / 2π; each pair of eigenvalues ​​corresponds to a low-frequency oscillation mode; the participation factor method is used to identify the low-frequency oscillation mode and calculate the damping ratio as follows: In the above formula, σ i,j is the real part of the ith eigenvalue of the jth operating mode, ζ i,j is the damping ratio of the i-th low-frequency oscillation mode under the j-th operation mode, ω i,j is the angular frequency of the i-th low-frequency oscillation mode in the j-th operating mode.

4. According to claim 1, a method for optimizing parameters of a power system stabilizer based on an NLP system, Its characteristics are: The NLP system optimizes the optimization problem, including: Step 31. Initialize the parameters of the NLP system: including the number of iterations, neurons, spikes, spikes and evolution probabilities; Step 32. Establishing a membrane structure: Establishing a membrane structure containing Z neurons in the skin membrane, and generating an initial spike in each neuron in the membrane, each neuron has H spikes, and each spike W has a set of PSS parameters; applying the QR method to calculate the eigenvalue of the system state equation, and calculating the damping ratio of the power system according to the eigenvalue; Step 33. Calculate the fitness of each peak according to the objective function and save the optimal parameters; Step 34. Use the forgetting rule, the spike rule, and the evolution rule to update the parameters in each neuron; In the spike rule, for any two spikes, if the update probability P0 < Pc, two new spikes are obtained, and the spikes are updated according to the fitness value; in the evolution rule, the worst t spikes evolve according to a certain probability Pm, and the first t spikes with the best fitness are selected; Step 35. Use the transfer rule to transfer the best L spikes of each membrane i in step 32 from the neuron, and bring the best L spikes of the foreign membrane into the neuron i; Step 36. Determine whether the objective function meets the termination condition, reaches the maximum number of iterations or converges to the theoretical optimal value, otherwise return to step 32; Step 37. Output the best peak and optimal PSS parameters from the skin membrane in step 32.

5. According to claim 1, a method for optimizing parameters of a power system stabilizer based on an NLP system, Its characteristics are: The peak rule is as follows: P 0 <P c In the above formula, P c is the probability given by the spike rule, P 0 and r is a random number in (0, 1), o′ i1 is the first parameter to be optimized for the i-th new peak, o′ i2 is the second parameter to be optimized for the i-th new spike, The Nth peak to be optimized for the i-th new peak 0 Parameters, o i1 is the first parameter to be optimized for the ith peak, o j2 is the second parameter to be optimized for the j-th spike, The Nth peak to be optimized for the i-th peak 0 parameters, The Nth peak to be optimized for the jth peak 0 parameters, The Nth value to be optimized for the jth new peak 0 parameters; Update the peak O according to the value of fitness i and O j : The evolution rules are: P 1 <P m R i，evolving :{THE min1 ,THE min2 ,...,THE mint } →{O′ min1 ,O' min2 ,...,O' maxt } In the above formula, P m is the probability O given by the evolutionary rule best is the best peak, P 1 is a (0, 1) random number, O min1 is the first peak with the worst fitness, O min2 is the second peak with the worst fitness, O mint is the tth peak with the worst fitness, O′ min1 is the first peak with the worst fitness after evolution, O min2 is the second peak with the worst fitness after evolution, O′ minj is the jth peak with the worst fitness after evolution; Choose the peak value with the best fit: {O′ new1 ,O' new2 ,...,O' newt } In the above formula: O min1 is the first peak with the worst fitness, O min2 is the second peak with the worst fitness, O mint is the,th peak with the worst fitness, O′ min1 is the first peak with the worst fitness after evolution, O′ min2 is the second peak with the worst fitness after evolution, O′ minj is the jth peak with the worst fitness after evolution, O′ new1 O′ is the first peak with the best fitness selected from all the peaks with poor fitness. new2 O′ is the second best peak in fitness selected from all the peaks with poor fitness. newt is the tth peak with the best fitness selected from all the peaks with poor fitness mentioned above.

6. According to claim 1, a method for optimizing parameters of a power system stabilizer based on an NLP system, Its characteristics are: The transfer rules are: R i，c =R i，c1 and R i，c2 R i,cl :[THE max1 ,THE max2 ,...,THE maxL ] i →[] i THE max1 ,THE max2 ,...,THE maxL R i，c2 :[]I' *max1 ,OR' *max2 ,...,OR' *maxL →[O′ max1 ,OR' max2 ,...,OR' maxL ] i In the above formula, [] i represents membrane i, {O max1 , O max2 , ..., O maxL } is the best L peak in membrane i, {O′ max1 , O′ max2 , ..., O′ maxL } is the best L peak outside membrane i.

7. A power system stabilizer parameter optimization device based on NLP system, Its characteristics are: include: The construction and processing module is used to connect the power system stabilizer PSS to the power system, set different operation modes, build the dynamic model of the power system, and linearize the dynamic model to obtain the state equation of the power system; A construction module is used to calculate the eigenvalues ​​of the state matrix of the state equation of the power system by using the QR method, calculate the system damping ratio according to the eigenvalues, construct an objective function based on the system eigenvalues ​​and the damping ratio, and construct an optimization problem using the objective function with the control parameters of the power system stabilizer PSS as constraints; The optimization module is used to optimize the optimization problem using the NLP system based on the forgetting rule, the spike rule, the evolution rule and the transfer rule adjustment strategy to obtain the optimal control parameter value of the power system stabilizer PSS.

8. The power system stabilizer parameter optimization device based on NLP system according to claim 7, Its characteristics are: The device for optimizing parameters of a power system stabilizer based on an NLP system is used to implement the operating steps of the method for optimizing parameters of a power system stabilizer based on an NLP system as described in any one of claims 1 to 6.

9. A computer device comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, It is characterized in that When the processor executes the computer program, the processor implements the steps of a method for optimizing parameters of a power system stabilizer based on an NLP system as described in any one of claims 1 to 6.

10. A computer storage medium, Its characteristics are: The computer storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the method for optimizing parameters of a power system stabilizer based on an NLP system as described in any one of claims 1 to 6 are implemented.

Images