Power transmission section quantum search method, device and system based on topology specific cut set search

Abstract

The invention relates to the technical field of power system optimization scheduling and quantum computing, in particular to a power transmission section quantum search method, device and system based on topology specific cut set search. The key power transmission section searching method based on quantum approximate optimization comprises the following steps: firstly, obtaining a power grid model and a corresponding quantum line model by combining power grid operation data, and performing quantum evolution and measurement on the quantum line model to obtain a quantum state result of a power transmission section; setting a starting value en0 and a final value en2 of a target cut value and a cut value margin em; enabling en1 to traverse a numerical interval [en2, en0] at a set interval, solving a quantum state result for each en1, and obtaining an expected cut value and a corresponding section result; and screening a section result meeting the screening condition as an optimal power transmission section and outputting the optimal power transmission section. According to the method, improvement is carried out on the basis of a quantum approximate optimization algorithm, a multi-stage topology specific cut set search optimization strategy is introduced, the optimal solution of the quantum loss function is approached stage by stage, and the accuracy of a quantum topology search result is improved.

Description

Technical Field

[0001] The present invention relates to the field of power system optimization dispatching and quantum computing technology, and in particular to a transmission section quantum search method, device and system based on topology specific cut set search. Background Art

[0002] As the scale of new power systems continues to expand and the grid architecture becomes more complex, the search for transmission sections of the grid is a critical task. In the real-time monitoring and optimal dispatching of power systems, whether the transmission sections of the grid can be searched quickly and accurately can directly affect the reliability and economy of the power system. The essence of the power system transmission section search problem is to search for a cut set containing a specific number of transmission lines from the grid topology. When dealing with the optimization of transmission sections of large-scale power systems, specific traditional section search algorithms face problems such as high computational complexity, slow convergence, missed sections, and huge search space. Topology-specific cut set search is a typical NP-hard mathematical combinatorial optimization problem and a common mathematical problem in many industries.

[0003] Quantum computing, as a new type of computing method, especially quantum approximate optimization algorithm, has shown its unique advantages in large-scale data storage and parallel computing, and has been proven to have exponential acceleration capabilities for solving traditional combinatorial optimization problems. Some scholars have applied quantum approximate optimization algorithm to the transmission section search problem, but in practical applications, the algorithm cannot accurately search for the transmission section with a specified number of lines, and can only search for the transmission section near the minimum cut set. The probability of the target transmission section with a specified number of transmission lines being observed is low.

[0004] When quantum algorithms deal with large-scale, high-dimensional problems, the depth of the quantum circuit has an important impact on the optimization results. If the depth is too large, it may lead to a waste of quantum computing resources; if the depth is too small, it will be difficult to converge to an accurate solution. Summary of the invention

[0005] In order to overcome the defects of the above-mentioned existing quantum section search technology, such as the difficulty in determining the quantum line depth, unstable convergence effect, and inaccurate section, the present invention proposes a transmission section quantum search method based on topological specific cut set search, which can overcome the problem that the existing quantum search algorithm cannot determine the number of lines in the transmission section, and can make the target transmission section be observed with the maximum probability.

[0006] The present invention proposes a key transmission section search method based on quantum approximate optimization. First, the power grid model and the corresponding quantum circuit model are obtained by combining the power grid operation data, and the quantum circuit model is quantum evolved and measured to obtain the quantum state result of the transmission section. Set the starting value of the target cut value e n0 and terminal value e n2 And the cut-off margin em , e n0 >e n2 ; Let the target cut value e n1 Traverse the value range from large to small with the set interval [e n2 ,e n0 ], for each e n1 , solve the quantum state result, and obtain the expected cut value and the corresponding cross-section result; The section results that meet the screening conditions are selected as the optimal transmission section and output.

[0007] Preferably, the method of using quantum algorithms to solve quantum state results is: combining optimization algorithms, taking minimizing the quantum loss function as the optimization goal, and calculating the expected cut value and the corresponding cross-section result.

[0008] Preferably, the quantum loss function C P (β,γ) is: C P (β,γ)=<ψ P (β,γ)|Hc|ψ P (β,γ)> Where P represents the quantum depth; |ψ P (β,γ)> is the quantum state result, <ψ P (β,γ)| is the quantum state result |ψ P (β,γ)>, β and γ represent the quantum parameters to be optimized; Hc represents the target Hamiltonian for the specific topological cut set search problem.

[0009] Preferably, the optimization algorithm adopts the gradient descent method, which is expressed as follows: (β,γ) t =(β,γ) t-1 -ηg t ; g t =[C p (β,γ) t -C p (β,γ) t-1 ] / [(β,γ) t -(β,γ) t-1 ]; Among them, η is the iteration step size, C p (β,γ) t-1 is the quantum loss function after the t-1th iteration, C p (β,γ) t is the quantum loss function after the tth iteration, (β,γ) t-1 is the quantum parameter to be optimized after the t-1th iteration, (β,γ) t is the quantum parameter to be optimized after the tth iteration; gt is the gradient of the quantum loss function.

[0010] Preferably, the quantum circuit model is constructed in the following manner: the nodes in the power grid model are converted into quantum bits, and an H gate is applied to each quantum bit; the circuits of the power grid model are then converted into quantum parameter circuits in quantum computing, in which each circuit is represented by a CNOT-RZ-CNOT gate, and an RX gate is applied to each quantum bit; a quantum parameter circuit corresponding to the quantum depth is constructed after the H gate; and finally a quantum measurement gate is applied to each quantum bit.

[0011] Preferably, the specific steps are as follows: S1. Obtain quantum depth, combine the power grid model and quantum depth to build a quantum circuit model, perform quantum evolution and measurement on the quantum circuit model, and obtain the quantum state results of the transmission section; S2. Solve the quantum state result to obtain the expected cut value e and the corresponding cross-section result; S31, determine whether e is satisfied n1 -e m ≤e≤e n1 +e m ;e n1 The initial value is e n0 ; If not, go to S32; If yes, let e n1 Update to e n1 -1 and turn to S33; S32, determine whether the number of iterations t satisfies 1≤t<t max ;t max is the set value; If yes, update t to t+1 and return to S2; if no, end; S33, determine whether e is satisfied n1 <e n2 ; If yes, all cross-section results will be output; If not, go to step S31.

[0012] Preferably, when the quantum depth is not determined, the quantum depth in step S1 is iterated from an initial value, and the initial value of the quantum depth is greater than or equal to 1; In step S32, if t>t max , then update the quantum depth p to p+1, and then return to S1; In step S33, if e n1 <e n2 , then determine whether |ee is satisfied n2 |≤e m ; |ee n2 |≤e m , then output the quantum depth p and the cross-section results of each stage; |ee n2 |>e m , then update the quantum depth p to p+1, and then return to S1.

[0013] The present invention proposes a key transmission section search device based on quantum approximate optimization, comprising: Data input module, used to obtain power grid operation data and establish power grid model; A quantum modeling module is connected to the data input module; the quantum modeling module is used to build a quantum circuit model by combining the input quantum depth and the power grid model; The quantum solution module is connected to the quantum modeling module; the quantum solution module is used to execute the key transmission section search method based on quantum approximate optimization to obtain the transmission section that meets the screening conditions.

[0014] The present invention proposes a key power transmission section search system based on quantum approximate optimization, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is connected to the memory, and the processor is used to execute the computer program to implement the key power transmission section search method based on quantum approximate optimization.

[0015] The present invention proposes a storage medium storing a computer program, which, when executed, is used to implement the key power transmission section search method based on quantum approximate optimization.

[0016] The advantages of the present invention are: (1) The transmission section quantum search method based on topological specific cut set search proposed in the present invention is improved on the basis of the quantum approximate optimization algorithm, and a multi-stage topological specific cut set search optimization strategy is introduced to approximate the optimal solution of the quantum loss function in stages, thereby improving the accuracy of the quantum topological search results.

[0017] (2) In the present invention, the target cut-off value and cut-off value margin of each stage can be adjusted according to actual needs, which increases the flexibility of the optimization process and can better adapt to power system transmission section search problems of different complexities.

[0018] (3) In view of the problem that the quantum circuit depth of the existing quantum cross-section search technology is difficult to determine, the present invention proposes an adaptive quantum circuit depth adjustment mechanism, which avoids the waste of quantum computing resources or poor convergence effect caused by using a fixed depth, and gives full play to the quantum evolution capability of the quantum circuit.

[0019] (4) The present invention introduces two convergence criteria. The first convergence criterion en1 -e m ≤e≤e n1 +e m Used to determine whether the target cut value and cut value margin of the current stage are met; the second convergence criterion |ee n2 |≤e m It is used to determine whether the overall optimization goal has been achieved and to ensure the accuracy and effectiveness of the optimization results. The combination of the two convergence criteria can ensure the reliability and accuracy of the search method, and can quickly reach the convergence standard under limited quantum computing resources, which is especially suitable for the trend of expanding the scale and increasing complexity of the new power system grid.

[0020] (5) Based on the quantum approximate optimization algorithm, the present invention controls the variational parameter optimization process and the depth of quantum circuits to search for topological specific cut sets. This method can overcome the problem that the existing quantum search algorithm cannot determine the number of lines in the transmission section, and can make the target transmission section be observed with the maximum probability, reducing the occupation and waste of quantum computing resources, and greatly improving the accuracy and stability of the power system quantum transmission section search algorithm.

[0021] (6) By setting the target cut value and cut value margin, and adaptively adjusting the quantum circuit depth, combined with two key convergence criteria, the optimal solution set for searching the power grid transmission section is ensured under limited quantum computing resources. The present invention can quickly and effectively search the power grid transmission section, improve the real-time monitoring and optimization scheduling capabilities of the power system, and ensure the safe and stable operation of the future power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a flow chart of the power transmission section quantum search method based on topology specific cut set search of the present invention; Figure 2 A detailed flow chart of the power transmission section quantum search method based on topological specific cut set search at a specified quantum depth of the present invention; Figure 3 A detailed flow chart of the power transmission section quantum search method based on topological specific cut set search under undetermined quantum depth of the present invention; Figure 4 Schematic diagram of the quantum topological cut set search model with a quantum depth of P in the present invention; Figure 5 It is the network topology diagram of IEEE14 node system of the present invention; Figure 6 A line graph showing the quantum depth and number of iterations required for the example proposed in the present invention to reach the final target cut value; Figure 7 This is a probability histogram of the specific section calculation results of the quantum section search algorithm mentioned in the implementation example of the present invention. DETAILED DESCRIPTION

[0023] 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.

[0024] Reference Figure 1 , Figure 2 , Figure 4 , a key transmission section search method based on quantum approximate optimization proposed in this embodiment includes the following steps: S1. Combine the grid operation data to obtain the grid model and the corresponding quantum circuit model, perform quantum evolution and measurement on the quantum circuit model, and obtain the quantum state result of the transmission section; like Figure 4 As shown, the quantum circuit model includes 2n+Pm+Pn quantum gates, where n is the number of nodes in the power grid model, m is the number of circuits in the power grid model, and P is the set quantum depth; The construction of the quantum circuit model includes: The first step is to convert the nodes in the power grid model into quantum bits, with the quantum bits corresponding to the nodes in the power grid model one by one. The second step is to apply an H gate to each quantum bit to transform it into a quantum superposition state as the quantum initial state. The third step is to convert the circuit of the power grid model into a quantum parameter circuit in quantum computing. In the quantum parameter circuit, each circuit is represented by a CNOT-RZ-CNOT gate, and each quantum bit is acted on by an RX gate; thus, a quantum parameter circuit consisting of m CNOT-RZ-CNOT gates and n RX gates is obtained, and P quantum parameter circuits are repeatedly constructed to generate Pm CNOT-RZ-CNOT gates and Pn RX gates; In the fourth step, each qubit acts as a quantum measurement gate, thus completing the quantum circuit model containing n H gates, Pm CNOT-RZ-CNOT gates, Pn RX gates and n quantum measurement gates, as shown in the following figure. Figure 4 shown.

[0025] The quantum circuit model is measured through a quantum measurement gate, and the resulting quantum state is recorded as |ψ P (β,γ)>, where P is the quantum depth, β is the quantum parameter to be optimized in the RX gate, and γ is the quantum parameter to be optimized in the RZ gate; S2. The measured quantum state result of the transmission section |ψ P(β,γ)>, combined with classical optimization algorithms to minimize the quantum loss function C p (β,γ) is the optimization target, the expected cut value e is calculated, and the nodes in the quantum power grid model are divided into set A and set B; Classic optimization algorithms include Adam optimization algorithm (Adam), Newton method or gradient descent method.

[0026] Quantum loss function C p The calculation formula (1) of (β,γ) is as follows: (1); Where: P represents quantum depth; <ψ P (β,γ)| is the quantum state result |ψ P (β,γ)>, i.e., the conjugate transpose; β and γ represent the quantum parameters to be optimized; Hc represents the target Hamiltonian of the specific topological cut set search problem; E is the line set of the power grid model, e rj is the connection line between node r and node j; σ r Z represents the Pauli Z-matrix operator corresponding to node r, σ j Z Represents the Pauli Z-matrix operator corresponding to node j.

[0027] To facilitate the explanation of the meaning of the expression, the following simplified operations are performed as shown in formulas (2) and (3): (2); (3); The concise formula (4) derived from formulas (1), (2) and (3) is as follows: (4); Where: <ψ P (β,γ)|is|ψ P (β,γ)> is the right vector, i.e. the conjugate transpose; m is the number of lines; C rj To represent the node r and nodes j The binary value of the distribution set relation between ; C rj =1, indicating a node r and nodes j are assigned to the same set; C rj =-1, indicating that nodes r and j are assigned to different sets; after the iteration of each node assignment is completed, the nodes assigned to 1 are assigned to set A, and the nodes assigned to -1 are assigned to set B; e is the expected cut value of the solution; It should be noted that the expected cut value e is the topological cut value of the power network. The expected size is minimized using the classical optimization algorithm to meet the convergence criterion.

[0028] In this embodiment, the calculation formula (5) for tuning the quantum parameters β and γ using the classical optimization algorithm of gradient descent is as follows: (β,γ) t =(β,γ) t-1 -ηg t (5); Where: t is the number of iterations, the initial value is t=1, and the maximum number of iterations is t max , t max The size can be set by yourself; η is the iteration step; g t is the gradient of the quantum loss function. When the power network is too complex and the number of quantum gates in the quantum circuit is too large, the quantum loss function C p The gradient g of (β,γ) t The solution will be very difficult, so the difference formula (6) can be used to simply calculate the gradient size.

[0029] g t =[C p (β,γ) t -C p (β,γ) t-1 ] / [(β,γ) t -(β,γ) t-1 ] (6); Among them, C p (β,γ) t-1 is the quantum loss function after the t-1th iteration, C p (β,γ) t is the quantum loss function after the tth iteration, (β,γ) t-1 is the quantum parameter to be optimized after the t-1th iteration, (β,γ) t is the quantum parameter to be optimized after the tth iteration. Specifically, β is the quantum parameter to be optimized in the RX gate, and γ is the quantum parameter to be optimized in the RZ gate.

[0030] S3, set the starting value of the target cut value e n0 and terminal value e n2 And the cut-off margin e m , e n0 >e n2 ; Let e n1 Traverse the value range from large to small with the set interval [e n2 ,e n0 ], loop step S2, list the cross-section results corresponding to each expected cut value e; the cross-section result is essentially a topological cut set, that is, dividing the nodes into set A and set B.

[0031] specific,e n0 、e n2 and e m All can be set manually, and can be set based on experience and task requirements, such as e n0 It can be specifically set as the maximum number of connected lines on the transmission section, e n2 It can be specifically set as the minimum number of connected lines on the transmission section, e m It can take values ​​in the interval (0,1).

[0032] S4, selecting the section results that meet the selection conditions as the optimal transmission section; The screening conditions are the set electrical partition properties, for example: (1) the number of nodes in the two electrical partitions must be greater than or equal to the set first threshold, and the first threshold is taken from the number set {3, 4, 5}; (2) The number of interconnecting lines between electrical partitions must be less than or equal to a set second threshold value, and the second threshold value is taken from the number set {3, 4, 5}; (3) The two electrical partitions are internally connected; (4) The connection lines between electrical partitions have consistent directions.

[0033] Specifically, when the P value is determined, step S3 includes the following steps.

[0034] S31, determine whether e is satisfied n1 -e m ≤e≤e n1 +e m ; Among them, e n1 is the target cut value, e n2 is the final value of the target cut value, e m is the cut-off margin; e n2 and e m All are set constants, e n1 The initial value is the set constant e n0 , e n0 Greater than e n2 ; If not, go to step S32; If yes, let e n1 Update to e n1 -1, and go to step S33; S32, determine whether the number of iterations t satisfies 1≤t<t max ;t max is the maximum number of iterations set; If yes, update t to t+1, and go to step S2 to recalculate the expected cut value e; If not, the output quantum depth is insufficient; S33, determine whether e is satisfied n1 <e n2 ; If yes, then all cross-section results are output, that is, the node segmentation set corresponding to each expected cut value e; If not, go to step S31; Reference Figure 3 , when the quantum depth P is not determined, the quantum depth can also be iterated at the same time to determine the quantum depth P and search for the quantum cross section at the same time. At this time, the quantum depth p can be iterated starting from 1; that is: In step S1, a quantum circuit model is constructed based on the quantum depth p, which includes n H gates, pm CNOT-RZ-CNOT gates, pn RX gates and n quantum measurement gates; Step S3 specifically includes the following steps: S31', determine whether e is satisfied n1 -e m ≤e≤e n1 +e m ; Among them, e n1 is the target cut value, e n2 is the final value of the target cut value, e m is the cut-off margin; e n2 and e m All are set constants, e n1 The initial value is the set constant e n0 , e n0 Greater than e n2 ; If not, go to step S32'; If yes, let e n1 Update to e n1 -1, and go to step S33'; S32', determine whether the number of iterations t satisfies 1≤t<t max ; If yes, update t to t+1, and go to step S2 to recalculate the expected cut value e; If not, update the quantum depth p to p+1, and then return to step S1 to rebuild the quantum circuit model; S33', determine whether e is satisfied n1 <e n2 ; If yes, go to step S34'; If not, go to step S31'; S34', determine whether it satisfies |ee n2 |≤e m ; If yes, then record the quantum depth p as the quantum depth target value P, and output the cross-section results of each stage; If not, update the quantum depth p to p+1, and then return to step S1 to rebuild the quantum circuit model.

[0035] Thus, when searching for a transmission section, if the quantum depth target value P is determined, the following Figure 1 Search for transmission sections; if the quantum depth target value P is not determined, follow the following Figure 2 Search for transmission sections.

[0036] It is worth noting that the implementation Figure 2 When searching for the transmission section shown, the section result output in step S34' may only include the section result corresponding to the e value solved under the last p value; or may include the section results corresponding to all e values ​​generated during the p value iteration process.

[0037] The above-mentioned transmission section quantum search method based on topology specific cut set search is verified in combination with specific embodiments below.

[0038] In this embodiment, the IEEE 14-node system is used for verification. The network topology diagram of the IEEE 14-node system is as follows: Figure 5 As shown, it includes 14 nodes and 20 lines; the nodes include generator nodes and load nodes, and the lines include transmission lines and transformer branches.

[0039] In this embodiment, the initial value of quantum depth P is set to 1. In the subsequent iteration process, the quantum depth value will increase adaptively with the convergence criterion. The maximum number of optimized iterations t max is 200 times; the cut value margin e m is 0.2; target cut value e n1 The initial value of e n0 Set to 4; the final value of the target cut value is e n2 Set to 1; set the quantum topological cut set results for each quantum measurement sampling to 10,000; the first threshold and the second threshold are both set to 4.

[0040] In this embodiment, implementation Figure 3 The method shown in the figure shows that the quantum depth iteration reaches 6 and achieves |ee for the first time. n2 |≤e m In this process, the line graph of quantum depth and number of iterations is as follows: Figure 6 As shown; it can be seen that when the quantum depth p<6, the algorithm cannot converge; this is because the lower quantum depth cannot provide enough gradient information, the quantum evolution is too simple, and it cannot reach the complexity required for the target expectation, resulting in the algorithm failing to converge, so it is necessary to increase the p value.

[0041] When p=6, Figure 6It can be seen that the number of required iterations changes and no longer stays at the maximum number of iterations. Only 80 iterations are needed to complete the quantum evolution process. When p is greater than 6, although the evolution process can be successfully completed, the number of quantum gates in the quantum circuit model will increase, resulting in higher computing costs and more complex circuit design.

[0042] In this embodiment, the quantum depth target value P is set to 6, and then the Figure 2 The method shown. At this time, the number of quantum gates required for the entire quantum circuit model is 2n+Pm+Pn=2×14+6×20+6×14=232. Quantum bit string |ψ P After solving (β,γ)>, all cross-section results are shown in Table 1.

[0043] Table 1 Number of quantum transmission section results at each stage ;

[0044] In this embodiment, 10,000 quantum state results are sampled in each stage, and the probability bar graph of the calculation results of a specific section is as follows: Figure 7 As shown, the first stage (e n1 =4) out of the 10,000 quantum state results searched, 4,426 satisfy the cross-section value of 4. The algorithm proposed in the present invention is based on two convergence criteria (e n1 -e m ≤e≤e n1 +e m and|ee n2 |≤e m ) Repeated searches can greatly increase the probability of effective cross-section results at each stage and improve search accuracy.

[0045] In this embodiment, by screening conditions (ie, electrical partition properties), partitioning methods that meet the conditions are obtained as shown in Table 2.

[0046] The electrical partitioning properties set in this embodiment are: "The number of nodes within the two electrical partitions must be greater than or equal to the first threshold 4 and the number of communication lines between the electrical partitions must be less than or equal to the second threshold 4" and "The two electrical partitions are internally connected and the communication lines between the electrical partitions have directional consistency."

[0047] Table 2 Cross-sectional results of IEEE14-bus system and grid partitioning ;

[0048] In Table 2, l i-j Represents the connection line between node i and node j.

[0049] Of course, it is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, but also includes the same or similar structures that can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-limiting, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any reference numerals in the claims should not be regarded as limiting the claims involved.

[0050] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

[0051] The techniques, shapes, and structural parts not described in detail in the present invention are all well-known techniques.

Claims

1. A key transmission section search method based on quantum approximate optimization, characterized in that: First, the power grid model is obtained by combining the power grid operation data, and the quantum circuit model corresponding to the power grid model is constructed. The quantum circuit model is quantum evolved and measured to obtain the quantum state results of the transmission section. Set the starting value of the target cut value e n0 and terminal value e n2 And the cut-off margin e m , e n0 >e n2 ; Let the target cut value e n1 Traverse the value range from large to small with the set interval [e n2 ,e n0 ], for each e n1 , solve the quantum state results to obtain the expected cut value and the corresponding cross-section results; The section results that meet the screening conditions are selected as the optimal transmission section and output.

2. The key transmission section search method based on quantum approximate optimization according to claim 1 is characterized in that: The method of using quantum algorithms to solve quantum state results is: combining optimization algorithms, taking minimizing the quantum loss function as the optimization goal, and calculating the expected cut value and the corresponding cross-section results.

3. The key transmission section search method based on quantum approximate optimization as claimed in claim 2 is characterized in that quantum Loss function C P (β,γ) is: C P (b,c)=<ψ P (b,c)|Hc|ψ P (b,c)>; Where P represents the quantum depth; |ψ P (β,γ)> is the quantum state result, <ψ P (β,γ)| is the quantum state result |ψ P (β,γ)>, β and γ represent the quantum parameters to be optimized; Hc represents the target Hamiltonian for the specific topological cut set search problem.

4. The key transmission section search method based on quantum approximate optimization according to claim 2 is characterized in that: The optimization algorithm uses the gradient descent method, and the formula is expressed as: (b,c) t =(b,c) t-1 -ηg t ; g t =[C p (b,c) t -C p (b,c) t-1 ] / [(b,c) t -(b,c) t-1 ]; Among them, η is the iteration step size, C p (β,γ) t-1 is the quantum loss function after the t-1th iteration, C p (β,γ) t is the quantum loss function after the tth iteration, (β,γ) t-1 is the quantum parameter to be optimized after the t-1th iteration, (β,γ) t is the quantum parameter to be optimized after the tth iteration; g t is the gradient of the quantum loss function.

5. The key transmission section search method based on quantum approximate optimization according to claim 1 is characterized in that: The quantum circuit model is constructed in the following way: the nodes in the power grid model are converted into quantum bits, and an H gate is applied to each quantum bit; the circuits of the power grid model are then converted into quantum parameter circuits in quantum computing. In the quantum parameter circuits, each circuit is represented by a CNOT-RZ-CNOT gate, and then an RX gate is applied to each quantum bit; a quantum parameter circuit corresponding to the quantum depth is constructed after the H gate; and finally, a quantum measurement gate is applied to each quantum bit.

6. The key transmission section search method based on quantum approximate optimization according to any one of claims 1 to 5, characterized in that: The specific steps are as follows: S1. Obtain quantum depth, combine the power grid model and quantum depth to build a quantum circuit model, perform quantum evolution and measurement on the quantum circuit model, and obtain the quantum state results of the transmission section; S2. Solve the quantum state result to obtain the expected cut value e and the corresponding cross-section result; S31, determine whether e is satisfied n1 -e m ≤e≤e n1 +e m ;e n1 The initial value is e n0 ; If not, go to S32; If yes, let e n1 Update to e n1 -1 and turn to S33; S32, determine whether the number of iterations t satisfies 1≤t<t max ;t max is the set value; If yes, update t to t+1 and return to S2; if no, end; S33, determine whether e is satisfied n1 <e n2 ; If yes, all cross-section results will be output; If not, go to step S31.

7. The key transmission section search method based on quantum approximate optimization according to claim 6 is characterized in that: When the quantum depth is not determined, the quantum depth is iterated from an initial value in step S1, and the initial value of the quantum depth is greater than or equal to 1; In step S32, if t>t max , then update the quantum depth p to p+1, and then return to S1; In step S33, if e n1 <e n2 , then determine whether |ee is satisfied n2 |≤e m ; |ee n2 |≤e m , then output the quantum depth p and the cross-section results of each stage; |ee n2 |>e m , then update the quantum depth p to p+1, and then return to S1.

8. A key transmission section search device based on quantum approximate optimization, comprising: Data input module, used to obtain power grid operation data and establish power grid model; A quantum modeling module, connected to the data input module; The quantum modeling module is used to build a quantum circuit model by combining the input quantum depth and power grid model; A quantum solution module is connected to the quantum modeling module; the quantum solution module is used to execute the key transmission section search method based on quantum approximate optimization as described in any one of claims 1 to 7 to obtain a transmission section that meets the screening conditions.

9. A key transmission section search system based on quantum approximate optimization, characterized in that: It includes a memory and a processor, the memory stores a computer program, the processor is connected to the memory, and the processor is used to execute the computer program to implement the key transmission section search method based on quantum approximate optimization as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: A computer program is stored, and when the computer program is executed, it is used to implement the key power transmission section search method based on quantum approximate optimization as described in any one of claims 1 to 7.

Images