Power grid relay protection constant value optimization method based on quantum genetic algorithm

Abstract

The invention relates to a power grid relay protection constant value optimization method based on a quantum genetic algorithm. The method comprises the following steps: constructing a setting optimization model by taking the total action time of all main protection devices of a power grid for coping and processing a near-zone fault or a far-zone fault as a target function; constructing a quantum genetic algorithm model to carry out iterative optimization on the setting optimization model, designing a quantum bit representation method according to quantum state characteristics, and realizing improvement of coverage and convergence performance through linear superposition of different state probabilities of each analytic solution; an operation process is established, dynamic iteration is carried out on a quantum bit chromosome population and a two-state solution based on a quantum gate, the problems that a traditional algorithm is complex in operation process and low in calculation efficiency are solved, and therefore the safety and reliability of a relay protection system are further promoted.

Description

Technical Field

[0001] The present invention belongs to the field of power system relay protection, and specifically relates to a distributed power grid relay protection setting optimization method based on a quantum genetic analysis algorithm. Background Art

[0002] As the first line of defense for the safety and stability of the power grid, relay protection is responsible for the task of quickly and sensitively removing faults. Relay devices based on different principles in the protection system can be used to detect abnormal operation of the system and quickly isolate only the fault location from other normal operating areas. One of the key guarantees for this goal is to ensure that the relay device operates in the correct manner and coordinates with other devices. Therefore, exploring the setting values ​​of each relay device in the protection system to ensure the sensitivity, selectivity, stability and speed of the system protection is the focus of coordinating the actions of each device. However, the calculation process of the existing power grid relay protection setting algorithm is complex and the calculation efficiency is low. Summary of the invention

[0003] In order to solve the problems existing in the prior art, the present invention provides a distributed optimization method for relay protection settings of a power system based on a quantum genetic analysis algorithm.

[0004] In order to achieve the above object, the technical solution proposed by the present invention is: a method for optimizing the setting value of power grid relay protection based on quantum genetic algorithm, specifically:

[0005] The total action time of the relay protection device to deal with and handle faults is calculated based on the time dial setting value and interpolation setting value of all relay protection devices in the power grid, and a setting optimization model is constructed with the minimum total action time as the objective function;

[0006] A quantum genetic algorithm is constructed to iteratively optimize the tuning optimization model, specifically:

[0007] The different value combinations of the time dial setting value and the interpolation setting value of the relay protection device are represented by the quantum bit representation method to form the quantum bit chromosome population Q(t);

[0008] Initialize the qubit chromosome population Q(t);

[0009] Solve a set of two-state solutions P(t) based on Q(t); evaluate P(t) and store the current optimal solution;

[0010] Generate P(t) through the previous time series Q(t-1) of the qubit chromosome population;

[0011] Update Q(t) based on quantum gate U(t), and continue to iteratively solve and evaluate the two-state solution P(t) until the termination condition is reached and output the current optimal solution.

[0012] A further design of the above technical solution is: the total action time of the relay protection device to respond to and handle faults is:

[0013]

[0014] Among them, a, b, c are constant coefficients; I f is the fault current at the primary port of the current transformer; CT P is the primary setting value at the current transformer; TDS is the time dial setting value, and PS is the interpolation setting value.

[0015] The constant coefficients a, b, and c are 0.14, 0.02, and 1.0, respectively.

[0016] The objective function of the tuning optimization model is:

[0017]

[0018] in,

[0019]

[0020] T b -T P ≥CTI

[0021] T i p-cl and T i p-fa T represents the operating time of the main protection relay for local fault and the operating time of the main protection relay for remote fault respectively; b ,T p are the action time of the backup protection relay and the main protection relay respectively, CTI is the protection coordination time interval, i is the fault number of the nearby area in the system, and j is the fault number of the far area in the system.

[0022] The t-th generation qubit chromosome population Q(t) is:

[0023]

[0024] When initializing Q(t), all chromosomes in Q(t) are Default is

[0025] Compared with the prior art, the present invention has the following advantages:

[0026] The QGA algorithm of the present invention solves the global optimization problem of power grid relay protection setting value through steps such as quantum bit representation, generation and evaluation of two-state solutions, and quantum gate update. QGA optimizes the overall protection device action time by minimizing the objective function, which comprehensively considers the action time of all relay protection devices.

[0027] The population size in QGA, i.e., the number of qubit chromosomes, always remains constant. Therefore, QGA using qubit representation has better diversity and convergence than existing GA algorithms.

[0028] The QGA algorithm of the present invention optimizes the action time and accuracy of the relay protection device by constructing a global setting optimization model and combining the characteristics of near-field and far-field faults. This is different from the traditional relay protection setting method, which usually relies on step-by-step coordination setting. The quantum genetic model uses a global perspective for optimization, which improves the efficiency and accuracy of the setting.

[0029] The QGA algorithm of the present invention uses a quantum bit representation method to reduce the number of optimized chromosomes, thereby reducing the computational complexity. By reducing the number of necessary optimized chromosomes through the linear superposition of different state probabilities of each optimization solution of the quantum state, QGA can achieve better convergence and diversity, and improve the ability of global optimization. This method has a significant difference in efficiency when dealing with complex optimization problems compared to traditional tuning models. At the same time, quantum gates are used to implement adaptive iterations for the quantum bit chromosome population and the two-state solution, which greatly improves efficiency in the global optimization process. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flow chart of a method for optimizing relay protection settings of a power grid based on a genetic analysis algorithm provided by the present invention;

[0031] Figure 2 It is a comparison chart of optimization convergence curves. DETAILED DESCRIPTION

[0032] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0033] Embodiment 1

[0034] To facilitate understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings.

[0035] Step 1: Calculate the total action time of the relay protection device to respond to and handle faults according to the time dial setting value and interpolation setting value of all relay protection devices in the power grid, and build a setting optimization model with the minimum total action time as the objective function;

[0036] Step 1.1: The relay device action time is related to the fault current, which is determined by the relay type and contact current, and can be solved by the corresponding numerical characteristic curve. Since the total protection action time is generally nonlinear, its mathematical expression is set as follows:

[0037]

[0038] Where a, b, and c are constant coefficients, which can be taken as 0.14, 0.02, and 1.0 according to the IEEE standard; I f is the fault current at the primary port of the current transformer; CT P is the primary setting value at CT; TDS is the time dial setting value, and PS is the interpolation setting value;

[0039] Step 1.2: Establish the objective function and constraints. In order to achieve global protection, in the process of setting and optimizing each device, the total action time of all main protection devices to deal with and handle local faults or remote faults is used as the objective function. The constructed objective function is as follows:

[0040]

[0041] in,

[0042]

[0043] T b -T P ≥CTI

[0044] T b ,T p are the action time of the backup protection relay and the main protection relay respectively, CTI is the protection coordination time interval, i is the fault number of the nearby area in the system, and j is the fault number of the far area in the system.

[0045] In this embodiment, the setting values ​​of the protection device include TDS and PS, as shown in Table 1.

[0046] Table 1 Setting values ​​of protection devices

[0047]

[0048] Step 2: Construct a quantum genetic model;

[0049] Step 2.1: QGA modeling. In evolutionary algorithms, the optimization scheme is encoded by some different representations and stored on chromosomes. This embodiment designs a representation based on the concept of quantum bits. A quantum bit is defined by a pair of complex numbers and can be written as (α, β) T , define a quantum bit space as follows:

[0050]

[0051] Where |α i | 2 +|β i | 2=1,i=1,2,...,m; α is the probability amplitude of the qubit being in state "0", β is the probability amplitude of the qubit being in state "1", ensuring that the total probability of the quantum state is 1

[0052] The advantage of this expression is that it can cover superposition states other than "1" and "0". For example, for a three-qubit system with three states, its probability amplitude can be written as:

[0053]

[0054] Then the system state can be expressed as:

[0055]

[0056] The three-qubit system contains information about eight states, namely, (000), (001), (010), (011), (100), (101), (110), and (111), with probabilities of 1 / 16, 3 / 16, 1 / 16, 3 / 16 / , 1 / 16, 3 / 16 / , 1 / 16, and 3 / 16, respectively. Compared with other representations, the qubit-based representation has better performance in dealing with multi-state scenarios.

[0057] Step 3: Use the quantum genetic model to iteratively optimize the tuning optimization model;

[0058] Step 3.1: In this embodiment, QGA adds a probability analysis model based on the genetic algorithm. QGA usually covers a qubit chromosome population, and the population at the tth generation can be expressed as:

[0059]

[0060] In this embodiment, the quantum bit corresponds to different value combinations of the time dial setting value (TDS) and the interpolation setting value (PS) in the setting optimization model. Through the superposition of the quantum bit, one quantum bit can simultaneously represent multiple different value combinations.

[0061] Step 3.2: In the step of initializing Q(t), all Q(t) are preset as Right now Refers to a qubit chromosome that linearly superimposes all possible states with the same probability, and its state can be written as:

[0062]

[0063] In the formula, S k is a binary string x1,x2,…,x m Represents the kth state.

[0064] The state represents the initial state of the quantum bit chromosome. Its function is to initialize the population and provide a wide search space for subsequent optimization and iteration. The equal superposition state after initialization ensures the diversity in the initial stage and avoids falling into the local optimum, thereby improving the ability of global optimization.

[0065] Step 3.3: Find a set of two-state solutions based on Q(t) The binary solution corresponds to a specific combination of TDS and PS values, that is, the specific setting of the relay protection device. is a two-state string of length m. Using the probability of each qubit, that is The status is |a| 2 or |β| 2 . Each solution All can be used to evaluate their fitness, and then the best solution is selected and stored in the two-state solution P(t).

[0066] Step 3.4: During the iteration process, the two-state solution P(t) is solved based on the state of the previous time series Q(t-1) of the qubit chromosome population, and each of the two-state solutions is evaluated to obtain the fitness value, that is, these solutions are substituted into the objective function to calculate the action time of its protection device; next, each group of qubit chromosomes Q(t) is updated by constructing a quantum gate U(t). U(t) is formed by the two-state solution P(t) and the optimal storage solution, which can be designed according to the needs, and its mathematical expression is as follows:

[0067]

[0068] θ is the rotation angle in the quantum genetic algorithm (QGA) and is used to update the quantum gate. θ represents the rotation angle when the qubit chromosome is updated through the quantum gate, which determines the change in the state of the qubit chromosome.

[0069] In the iteration step, the best solution in a set of two-state solutions P(t) will be stored and compared with the next set of solutions in the next cycle, and the better solution will be retained. The iteration step will be repeated until the iteration termination condition is reached and the best solution is obtained.

[0070] In this embodiment, the quantum chromosome represents a fixed value combination of multiple protection devices, and the quantum gate is used to update these solutions so that they gradually approach the global optimal solution. The fitness value is used to evaluate the performance of a solution in minimizing the objective function (total action time). In the setting optimization model, the fitness value is calculated based on the objective function (minimization of action time). The quantum gate is used to update the quantum chromosome and adjust the probability of the quantum bit so that the solution gradually converges to the optimal fixed value combination. The quantum gate updates the state of the quantum bit, thereby affecting the binary solution generated in each round of iteration. The quantum gate indirectly affects the values ​​of TDS and PS by adjusting the probability of the quantum bit, thereby adjusting the minimization process of the action time in the objective function.

[0071] from Figure 2 It can be concluded that in terms of convergence speed, QGA has certain advantages over the standard GA method, and the number of iterations is reduced by 37.85%. Secondly, the QGA model is compared with the particle swarm algorithm, differential evolution algorithm and BP neural network in terms of operating efficiency. The comparison results based on the same example are shown in Table 2. It can be seen from the table that the QGA algorithm of this embodiment has higher computational efficiency and can improve the comprehensiveness and convergence efficiency of the optimization process. This dynamic optimization mechanism is in contrast to the static setting method of the traditional model, showing its advantages in adapting to complex power grid fault environments. Through the above-mentioned global optimization example results, it is verified that the proposed method can take into account the requirements of sensitivity, selectivity, stability and speed in the constant value optimization process, while improving the optimization efficiency.

[0072] Table 2 Setting values ​​of protection devices

[0073]

[0074] The above embodiments are only preferred specific implementations of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for optimizing relay protection settings of a power grid based on a quantum genetic algorithm, characterized in that: The total action time of the relay protection device to deal with and handle faults is calculated based on the time dial setting value and interpolation setting value of all relay protection devices in the power grid, and a setting optimization model is constructed with the minimum total action time as the objective function; A quantum genetic algorithm is constructed to iteratively optimize the tuning optimization model, specifically: The different value combinations of the time dial setting value and the interpolation setting value of the relay protection device are represented by the quantum bit representation method to form the quantum bit chromosome population Q(t); Initialize the qubit chromosome population Q(t); Solve a set of two-state solutions P(t) based on Q(t); evaluate P(t) and store the current optimal solution; Generate P(t) through the previous time series Q(t-1) of the qubit chromosome population; Update Q(t) based on quantum gate U(t), and continue to iteratively solve and evaluate the two-state solution P(t) until the termination condition is reached and output the current optimal solution.

2. The method for optimizing the setting value of power grid relay protection based on quantum genetic algorithm according to claim 1 is characterized in that: The total action time of the relay protection device to respond to and handle faults is: Among them, a, b, c are constant coefficients; I f is the fault current at the primary port of the current transformer; CT P is the primary setting value at the current transformer; TDS is the time dial setting value, and PS is the interpolation setting value.

3. The method for optimizing the setting value of power grid relay protection based on quantum genetic algorithm according to claim 2 is characterized in that: The constant coefficients a, b, and c are 0.14, 0.02, and 1.0, respectively.

4. The method for optimizing the setting value of power grid relay protection based on quantum genetic algorithm according to claim 3 is characterized in that: The objective function of the tuning optimization model is: in, T b -T P ≥CTI T i p-cl and T i p-fa T represents the operating time of the main protection relay for local fault and the operating time of the main protection relay for remote fault respectively; b ,T p are the action time of the backup protection relay and the main protection relay respectively, CTI is the protection coordination time interval, i is the fault number of the near zone in the system, and j is the fault number of the far zone in the system.

5. The method for optimizing the setting value of relay protection of power grid based on quantum genetic algorithm according to claim 1 is characterized in that: The t-th generation qubit chromosome population Q(t) is: When initializing Q(t), all chromosomes in Q(t) are The default value is 1 / √2.

Images