Search

CN-122022019-A - Relay protection starting point intelligent selection and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm

CN122022019ACN 122022019 ACN122022019 ACN 122022019ACN-122022019-A

Abstract

The invention discloses a relay protection starting point intelligent selection and high-efficiency setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm, which comprises the following steps of 1) constructing a protection coordination importance comprehensive index to quantify the influence of misoperation caused by protection mismatch on a power grid, 2) constructing a starting point optimization and setting calculation model based on minimum mismatch importance and protection adjustment quantity based on the protection coordination importance comprehensive index, and 3) solving a starting point optimization and setting calculation model based on the minimum mismatch importance and the protection adjustment quantity to obtain a relay protection starting point setting result. The method and the device remarkably improve the calculation efficiency of relay protection setting, shorten the time consumption of starting point optimization and setting calculation from the order of hours to the order of tens of seconds through QIRO-DAGA fusion algorithm, and effectively solve the problems of slow response and long period of manual setting.

Inventors

  • DU ZHENFENG
  • YU JUAN
  • HE QIAN
  • Liao Zihe
  • YANG ZHIFANG

Assignees

  • 重庆大学

Dates

Publication Date
20260512
Application Date
20260109

Claims (10)

  1. 1. The intelligent relay protection starting point selecting and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm is characterized by comprising the following steps of: step 1) constructing a protection coordination importance comprehensive index to quantify the influence of misoperation caused by protection mismatch on a power grid; step 2) constructing a starting point optimization and setting calculation model based on minimum mismatch importance and protection adjustment quantity based on protection coordination importance comprehensive indexes; And 3) solving a starting point optimization and setting calculation model based on the minimum mismatch importance and the protection adjustment quantity to obtain a relay protection starting point setting result.
  2. 2. The intelligent relay protection starting point selecting and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm as claimed in claim 1, wherein in step 1), the comprehensive index of protection coordination importance is as follows: (1) (2) In the formula, Respectively the weight coefficients of the indexes; ; Is a branch power index; is a load loss index; is a branch overload index; Is a tide transfer degree index; Is an exponential utility function; Is an input index; to protect the comprehensive index of the matching importance degree.
  3. 3. The intelligent relay protection starting point selecting and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm as claimed in claim 2, wherein the branch power index is characterized in that The method is used for reflecting the power load condition of the branch circuit where the protection coordination relation is located in an initial operation mode, namely: (3) wherein: is the first The sum of the active powers of the branches associated with the protection coordination relations; the total active power of the original system; Load loss index The method is used for representing the load loss degree caused by override action due to mismatch of the protection coordination relationship, namely: (4) wherein: To break off the first The total active power of the system after the branch is associated with the protection coordination relation; branch overload indicator For evaluating the average overload severity of the branches in the post-mismatch system, namely: (5) wherein: Is an overload branch set; And Is that Overload branch in network The size of the active power transmitted and the power quota; index of degree of transfer of tide The method is used for measuring the health degree and stability risk of the system operation after the protection mismatch, namely: (6) (7) wherein: Is a branch set; Is a generator set; And Respectively branch circuits Reactance of (d) and branch active power; Is a branch circuit A power quota; Is an electric generator Maximum active force; 、 、 For the average transmission distance of network power, the average load rate of branches and the average load rate of generators, subscripts And Respectively correspond to the initial positions Network and disconnect 1 st Formed by protection of the fit relationship A network.
  4. 4. The intelligent relay protection starting point selecting and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm according to claim 1, wherein a starting point optimizing and setting calculation model based on minimum mismatch importance and protection adjustment quantity is as follows: (8) wherein: decision variables are starting points; is a comprehensive objective function; As the weight coefficient of the light-emitting diode, For balancing mismatch risk and adjustment cost; The method is a set of all protection coordination relations in the system; is the first Importance comprehensive indexes of the protection coordination relationship; to indicate the function, when the starting point scheme Resulting in the first The value is 1 when the protection coordination relation is not matched, otherwise, 0 is taken; The total number of protections is protected for the system; to indicate the function, when protecting Segment II constant value of (2) Compared with the initial fixed value When the change occurs, the value is 1, otherwise, 0 is taken; Expressed in terms of As input, a complete set of protection fixed value scheme is obtained after constraint processing strategy is executed; representing all feasible tuning scheme sets meeting the actual tuning principles.
  5. 5. The intelligent relay protection starting point selecting and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm as claimed in claim 4, wherein the decision variable is as follows The following is shown: (9) In the first step Column vector Represents the first Pairs of starting points of the ring network, components And (3) with Are integers and respectively represent the initial protection numbers selected by the forward and reverse setting calculation in the ring network.
  6. 6. The method for intelligently selecting and efficiently setting a relay protection starting point based on quantum heuristic recombination optimization and dynamic adaptive genetic algorithm according to claim 4, wherein the step of executing a constraint processing strategy comprises the following steps: Step 1) calculating fixed setting values of all protection according to the power grid parameters and the setting principle, wherein the fixed setting values comprise I-section setting values Sensitivity requirement value of section II Initial fixed value matched with lower-level protection I section ; Step 2) initializing a coordination relation matrix RDR, wherein the dimension of the coordination relation matrix RDR is as follows Elements of Characterization protection And protection The number of the matching sections between the two parts, if protection And protection Is matched with to satisfy Then put into 1, Otherwise set to 0; Step 3) deciding a variable by a starting point To guide and determine the setting sequence, the setting calculation is carried out step by step along the forward and reverse directions of the ring network, if in the calculation process 1, Adopting the matching value with the lower protection I section If (1) Is 0, is adjusted to be matched with the lower protection II section, and adopts the constant value And will 2 Is changed to; step 4) taking the minimum value in all possible matching values as a final II section constant value; Each time a segment II constant of a lower level protection is determined, the associated mating constant of the upper level protection is updated And checking whether the upper-level protection sensitivity requirement is met, if the sensitivity requirement is still not met by the combination of the II sections, then Setting 0, and considering the matching relation mismatch; step 5) based on the final RDR matrix and the complete set of fixed values, calculating the total mismatch importance and the protection adjustment quantity of the system, and calculating a candidate solution If a complete set-value scheme cannot be generated in the process, the candidate solution is determined to be an infeasible solution.
  7. 7. The intelligent relay protection starting point selecting and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm according to claim 1, wherein in the step 3), a starting point optimizing and setting calculation model based on minimum mismatch importance and protection adjustment quantity is solved through a QIRO-DAGA fusion algorithm.
  8. 8. The method for intelligently selecting and efficiently setting a relay protection starting point based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm according to claim 1, wherein in the step 3), the step of solving a starting point optimization and setting calculation model based on the minimum mismatch importance and the protection adjustment quantity comprises the following steps: Step 3.1) inputting power grid parameters and setting rules, and calculating the importance of all protection coordination relations based on a protection coordination relation importance evaluation strategy; Setting algorithm parameters including population scale and maximum iteration times; Step 3.2) executing QIRO global exploration to generate a high-quality population; step 3.3) taking the high-quality population output by QIRO as an initial population, and carrying out local search by utilizing an adaptive evolution mechanism to generate an optimal starting point scheme ; Step 3.4) optimal starting point based scheme And (3) performing setting calculation, and outputting a complete scheme comprising each protection fixed value, action time and coordination relation.
  9. 9. The method for intelligently selecting and efficiently setting a relay protection starting point based on quantum heuristic recombination optimization and dynamic adaptive genetic algorithm according to claim 8, wherein the step 3.2) of executing QIRO global exploration, the step of generating a high-quality population comprises: Step 3.2.1) initializing a quantum population, encoding each protected coordination state into a quantum state, and collapsing each quantum individual to a starting point scheme by quantum measurement ; Wherein, the first Quantum state of each protection coordination relation The following is shown: (10) wherein: the probability amplitude corresponds to the three states; To protect the coordination relationship At the first position Probability of seed status; Starting point scheme In (1) The state of the protection coordination relationship The following is shown: (11) wherein: is the first A protection fit state; Step 3.2.2) executing constraint processing strategy for each starting point scheme, judging feasibility and calculating fitness value ; Step 3.2.3) based on fitness value Updating a quantum state probability amplitude by utilizing quantum rotation gate probability and interference operation; Wherein, for the first The quantum states of the coordination relation are protected, and the updating operation is as follows: (12) For the first The quantum states of the coordination relationship are protected, and the interference operation is as follows: (13) wherein: respectively represent the post-interference and the pre-interference The first of the candidate solutions Probability breadth of a certain state of the protection coordination relation; Is the population scale; is the first And the first and second Random phase angles between candidate solutions, subject to Uniformly distributed on the upper part; is a constant interference factor for controlling the intensity of the interference effect; is the rotation angle; Step 3.2.4) repeating steps 3.2.2) -3.2.3), and terminating and outputting the high-quality population when the algorithm reaches the maximum iteration number.
  10. 10. The method for intelligently selecting and efficiently setting the relay protection starting point based on quantum heuristic recombination optimization and dynamic adaptive genetic algorithm according to claim 8, wherein in step 3.3), the step of performing local search by using an adaptive evolution mechanism comprises the following steps: Step 3.3.1) taking the high-quality population output by QIRO as an initial population; Step 3.3.2) dynamically calculating self-adaptive crossover and mutation probabilities according to a self-adaptive mechanism, and executing selection, crossover and mutation operations, wherein the mutation probabilities introduce importance of protection coordination relations, and preferentially adjust mismatch points with high risks; Crossover probability Probability of variation The dynamic adjustment mode of (a) is as follows: (14) (15) wherein: maximum and average fitness of contemporary population; the fitness is larger in the individuals to be crossed; the fitness of the individual to be mutated; respectively a crossover probability boundary and a mutation probability boundary; First, the The actual probability of variation of the individual protection fit relationships is as follows: (16) wherein: is the first Actual variation probability of the individual protection coordination relationships; is the first The importance of the protection coordination relationship; for adjusting the coefficient; Step 3.3.3) adopting elite retention strategy to ensure that the optimal solution is not lost, starting local search when the algorithm is multi-generation and the optimal solution is not improved, and finally outputting the optimal starting point scheme ; The operation of performing a local search on elite individuals is as follows: (17) wherein: Is a new individual; is elite individual; gaussian disturbance is adopted, and the average value is 0; is a time-varying variance; the current iteration number; Is the total number of iterations.

Description

Relay protection starting point intelligent selection and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm Technical Field The invention relates to the field of power systems and automation thereof, in particular to a relay protection starting point intelligent selection and efficient setting method based on quantum heuristic recombination optimization and dynamic self-adaptive genetic algorithm. Background In recent years, with the large-scale investment of power generation and transformation facilities and the dynamic adjustment of power grid operation lines, the power grid scale is continuously enlarged, the structure is increasingly complex, and the design and setting work of relay protection face greater challenges. In engineering practice, improper protection configuration setting may cause mismatch between devices, i.e., protection mismatch. When a fault occurs on a line where a mismatch point is located, protection override action is easy to be initiated, the power failure range is enlarged, and loss is increased. This problem is particularly pronounced in complex ring structures, because the protection coordination is in a closed loop form, and the fixed value of any protection is closely associated with the selection of the starting point. The traditional starting point selection depends on manual experience, if improper selection is performed, the mismatch risk of the subsequent setting process is aggravated, and the large-range linkage adjustment of the protection constant value can be triggered, so that the setting workload is increased, and new operation uncertainty is introduced. Therefore, the setting strategy capable of intelligently optimizing and selecting the starting point is studied, and the method has important significance for improving the running reliability of the power grid. For protection setting, the industry still relies on manual experience to repeatedly calculate, check and correct. Although the method can deal with conventional scenes by experience, under the complex conditions of frequent switching of lines, temporary overhaul and the like, a plurality of starting points are difficult to quickly and scientifically decide from a global view, so that the method has a long setting period and a slow response, and is difficult to meet the balance requirement of the modern power grid on the setting precision and speed. In terms of starting point optimization selection, the existing research is mainly developed around protection importance evaluation and optimization model construction. At the evaluation level, researchers evaluate protection importance mainly by analyzing the topology where protection is located, such as by using the node density associated with protection, or the consequences of protection failures, such as power flow transfer, etc. Although the method provides a certain basis for the selection of the setting mismatch points, the method is often focused on a single index, the adverse effects caused by initial steady-state power flow and mismatch results cannot be systematically integrated, and the comprehensive evaluation of the importance of the protection coordination relationship is difficult to support. At the model level, the research is based on the integer planning model of the starting point 0-1, and the minimum coordination action time limit, the maximum independent starting point number, the minimum starting point importance and the like are introduced as optimization targets. Although the researches are helpful to avoid overlong time limit of ring network protection setting action and consider the problem of starting point protection matching, the method can not effectively combine the relay protection setting principle to comprehensively stage the global mismatch importance and the original protection adjustment quantity, so that the obtained scheme has limited applicability in actual setting, a complete closed loop of 'starting point selection-setting calculation' is difficult to form, and the result is difficult to be directly applied to engineering practice. Therefore, it is necessary to combine the relay protection setting principle to build a mixed integer programming model aiming at minimizing the mismatch importance and the adjustment quantity. In the aspect of solving algorithm, for the integer programming model with the starting point of 0-1, researchers have applied various intelligent optimization algorithms such as genetic algorithm, particle swarm optimization algorithm, ant colony algorithm and the like to solve. The method has a certain progress in improving the calculation efficiency, but the ring network and the protection quantity are increased rapidly as the power grid scale is enlarged, the algorithms generally face the limitations of difficult super parameter setting, easy sinking into local optimum and the like, and difficult dynamic adaptation to complex