Signal analysis method, system, device and medium based on solution prediction of a quadratic constraint quadratic programming problem

Abstract

The application discloses a signal analysis method, system, device and medium based on quadratic constraint quadratic programming problem solution prediction, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring a target programming problem of a power transmission signal, wherein the target programming problem comprises a quadratic objective function and a quadratic constraint condition; encoding the quadratic objective function and the quadratic constraint condition to generate a target tripartite graph of the target programming problem; inputting the target tripartite graph into a message passing model; analyzing a plurality of nodes in the target tripartite graph and feature edges between the plurality of nodes by using the message passing model to obtain problem attributes of the target programming problem output by the message passing model, wherein the problem attributes comprise feasibility, an objective value and an optimal solution; and performing signal analysis on the power transmission signal based on the problem attributes. The method can process quadratic constraint quadratic programming problems in real time, and improves the analysis efficiency of the power transmission signal.

Description

Technical Field

[0001] This application belongs to the field of artificial intelligence technology, and in particular relates to a signal analysis method, system, device and medium based on the prediction of solutions to quadratic programming problems with quadratic constraints. Background Technology

[0002] In the process of power transmission signal analysis in the power grid, the Quadratically Constrained Quadratic Programming (QCQP) problem often appears in optimization and control problems, especially in power system optimization, signal processing and power dispatch, such as power system state estimation, current optimization or voltage control.

[0003] Quadratic programming (QP) is a class of optimization problems whose objective function consists of a combination of quadratic and linear terms, and whose constraints are typically linear inequalities or equations. QCQP is an extension of quadratic programming, where not only is the objective function quadratic, but the constraints also contain quadratic terms.

[0004] Solving quadratic programming (QP) problems, especially those with quadratic constraints, presents significant challenges. Traditional algorithms typically involve computationally intensive processes, exhibiting clear limitations in handling large-scale problems and meeting real-time computing requirements. For example, the interior-point method, which iteratively approximates the optimal solution within feasible solutions, requires solving a series of linear equations for a typical quadratic programming problem with n variables, relying on matrix factorization operations, resulting in a time complexity of O(n^2). 3 Therefore, this algorithm incurs a huge computational burden in large-scale problems. Similarly, the active-set algorithm solves the problem step by step by iteratively adjusting the active constraints in the constraint set, but as the number of constraints and variables increases, the computational cost of this algorithm also increases significantly, affecting the efficiency of transmission signal analysis.

[0005] Therefore, there is an urgent need for a method for analyzing power grid signals to handle the quadratic constraint quadratic programming problem in the signal analysis process in real time and improve the efficiency of power transmission signal analysis. Summary of the Invention

[0006] This application aims to address at least one of the technical problems existing in the prior art. To this end, this application proposes a signal analysis method, system, device, and medium based on the prediction of solutions to a quadratic constraint quadratic programming problem, which can process quadratic constraint quadratic programming problems in real time and improve the analysis efficiency of power transmission signals.

[0007] Firstly, this application provides a signal analysis method based on the prediction of solutions to a quadratic constrained quadratic programming problem, the method comprising:

[0008] The objective planning problem for acquiring power transmission signals includes a quadratic objective function and quadratic constraints.

[0009] Encode the quadratic objective function and the quadratic constraints to generate a target triangulation graph of the objective programming problem;

[0010] The target trigonometric graph is input into a message passing model. The message passing model analyzes multiple nodes in the target trigonometric graph and the feature edges between the multiple nodes to obtain the problem attributes of the target planning problem output by the message passing model. The problem attributes include feasibility, target value, and optimal solution.

[0011] Based on the aforementioned problem attributes, signal analysis is performed on the power transmission signal.

[0012] According to one embodiment of this application, encoding the quadratic objective function and the quadratic constraints to generate a target triangulation graph of the objective programming problem includes:

[0013] The optimization quantities are extracted from the quadratic objective function and the quadratic constraints to form the multiple nodes, and the coefficient terms are extracted as the feature edges;

[0014] The target triangulation is constructed based on the correspondence between the multiple nodes and the feature edges.

[0015] According to one embodiment of this application, the plurality of nodes includes variable nodes, quadratic term nodes, and constraint nodes, and the feature edges include a first feature edge between the variable node and the quadratic term node, a second feature edge between the variable node and the constraint node, and a third feature edge between the quadratic term node and the constraint node.

[0016] According to one embodiment of this application, the goal planning problem includes:

[0017]

[0018] Where, x∈R n The decision variables are a symmetric matrix Q, a vector p, and a constant b. i The problem provides a set of coefficients. b i ∈R;

[0019] The variable node v jj based on We obtain the variable node v jk Based on Q jk The quadratic term node u is obtained. j Based on p j , Obtain; constraint node c i Based on b i get;

[0020] The variable node v jj and the quadratic term node u j The first characteristic edge between them is 2; the quadratic term node u j and the quadratic term node u j The first feature edge between them is 1; the variable node v jj and constraint node c i The second characteristic edge between them is The quadratic term node u j and constraint node c i The second characteristic edge between them is The quadratic term node u j and constraint node c i The third characteristic edge between them is

[0021] According to one embodiment of this application, the messaging model includes an embedding layer, a multi-layer messaging layer, and a reading layer connected in sequence.

[0022] The step of analyzing multiple nodes and feature edges between the nodes in the target triangulation graph using the message passing model to obtain the problem attributes of the target planning problem output by the message passing model includes:

[0023] The initial features of multiple nodes in the target triangulation are mapped to the hidden state space through the embedding layer to obtain the state vector;

[0024] The state of the multiple nodes is updated through the multi-layer message passing layer based on the state vector;

[0025] The problem attributes are obtained by performing feature mapping on the updated state vectors of the multiple nodes through the reading layer.

[0026] According to one embodiment of this application, the messaging model is obtained based on the following steps:

[0027] Obtain a dataset, which includes multiple sample planning problems and attribute labels corresponding to the sample planning problems. The sample planning problems include a quadratic objective function and quadratic constraints.

[0028] The sample planning problem in the dataset is sequentially input into the GNNs network to obtain the predicted attributes output by the GNNs network;

[0029] The parameters of the GNNs network are updated based on the attribute labels and the predicted attributes.

[0030] The trained GNNs network is used as the message passing model.

[0031] Secondly, this application provides a signal analysis system based on the prediction of solutions to a quadratic constrained quadratic programming problem, the system comprising:

[0032] The acquisition module is used to acquire the target planning problem of the power transmission signal, wherein the target planning problem includes a quadratic objective function and quadratic constraints.

[0033] The first processing module is used to encode the quadratic objective function and the quadratic constraints to generate a target triangulation graph of the target programming problem.

[0034] The second processing module is used to input the target three-part graph into the message passing model, and analyze multiple nodes in the target three-part graph and the feature edges between the multiple nodes through the message passing model to obtain the problem attributes of the target planning problem output by the message passing model. The problem attributes include feasibility, target value and optimal solution.

[0035] The third processing module is used to perform signal analysis on the power transmission signal based on the problem attributes.

[0036] Thirdly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the signal analysis method for predicting solutions to a quadratic programming problem based on quadratic constraints as described in the first aspect above.

[0037] Fourthly, this application provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the signal analysis method for predicting solutions to a quadratic programming problem based on quadratic constraints as described in the first aspect above.

[0038] Fifthly, this application provides a chip including a processor and a communication interface, the communication interface being coupled to the processor, the processor being used to run programs or instructions to implement the signal analysis method based on the prediction of solutions to a quadratic programming problem with quadratic constraints as described in the first aspect.

[0039] In a sixth aspect, this application provides a computer program product, including a computer program that, when executed by a processor, implements the signal analysis method for predicting solutions to a quadratic programming problem based on quadratic constraints as described in the first aspect above.

[0040] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application.

[0041] This application provides a signal analysis method, system, device, and medium based on the prediction of solutions to a quadratic constrained quadratic programming problem, which has the following advantages over existing technologies:

[0042] (1) By encoding the target programming problem into a three-part graph and predicting the feasibility, target value and optimal solution of the quadratic constraint quadratic programming problem by using the message passing model, the complex interaction relationship brought about by the quadratic constraint can be effectively captured, the optimization problem can be processed in real time, and the analysis efficiency of power transmission signals can be improved.

[0043] (2) The quadratic constraint quadratic programming problem is represented by multiple nodes in the three-part graph and the feature edges between the nodes, so that all the information of the quadratic constraint quadratic programming problem can be represented in the three-part graph. This can effectively capture the complex interaction relationship brought about by the quadratic constraint, thus laying the foundation for subsequent solution based on the message passing model. Attached Figure Description

[0044] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which:

[0045] Figure 1 This is one of the flowcharts illustrating the signal analysis method based on the prediction of solutions to a quadratic programming problem provided in this application embodiment;

[0046] Figure 2 This is a flowchart illustrating the conversion of a QCQP problem into a three-part diagram, as provided in an embodiment of this application.

[0047] Figure 3 This is a schematic diagram of the message passing model provided in the embodiments of this application;

[0048] Figure 4 This is the second flowchart illustrating the signal analysis method based on the prediction of solutions to a quadratic programming problem provided in this application embodiment;

[0049] Figure 5 This is a schematic diagram of the structure of a signal analysis system based on the prediction of solutions to a quadratic programming problem with quadratic constraints, provided in an embodiment of this application.

[0050] Figure 6This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0051] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0052] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0053] The following description, in conjunction with the accompanying drawings, details the signal analysis method, system, electronic device, and readable storage medium based on the prediction of solutions to quadratic constrained quadratic programming problems provided in this application, through specific embodiments and application scenarios.

[0054] Among them, the signal analysis method based on the prediction of solutions to quadratic programming problems with quadratic constraints can be applied to the terminal, specifically executed by the hardware or software in the terminal.

[0055] The terminal includes, but is not limited to, portable communication devices such as mobile phones or tablets with touch-sensitive surfaces (e.g., touchscreen displays and / or touchpads). It should also be understood that, in some embodiments, the terminal may not be a portable communication device, but rather a desktop computer with touch-sensitive surfaces (e.g., touchscreen displays and / or touchpads).

[0056] The following embodiments describe a terminal including a display and a touch-sensitive surface. However, it should be understood that the terminal may include one or more other physical user interface devices such as a physical keyboard, mouse, and joystick.

[0057] The signal analysis method based on the prediction of solutions to a quadratic constrained quadratic programming problem provided in this application embodiment can be executed by an electronic device or a functional module or entity within an electronic device that can implement the signal analysis method. The electronic devices mentioned in this application embodiment include, but are not limited to, mobile phones, tablets, computers, cameras, and wearable devices. The following description uses an electronic device as an example to illustrate the signal analysis method based on the prediction of solutions to a quadratic constrained quadratic programming problem provided in this application embodiment.

[0058] In recent years, machine learning (ML)-based quadratic programming solutions can learn the adaptive configuration of specific solving algorithms or optimizers to accelerate the solution process; they can also predict the initial solution of quadratic programming problems and use it as the initial input for the final solution or further algorithm optimization.

[0059] Graph Neural Networks (GNNs) are a type of neural network that can process data represented in graph form—that is, neural networks that can be deployed on graphs. They are widely used to leverage the structural information inherent in quadratic programming problems. GNNs excel at handling graph-structured data, making them particularly adept at representing the relationships and dependencies between variables in quadratic programming problems. By encoding quadratic programming instances as graph structures, GNNs can capture the complex features of the problem and provide adaptive guidance or efficient approximate solutions during the solution process.

[0060] like Figure 1 As shown, the signal analysis method based on the prediction of solutions to a quadratic programming problem with quadratic constraints includes:

[0061] Step 110: Obtain the target programming problem of the power transmission signal, wherein the target programming problem includes a quadratic objective function and quadratic constraints.

[0062] Quadratically Constrained Quadratic Programming (QCQP) is a class of optimization problems involving quadratic objective functions and quadratic constraints. Its standard form is as follows:

[0063]

[0064] Among them, Q, It is a symmetric matrix, p, It is a vector, b i ∈R is a constant, x L and x U Let Q and Q' represent the lower and upper bounds of the variables, respectively. If all matrices Q and Q''...i If all the matrices are positive semi-definite matrices, then such a QCQP problem is called convex.

[0065] use This represents the feasible set of solutions to the problem, i.e., the set of solutions that satisfy all constraints and boundary conditions. like If the set is not empty, the problem is feasible; if the set is empty, it is infeasible. If the objective function is in... If a problem has a finite lower bound, it is said to be bounded. In the bounded case, there is a feasible solution x that makes the objective function value take the infimum. * This is called the optimal solution.

[0066] Traditional graphical representation methods, such as bipartite graph representation in linear programming (LP) problems, have been widely used in solving optimization problems. However, since the constraints and objective functions in QCQP are quadratic, directly using bipartite graph modeling presents challenges. To address this, this application proposes a tripartite graph representation G... QCQP = (V, E) to accurately model QCQP without losing information.

[0067] Although existing research has demonstrated in theory and practice that graph neural networks (GNNs) can accelerate the computational efficiency of existing quadratic programming (QP) solvers and directly provide approximate solutions for various QP instances, their application in the field of quadratic constrained quadratic programming (QCQP) remains very limited, especially in the exploration of handling quadratic constraints. Existing graph representation methods for linearly constrained quadratic programming (LCQP) cannot effectively capture the complex interactions brought about by quadratic constraints.

[0068] Step 120: Encode the quadratic objective function and the quadratic constraints to generate the objective trigonometric diagram of the objective programming problem.

[0069] In some embodiments, encoding the quadratic objective function and the quadratic constraints to generate the objective trigonometric graph of the objective programming problem includes:

[0070] The optimization quantities are extracted from the quadratic objective function and the quadratic constraints to form the multiple nodes, and the coefficient terms are extracted as the feature edges;

[0071] The target triangulation is constructed based on the correspondence between the multiple nodes and the feature edges.

[0072] In some embodiments, the plurality of nodes include variable nodes, quadratic term nodes, and constraint nodes, and the feature edges include a first feature edge between the variable node and the quadratic term node, a second feature edge between the variable node and the constraint node, and a third feature edge between the quadratic term node and the constraint node.

[0073] In some embodiments, such as Figure 2 As shown, the goal programming problem includes:

[0074]

[0075] Where, x∈R n The decision variables are a symmetric matrix Q, a vector p, and a constant b. i The problem provides a set of coefficients. b i ∈R;

[0076] The variable node v jj based on We obtain the variable node v jk Based on Q jk The quadratic term node u is obtained. j Based on p j , Obtain; constraint node c i Based on b i get;

[0077] The variable node v jj and the quadratic term node u j The first characteristic edge between them is 2; the quadratic term node u j and the quadratic term node u j The first feature edge between them is 1; the variable node v jj and constraint node c i The second characteristic edge between them is The quadratic term node u j and constraint node c i The second characteristic edge between them is The quadratic term node u j and constraint node c i The third characteristic edge between them is

[0078] The QCQP formulation problem is transformed into a tripartite representation of the objective tripartite graph, which consists of three types of nodes: variable nodes, quadratic-term nodes, and constraint nodes.

[0079] Where, the quadratic term node v jj correspond v jk Corresponding to Qjk ; Variable node u j Corresponding to pj, Constraint node c i Corresponding to b i ;

[0080] The variable node is defined as V1 = {u1, u2, ..., u}. n}, each node u i Corresponding to variable x i and carries characteristics

[0081] The quadratic term node is defined as follows Each node v j,k Represents the quadratic term x j x k ,in It is the index set of all non-zero quadratic coefficients. Node v j,k The characteristic of is the coefficient of its corresponding quadratic term in the objective function, i.e.

[0082] The constraint nodes are defined as V3 = {c1, c2, ..., c m}, each node c i Corresponding to a constraint condition, carrying feature h c =b i .

[0083] The feature edges in a tripartite graph are also divided into three categories:

[0084] The first characteristic edge between the variable node and the quadratic term node: E 12 ={(u j′ ,v j,k ):j′=j or j′=k}

[0085] The weight of the first feature edge is assigned according to the relationship between j and k:

[0086] The second characteristic edge between the variable node and the constraint node: The weight of the second feature edge is

[0087] The third feature edge between the quadratic term node and the constraint node: The weight of the third feature edge is determined based on the coefficient of the quadratic term:

[0088] In this embodiment, the quadratic constraint quadratic programming problem is represented by multiple nodes in the tri-graph and the feature edges between the nodes, so that all the information of the quadratic constraint quadratic programming problem can be represented in the tri-graph. This can effectively capture the complex interaction relationship brought about by the quadratic constraint, thus laying the foundation for subsequent solution based on the message passing model.

[0089] Step 130: Input the target three-part graph into the message passing model, and analyze the multiple nodes in the target three-part graph and the feature edges between the multiple nodes through the message passing model to obtain the problem attributes of the target planning problem output by the message passing model. The problem attributes include feasibility, target value and optimal solution.

[0090] The message passing model is built on graph neural networks (GNNs).

[0091] In some embodiments, the messaging model is obtained based on the following steps:

[0092] Obtain a dataset, which includes multiple sample planning problems and attribute labels corresponding to the sample planning problems. The sample planning problems include a quadratic objective function and quadratic constraints.

[0093] The sample tripartite graphs corresponding to the sample planning problem in the dataset are sequentially input into the GNNs network to obtain the predicted attributes output by the GNNs network.

[0094] The parameters of the GNNs network are updated based on the attribute labels and the predicted attributes.

[0095] The trained GNNs network is used as the message passing model.

[0096] First, you need to prepare the dataset. Where M (i) It's a QCQP problem, x (i) These are the attribute labels corresponding to the problem, which can be feasibility, objective value, or optimal solution. These attribute labels can be obtained by solving the problem using the general solver Gurobi for a certain period of time (e.g., 1 hour). Then,

[0097] QCQP problems in the dataset The corresponding sample triangulation is input into the model prediction module to obtain the predicted attributes. These predictive attributes will be along with attribute labels. By comparing the results, the loss function is calculated and used to update the GNN network f. θ This process of updating is repeated cyclically until a predetermined number of iterations are reached, at which point the trained GNN network is output as a message passing model.

[0098] The gradient algorithm is used to optimize the GNN network. The specific process is shown in the algorithm below:

[0099]

[0100] Where T is the maximum number of training rounds; N is the number of convex quadratic programming problems in the dataset; It is the loss value, θ t These are the model parameters at time t, and α is the learning rate. It is the gradient of the model parameters with respect to the loss value.

[0101] This algorithm calculates the loss and gradient before each model update to optimize the model parameters. Here, GD represents the optimization algorithm, such as gradient descent, Adam, etc., and e is the loss function. If the task is to predict feasibility, i.e., x... (i) For feasibility (0 for infeasible, 1 for feasible), the loss function is binary classification cross-entropy loss. If the task is to predict the target value or the optimal solution, i.e., x... (i) If the target value or optimal solution is to be used, then the loss function adopts the least squares loss.

[0102] In some embodiments, the messaging model includes an embedding layer, a multi-layer messaging layer, and a reading layer connected in sequence.

[0103] The step of analyzing multiple nodes and feature edges between the nodes in the target triangulation graph using the message passing model to obtain the problem attributes of the target planning problem output by the message passing model includes:

[0104] The initial features of multiple nodes in the target triangulation are mapped to the hidden state space through the embedding layer to obtain the state vector;

[0105] The state of the multiple nodes is updated through the multi-layer message passing layer based on the state vector;

[0106] The problem attributes are obtained by performing feature mapping on the updated state vectors of the multiple nodes through the reading layer.

[0107] In actual implementation, such as Figure 3 As shown, the embedding layer maps the initial features of all nodes to the hidden state space, that is:

[0108] in, It is a learnable embedding function.

[0109] In some embodiments, the multi-layer messaging layer includes a first sub-layer that updates quadratic term nodes from variable nodes, a second sub-layer that updates constraint nodes from variable nodes and quadratic term nodes, a third sub-layer that updates quadratic term nodes from constraint nodes, and a fourth sub-layer that updates variable nodes from constraint nodes and quadratic term nodes.

[0110] The Message-Passing Layer consists of four sub-layers:

[0111] In the first sub-level, update the quadratic item nodes (from V1 to V2) from the variable nodes:

[0112] In the second sub-layer, update the constraint nodes (from V1 and V2 to V3) from the variable nodes and quadratic term nodes:

[0113] In the third sub-layer, update the quadratic term nodes (from V3 to V2) from the constraint nodes:

[0114] In the fourth sub-layer, update the variable nodes (from V3 and V2 to V1) from the constraint nodes and quadratic term nodes:

[0115] The T layer is used to transform the input data, performing linear or nonlinear transformations. It can be used to adjust the feature representation of nodes or change the depth / complexity of the network.

[0116] The Readout Layer maps the final node features to problem attributes and uses the problem attributes as the output.

[0117] Problem attributes can be graph-level scalars or node-level vectors. Graph-level scalar outputs are:

[0118]

[0119] The node-level vector output is:

[0120] Step 140: Based on the problem attributes, perform signal analysis on the power transmission signal.

[0121] In practical implementation, based on the analytical framework of feasibility, target values, and optimal solutions, the analysis of power transmission signals involves a comprehensive evaluation of various factors in the transmission process. Feasibility analysis ensures that the system's basic conditions and limitations are met; target values ​​clarify the specific performance standards the system should achieve; and optimal solution analysis optimizes the overall system performance through different technical means and strategies to achieve efficient, reliable, and energy-saving signal transmission.

[0122] According to the signal analysis method based on the prediction of solutions to quadratic constrained quadratic programming problems provided in the embodiments of this application, by encoding the target programming problem into a three-part graph and predicting the feasibility, target value and optimal solution of quadratic constrained quadratic programming by using a message passing model, it can effectively capture the complex interaction relationships brought about by quadratic constraints, process optimization problems in real time, and improve the analysis efficiency of power transmission signals.

[0123] like Figure 4 As shown, the dataset of the QCQP problem is input into the GNNs network, the GNNs network is trained, the loss is calculated, and the parameters of the GNNs network are updated until the number of training times reaches a threshold, and the trained GNNs network is obtained as a message passing model.

[0124] The goal programming problem is encoded as a goal trigonometric diagram. The attributes of the goal programming problem are predicted by a message passing model to obtain the problem attributes, which can be feasibility, target value, and optimal solution.

[0125] In this embodiment, the QCQP problem is encoded to obtain a trie graph. A Generative Neural Network (GNN) is constructed based on the trie graph and trained using a QCQP problem dataset to obtain a message passing model that can predict the properties of the QCQP problem. Through the design of a graph neural network under the new graph structure representation, its feasibility on convex QCQP problems is demonstrated from both theoretical and numerical experiments. It can effectively capture the complex interaction relationships brought about by quadratic constraints, solve the difficulty of representing QCQP problems with graph structures, realize accurate modeling of QCQP using graph structures, and handle optimization problems in real time.

[0126] The signal analysis method based on the prediction of solutions to a quadratic constrained quadratic programming problem provided in this application can be executed by a signal analysis system based on the prediction of solutions to a quadratic constrained quadratic programming problem. This application uses an example of a signal analysis system based on the prediction of solutions to a quadratic constrained quadratic programming problem executing the signal analysis method based on the prediction of solutions to a quadratic constrained quadratic programming problem to illustrate the signal analysis system based on the prediction of solutions to a quadratic constrained quadratic programming problem provided in this application.

[0127] This application also provides a signal analysis system based on the prediction of solutions to a quadratic programming problem with quadratic constraints.

[0128] like Figure 5 As shown, the signal analysis system based on the prediction of solutions to a quadratic programming problem with quadratic constraints includes:

[0129] The acquisition module 510 is used to acquire the target planning problem of the power transmission signal, wherein the target planning problem includes a quadratic objective function and quadratic constraints.

[0130] The first processing module 520 is used to encode the quadratic objective function and the quadratic constraints to generate a target triangulation graph of the target programming problem.

[0131] The second processing module 530 is used to input the target three-part graph into a message passing model, and analyze multiple nodes in the target three-part graph and the feature edges between the multiple nodes through the message passing model to obtain the problem attributes of the target planning problem output by the message passing model. The problem attributes include feasibility, target value and optimal solution.

[0132] The third processing module 540 is used to perform signal analysis on the power transmission signal based on the problem attributes.

[0133] According to the signal analysis system based on the prediction of solutions to quadratic constrained quadratic programming problems provided in the embodiments of this application, by encoding the target programming problem into a three-part graph and predicting the feasibility, target value and optimal solution of the quadratic constrained quadratic programming problem by using a message passing model, it can effectively capture the complex interaction relationships brought about by the quadratic constraints, process optimization problems in real time, and improve the analysis efficiency of power transmission signals.

[0134] The signal analysis system based on the prediction of solutions to a quadratic programming problem with quadratic constraints in this application embodiment can be an electronic device or a component within an electronic device, such as an integrated circuit or a chip. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc. It can also be a server, network attached storage (NAS), personal computer (PC), television (TV), ATM, or self-service machine, etc. This application embodiment does not specifically limit the specific implementation.

[0135] The signal analysis system based on the prediction of solutions to a quadratic programming problem with quadratic constraints in this application embodiment can be an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit it.

[0136] The signal analysis system based on the prediction of solutions to quadratic programming problems provided in this application can realize the various processes implemented in the above embodiments of the signal analysis method based on the prediction of solutions to quadratic programming problems. To avoid repetition, these processes will not be described again here.

[0137] In some embodiments, such as Figure 6 As shown, this application embodiment also provides an electronic device 600, including a processor 601, a memory 602, and a computer program stored in the memory 602 and executable on the processor 601. When the program is executed by the processor 601, it implements the various processes of the above-described signal analysis method embodiment based on the prediction of solutions to a quadratic constraint quadratic programming problem, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0138] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0139] This application also provides a non-transitory computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the above-described signal analysis method embodiment based on the prediction of solutions to a quadratic programming problem, and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0140] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0141] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described signal analysis method based on the prediction of solutions to a quadratic programming problem with quadratic constraints.

[0142] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0143] This application also provides a chip, which includes a processor and a communication interface. The communication interface and the processor are coupled. The processor is used to run programs or instructions to implement the various processes of the above-described signal analysis method embodiment based on the prediction of solutions to quadratic programming problems, and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0144] It should be understood that the chip mentioned in the embodiments of this application may also be referred to as a system-on-a-chip, system chip, chip system, or system-on-a-chip, etc.

[0145] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0146] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a computer software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, or network device, etc.) to execute the signal analysis method based on the prediction of solutions to quadratic constrained quadratic programming problems in the various embodiments of this application.

[0147] In the description of this application, "first feature" and "second feature" may include one or more of the features.

[0148] In the description of this application, "multiple" means two or more.

[0149] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

[0150] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0151] Although embodiments of this application have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of this application, the scope of which is defined by the claims and their equivalents.

Claims

1. A signal analysis method based on the prediction of solutions to a quadratic programming problem with quadratic constraints, characterized in that, include: The objective planning problem for acquiring power transmission signals includes a quadratic objective function and quadratic constraints. Encode the quadratic objective function and the quadratic constraints to generate a target triangulation graph of the objective programming problem; The target trigonometric graph is input into a message passing model. The message passing model analyzes multiple nodes in the target trigonometric graph and the feature edges between the multiple nodes to obtain the problem attributes of the target planning problem output by the message passing model. The problem attributes include feasibility, target value, and optimal solution. Based on the aforementioned problem attributes, signal analysis is performed on the power transmission signal; The step of encoding the quadratic objective function and the quadratic constraints to generate the objective trigonometric graph of the objective programming problem specifically includes: The optimization quantities are extracted into multiple nodes from the quadratic objective function and the quadratic constraints, and the coefficient terms are extracted as feature edges. Based on the correspondence between the multiple nodes and the feature edges, the target triangulation is constructed; The plurality of nodes include variable nodes, quadratic term nodes, and constraint nodes. The feature edges include a first feature edge between the variable node and the quadratic term node, a second feature edge between the variable node and the constraint node, and a third feature edge between the quadratic term node and the constraint node.

2. The signal analysis method based on the prediction of solutions to a quadratic programming problem with quadratic constraints as described in claim 1, characterized in that, The message passing model includes an embedding layer, a multi-layer message passing layer, and a reading layer connected in sequence; The step of analyzing multiple nodes and feature edges between the nodes in the target triangulation graph using the message passing model to obtain the problem attributes of the target planning problem output by the message passing model includes: The initial features of multiple nodes in the target triangulation are mapped to the hidden state space through the embedding layer to obtain the state vector; The state of the multiple nodes is updated through the multi-layer message passing layer based on the state vector; The problem attributes are obtained by performing feature mapping on the updated state vectors of the multiple nodes through the reading layer.

3. The signal analysis method based on the prediction of solutions to a quadratic programming problem with quadratic constraints as described in claim 2, characterized in that, The multi-layer message passing layer includes a first sub-layer that updates quadratic term nodes from variable nodes, a second sub-layer that updates constraint nodes from both variable nodes and quadratic term nodes, a third sub-layer that updates quadratic term nodes from both constraint nodes, and a fourth sub-layer that updates variable nodes from both constraint nodes and quadratic term nodes.

4. The signal analysis method based on the prediction of solutions to a quadratic programming problem with quadratic constraints as described in claim 1, characterized in that, The message passing model is obtained based on the following steps: Obtain a dataset, which includes multiple sample planning problems and attribute labels corresponding to the sample planning problems. The sample planning problems include a quadratic objective function and quadratic constraints. The sample tripartite graphs corresponding to the sample planning problem in the dataset are sequentially input into the GNNs network to obtain the predicted attributes output by the GNNs network. The parameters of the GNNs network are updated based on the attribute labels and the predicted attributes. The trained GNNs network is used as the message passing model.

5. A signal analysis system based on the prediction of solutions to a quadratic constrained quadratic programming problem, used to implement the signal analysis method based on the prediction of solutions to a quadratic constrained quadratic programming problem as described in any one of claims 1-4, characterized in that, include: The acquisition module is used to acquire the target planning problem of the power transmission signal, wherein the target planning problem includes a quadratic objective function and quadratic constraints. The first processing module is used to encode the quadratic objective function and the quadratic constraints to generate a target triangulation graph of the target programming problem. The second processing module is used to input the target three-part graph into the message passing model, and analyze multiple nodes in the target three-part graph and the feature edges between the multiple nodes through the message passing model to obtain the problem attributes of the target planning problem output by the message passing model. The problem attributes include feasibility, target value and optimal solution. The third processing module is used to perform signal analysis on the power transmission signal based on the problem attributes.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the signal analysis method based on the prediction of solutions to a quadratic programming problem as described in any one of claims 1-4.

7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the signal analysis method for predicting solutions to a quadratic programming problem based on quadratic constraints as described in any one of claims 1-4.

Images