Method, system and device for joint optimization of security constrained unit mixed integer programming

Abstract

The application discloses a kind of security constrained unit's mixed integer programming joint optimization method, system and equipment, it is related to power system unit combination optimization field, method includes: to the security constrained unit combination optimization model of pre-constructed model processing, obtain mixed integer linear programming model;According to prior rule, a plurality of variables in mixed integer linear programming model is fixed based on two-stage multivariate branch method;Artificial intelligence method is used to obtain solution information in the branch and bound solution process of the multivariate branch method, and the rule of the value of optimal solution and relevant parameter is learned, to predict the value of part of variables of optimal solution.The application is based on artificial intelligence method and large-scale integer programming solution interaction, obtain certain solution information in the branch and bound solution process, and learn the rule of the value of optimal solution and relevant parameter, to predict the value of part of variables of optimal solution, can reduce solution scale, and improve solution efficiency.

Description

Technical Field

[0001] This invention relates to the field of power system unit combination optimization, and specifically to a hybrid integer programming joint optimization method, system and equipment for safety-constrained units. Background Technology

[0002] Safety-constrained unit combination optimization is a crucial component of dispatching plans and spot markets. For large power systems, safety-constrained unit combination considers constraints such as power balance constraints, network security constraints, unit capacity constraints, unit operating reserve limitations, and unit ramp-up and ramp-down rates. After linearization, the power system unit combination problem primarily involves mixed-integer linear programming (MILP) modeling and solution techniques. Power system unit combination problems often involve millions of variables and constraints, while spot market construction places high demands on solution speed, accuracy, and convergence stability. Therefore, targeted research is necessary, taking into account the specific characteristics of the power system unit combination problem. Summary of the Invention

[0003] The purpose of this invention is to address the complex and demanding problem of solving the power system unit combination problem in the prior art by providing a hybrid integer programming joint optimization method, system, and equipment for safety-constrained units. Based on the multi-variable branching (MVB) method and combined with the characteristics of the power system unit combination problem, it can reduce the solution scale and achieve a comprehensive improvement in calculation speed, calculation accuracy, and convergence stability.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] Firstly, a hybrid integer programming joint optimization method for safety-constrained units is provided, including:

[0006] The pre-constructed safety-constrained unit combination optimization model is processed to obtain a mixed-integer linear programming model.

[0007] Based on the two-stage multivariate branching method, multiple variables in the mixed-integer linear programming model are fixed according to prior rules;

[0008] Artificial intelligence methods are used to obtain solution information during the branch and bound process of the multivariate branching method, and to learn the values ​​of the optimal solution and the patterns of related parameters, thereby predicting the values ​​of some variables in the optimal solution.

[0009] As a preferred embodiment, in the step of processing the pre-constructed safety-constrained unit combination optimization model, for the safety-constrained unit combination optimization model determined by the demand parameter d∈D, given the abstract function c:[K]:={2,...,K}, there exists f0: and Where I is the set of optimal solutions, and f0 is the set of... The mapping function, f1 is the mapping function from [K] to μ. μ are all subsets of the optimal solution set I, such that:

[0010]

[0011]

[0012] Based on the safety-constrained unit characteristics, as the demand parameter d varies within a certain range, the partial variable I of the optimal solution is obtained. j The value remains unchanged, i.e., it is always either 1 or 0. Therefore, it is assumed that there is a correlation between the demand parameter d and the optimal solution, and the values ​​of some variables are predicted based on the demand parameter d.

[0013] As a preferred embodiment, in the step of predicting the values ​​of some variables using the demand parameter d, the characteristic of probability is introduced, predicting the probability that a certain variable will take the value 1. The calculation expression is as follows:

[0014]

[0015] In the formula, It is a sigmoid function, and These are model parameters in the real number field, i.e. Use the abstract function c(·) to distinguish the model parameters for different k∈[K]. The abstract function c is defined as: c:[K]={2,...,K}, where K is the dimension of the requirement parameter d, j is the dimension of the optimal solution set I, and artificial intelligence methods are used to fit this parameter to predict the probability that the variable takes the value 1.

[0016] As a preferred option, the partial variable I of the optimal solution j Using model parameters {β j Calculate I using the following formula j =σ(β) j d), j∈[|I|], where β j d represents the model parameters and demand parameters determined based on the required unit combination model.

[0017] As a preferred embodiment, in the step of fixing multiple variables in the mixed-integer linear programming model based on the two-stage multivariate branching method according to prior rules, when variable {I j When the probability of taking the value 1 is high, take I. j =1, when the variable {I j When the probability of taking the value 0 is high, take I. j=0; Use machine learning methods to fit the corresponding model parameters β. j After that, with Use a threshold value to ensure that the predicted variable value is greater than a certain threshold. Take 1, which is less than the threshold. When the value is 0, the first stage of fixing the variable is completed, and the expression is as follows: I j =1, when And fix I j =0, Where k∈[K], K is the dimension of the requirement parameter d.

[0018] As a preferred embodiment, in the step of acquiring solution information using artificial intelligence methods during the branch and bound process of the multivariate branch and bound method, when the problem reaches an intermediate threshold, an MVB cut is added in the second stage to expand the feasible region of the problem. Specifically, this includes:

[0019] Branching based on infeasibility:

[0020] Branch upwards using the following formula:

[0021]

[0022] Where S u C u For the two cutting planes of the corresponding upward branch, δ u These are slack variables for the upward branch;

[0023] Branch downwards using the following formula:

[0024]

[0025] Where S l C l For the two cutting planes of the corresponding upward branch, δ l These are slack variables for the upward branch;

[0026] Solve the original problem and add the cutting planes in the following order:

[0027] a)S u ,S l ;

[0028] b)S l C u and S u C l ;

[0029] c)C u C l .

[0030] Secondly, a hybrid integer programming joint optimization system for safety-constrained units is provided, comprising:

[0031] The model processing module is used to process the pre-built safety-constrained unit combination optimization model to obtain a mixed-integer linear programming model.

[0032] The variable fixing module is used to fix multiple variables in a mixed integer linear programming model based on a priori rules using a two-stage multivariate branching method.

[0033] The optimal solution prediction module is used to acquire solution information during the branch and bound process of the multivariate branching method using artificial intelligence methods, and to learn the values ​​of the optimal solution and the patterns of related parameters, thereby predicting the values ​​of some variables in the optimal solution.

[0034] As a preferred embodiment, for the safety-constrained unit combination optimization model determined by the demand parameter d∈D, the model processing module, given the abstract function c:[K]:={2,...,K}, has f0: and Where I is the set of optimal solutions, and f0 is the set of... The mapping function, f1 is the mapping function from [K] to μ. μ are all subsets of the optimal solution set I, such that:

[0035]

[0036]

[0037] Based on the safety-constrained unit characteristics, as the demand parameter d varies within a certain range, the partial variable I of the optimal solution is obtained. j The value remains unchanged, i.e., it is always either 1 or 0. Therefore, it is assumed that there is a correlation between the demand parameter d and the optimal solution, and the values ​​of some variables are predicted based on the demand parameter d.

[0038] As a preferred embodiment, when the model processing module predicts the values ​​of some variables using the demand parameter d, it introduces the characteristic of probability, predicting the probability that a certain variable will take the value 1. The calculation expression is as follows:

[0039]

[0040] In the formula, It is a sigmoid function, and These are model parameters in the real number field, i.e. Use the abstract function c(·) to distinguish the model parameters for different k∈[K]. The abstract function c is defined as: c:[K]={2,...,K}, where K is the dimension of the requirement parameter d, j is the dimension of the optimal solution set I, and artificial intelligence methods are used to fit this parameter to predict the probability that the variable takes the value 1.

[0041] As a preferred embodiment, the model processing module uses model parameters {β} j Calculate the partial variable I of the optimal solution using the following formula. j :

[0042] I j =σ(β) j d)

[0043] In the formula, j∈[|I|], β j d represents the model parameters and demand parameters determined based on the required unit combination model.

[0044] As a preferred embodiment, the variable fixing module when variable {I j When the probability of taking the value 1 is high, take I. j =1, when the variable {I j When the probability of taking the value 0 is high, take I. j =0; Use machine learning methods to fit the corresponding model parameters β. j After that, with Use a threshold value to ensure that the predicted variable value is greater than a certain threshold. Take 1, which is less than the threshold. When the value is 0, the first stage of fixing the variable is completed, and the expression is as follows: I j =1, when And fix I j =0, Where k∈[K], K is the dimension of the requirement parameter d.

[0045] As a preferred approach, the optimal solution prediction module, during the branch and bound process of the multivariate branch method, adds an MVB cutting plane in the second stage when the solution reaches an intermediate threshold, and expands the feasible region of the problem. Specifically, this includes:

[0046] Branching based on infeasibility:

[0047] Branch upwards using the following formula:

[0048] and

[0049] Where S u C u For the two cutting planes of the corresponding upward branch, δ u These are slack variables for the upward branch;

[0050] Branch downwards using the following formula:

[0051] and

[0052] Where S l C l For the two cutting planes of the corresponding upward branch, δ l These are slack variables for the upward branch;

[0053] Solve the original problem and add the cutting planes in the following order:

[0054] a)S u ,S l ;

[0055] b)S l C u and S u C l ;

[0056] c)C u C l .

[0057] Thirdly, an electronic device is provided, 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 steps of the hybrid integer programming joint optimization method for the security-constrained unit.

[0058] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the hybrid integer programming joint optimization method for the security-constrained unit.

[0059] Compared with the prior art, the first aspect of the present invention has at least the following beneficial effects:

[0060] Based on the interaction between artificial intelligence methods and large-scale integer programming solutions, certain solution information is obtained during the branch and bound process. This information is then used to learn the values ​​of the optimal solution and some patterns in related parameters, thereby predicting the values ​​of some variables in the optimal solution. This reduces the solution scale and improves solution efficiency. A two-stage multivariate branching method is used to reduce the scale of integer programming problems and accelerate the solution speed of mixed integer programming models. This method not only fixes one variable in the branching process but also fixes multiple variables simultaneously based on certain prior rules. For large power systems, safety-constrained unit combinations consider power balance constraints, network security constraints, unit capacity constraints, unit operating reserve limits, and unit ramp-up and ramp-down speeds. Mathematically, this is a large-scale mixed integer programming (MIP) problem with complex models and high computational demands. The mixed integer programming joint optimization method of this invention can be customized and modified according to needs, enhancing the technical support capabilities of the integer programming optimization engine.

[0061] It is understood that the beneficial effects of the second to fourth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0063] Figure 1 Flowchart of the hybrid integer programming joint optimization method for safety-constrained units according to an embodiment of the present invention;

[0064] Figure 2 A block diagram of the hybrid integer programming joint optimization system for safety-constrained units according to an embodiment of the present invention. Detailed Implementation

[0065] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0066] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0067] Example 1

[0068] This invention proposes a joint optimization method for mixed-integer programming of power units with safety constraints, based on the branch-and-bound method and artificial intelligence methods. It also proposes an improved and upgraded method based on the branch-and-bound method—Multi-variable Branching (MVB). By considering the characteristics of power system problems, this method achieves a comprehensive improvement in computational scale, speed, accuracy, and convergence stability.

[0069] This invention presents a joint optimization method for mixed-integer programming in safety-constrained units. This method analyzes the characteristics of high-dimensional, multi-objective, and multi-time-period ultra-large-scale mixed-integer nonlinear programming. The scheme elucidates the relationships between variables, studies methods for the rational selection and prioritization of constraints, extracts key information from large-scale 0-1 variables and complex and numerous constraints, investigates methods to reduce the computational scale of mixed-integer programming problems, explores methods to improve branch-and-bound efficiency using artificial intelligence, and studies a decision-making method combining data-driven and physical model-driven approaches, thereby improving the accuracy and efficiency of solving mixed-integer programming problems.

[0070] This invention, taking into account the characteristics of power systems, primarily addresses the problem of accelerating the solution of combinatorial optimization of Security Constrained Unit Commitments (SCUCs) in power systems. It proposes a mixed-integer programming joint optimization method for SCUCs. The method of this invention mainly includes the following steps:

[0071] The pre-constructed safety-constrained unit combination optimization model is processed to obtain a mixed-integer linear programming model.

[0072] Based on the two-stage multivariate branching method, multiple variables in the mixed-integer linear programming model are fixed according to prior rules;

[0073] Artificial intelligence methods are used to obtain solution information during the branch and bound process of the multivariate branching method, and to learn the values ​​of the optimal solution and the patterns of related parameters, thereby predicting the values ​​of some variables in the optimal solution.

[0074] In one possible implementation, in the step of model processing the pre-built safety-constrained unit combination optimization model, for the safety-constrained unit combination optimization model determined by the demand parameter d∈D, given the abstract function c:[K]:={2,...,K}, there exists and Where I is the set of optimal solutions, and f0 is the set of... The mapping function, f1 is the mapping function from [K] to μ. μ are all subsets of the optimal solution set I, such that:

[0075]

[0076]

[0077] Based on the safety-constrained unit characteristics, as the demand parameter d varies within a certain range, the partial variable I of the optimal solution is obtained. j The value remains unchanged, i.e., it is always either 1 or 0. Therefore, it is assumed that there is a correlation between the demand parameter d and the optimal solution, and the values ​​of some variables are predicted based on the demand parameter d.

[0078] Furthermore, in the step of predicting the values ​​of some variables using the demand parameter d, the characteristic of probability is introduced, predicting the probability that a certain variable will take the value 1. The calculation expression is as follows:

[0079]

[0080] In the formula, It is a sigmoid function, and These are model parameters in the real number field, i.e. Use the abstract function c(·) to distinguish the model parameters for different k∈[K]. The abstract function c is defined as: c:[K]={2,...,K}, where K is the dimension of the requirement parameter d, j is the dimension of the optimal solution set I, and artificial intelligence methods are used to fit this parameter to predict the probability that the variable takes the value 1.

[0081] In the step of fixing multiple variables in the mixed-integer linear programming model based on the two-stage multivariate branching method according to prior rules, when variable {I j When the probability of taking the value 1 is high, take I. j =1, when the variable {I j When the probability of taking the value 0 is high, take I. j =0; Use machine learning methods to fit the corresponding model parameters β. j After that, with Use a threshold value to ensure that the predicted variable value is greater than a certain threshold. Take 1, which is less than the threshold. When the value is 0, the first stage of fixing the variable is completed.

[0082] Furthermore, in the step of obtaining solution information by using artificial intelligence methods in the branch and bound process of the multivariate branch method, when the problem is at an intermediate threshold, an MVB cutting plane is added in the second stage to expand the feasible region of the problem to enhance robustness.

[0083] like Figure 1As shown, the hybrid integer programming joint optimization method for safety-constrained units of the present invention is specifically implemented as follows:

[0084] 1. Model Processing. Based on the actual situation of the power system, a safe unit combination optimization problem model is formed. Through approximation and other methods, it is transformed into a mixed integer linear programming (MILP) problem.

[0085] 2. Initial solution calculation. Using model parameters {β} j Calculate I using the following formula j =σ(β) j d), j∈[|I|], where β j d represents the model parameters and demand parameters determined based on the required unit combination model.

[0086] 3. Fixed variables. j =1, when And fix I j =0, Where k∈[K], K is the dimension of the requirement parameter d.

[0087] 4. Branching according to infeasibility:

[0088] Branch upwards using the following formula:

[0089]

[0090] Where S u C u For the two cutting planes of the corresponding upward branch, δ u These are slack variables for the upward branch;

[0091] Branch downwards using the following formula:

[0092]

[0093] Where S l C l For the two cutting planes of the corresponding upward branch, δ l These are slack variables for the upward branch;

[0094] 5. Add cutting planes. Solve the original problem and add cutting planes in the following order:

[0095] a)S u ,S l

[0096] b)S l C u and S u Cl

[0097] c)C u C l

[0098] 6. Obtain the optimal solution. Solve and obtain the optimal solution.

[0099] The hybrid integer programming joint optimization method for safety-constrained units in this invention employs a multivariate branching method, which simultaneously fixes multiple variables in the optimization variables according to certain prior rules, thereby effectively reducing the scale of the integer programming problem; it also employs artificial intelligence methods to predict the values ​​of some variables in the optimal solution in order to reduce the solution scale and improve the solution efficiency.

[0100] In practical applications, a safety-constrained unit combination optimization model is formed based on the relevant power systems and actual needs involved in the electricity market. The method proposed in this invention can achieve rapid solution. This invention's hybrid integer programming joint optimization method for safety-constrained units expands upon a prototype algorithm set comprising preprocessing, initial linear programming, branch and bound method, cutting plane method, heuristic algorithm, and AI-assisted branching method, laying the algorithmic foundation for researching hybrid integer programming optimization engines.

[0101] Example 2

[0102] See Figure 2 This invention also proposes a hybrid integer programming joint optimization system for safety-constrained units, comprising:

[0103] Model processing module 1 is used to process the pre-built safety-constrained unit combination optimization model to obtain a mixed-integer linear programming model;

[0104] Variable fixing module 2 is used to fix multiple variables in a mixed integer linear programming model based on prior rules using a two-stage multivariate branching method.

[0105] The optimal solution prediction module 3 is used to obtain solution information during the branch and bound process of the multivariate branch method using artificial intelligence methods, and to learn the value of the optimal solution and the pattern of related parameters, thereby predicting the value of some variables of the optimal solution.

[0106] In one possible implementation, for the safety-constrained unit combination optimization model determined by the demand parameter d∈D, the model processing module 1, given the abstract function c:[K]:={2,...,K}, has the following... and Where I is the set of optimal solutions, and f0 is the set of... The mapping function, f1 is the mapping function from [K] to μ. μ are all subsets of the optimal solution set I, such that:

[0107]

[0108]

[0109] Based on the safety-constrained unit characteristics, as the demand parameter d varies within a certain range, the partial variable I of the optimal solution is obtained. j The value remains unchanged, i.e., it is always either 1 or 0. Therefore, it is assumed that there is a correlation between the demand parameter d and the optimal solution, and the values ​​of some variables are predicted based on the demand parameter d.

[0110] In one possible implementation, when model processing module 1 predicts the values ​​of some variables using the demand parameter d, it introduces the characteristic of probability, predicting the probability that a certain variable will take the value 1. The calculation expression is as follows:

[0111]

[0112] In the formula, It is a sigmoid function, and These are model parameters in the real number field, i.e. Use the abstract function c(·) to distinguish the model parameters for different k∈[K]. The abstract function c is defined as: c:[K]={2,...,K}, where K is the dimension of the requirement parameter d, j is the dimension of the optimal solution set I, and artificial intelligence methods are used to fit this parameter to predict the probability that the variable takes the value 1.

[0113] In one possible implementation, model processing module 1 uses model parameters {β} j Calculate the partial variable I of the optimal solution using the following formula. j :

[0114] I j =σ(β) j d)

[0115] In the formula, j∈[|I|], β j d represents the model parameters and demand parameters determined based on the required unit combination model.

[0116] In one possible implementation, variable fixing module 2 when variable {I j When the probability of taking the value 1 is high, take I. j =1, when the variable {I j When the probability of taking the value 0 is high, take I. j =0; Use machine learning methods to fit the corresponding model parameters β. j After that, with The threshold is set such that the predicted variable value is greater than a certain threshold. Take 1, which is less than the threshold. When the value is 0, the first stage of fixing the variable is completed, and the expression is as follows:

[0117] I j =1, when And fix I j =0, Where k∈[K], K is the dimension of the requirement parameter d.

[0118] In one possible implementation, during the branch and bound process of the multivariate branch method, the optimal solution prediction module 3, when at an intermediate threshold, adds an MVB cutting plane in the second stage and expands the feasible region of the problem, specifically including:

[0119] Branching based on infeasibility:

[0120] Branch upwards using the following formula:

[0121]

[0122] Where S u C u For the two cutting planes of the corresponding upward branch, δ u These are slack variables for the upward branch;

[0123] Branch downwards using the following formula:

[0124] and

[0125] Where S l C l For the two cutting planes of the corresponding upward branch, δ l These are slack variables for the upward branch;

[0126] Solve the original problem and add the cutting planes in the following order:

[0127] a)S u ,S l ;

[0128] b)S l C u and S u C l ;

[0129] c)C u C l .

[0130] Under the energy-saving power generation dispatch and electricity market model, dispatching agencies are required to fully explore the economic operation potential of the power grid while ensuring its safe operation, thus achieving an organic unity between safe and economic operation. Safety-constrained unit combination optimization is a crucial component of dispatching plans and the spot market. For large-scale power systems, safety-constrained unit combination considers constraints such as power balance constraints, network security constraints, unit capacity constraints, unit operating reserve limitations, and unit ramp-up and ramp-down rates. Mathematically, this problem is a large-scale mixed integer programming (MIP) problem, characterized by complex models and high computational demands, requiring the use of mixed integer programming software for solution. The MIP joint optimization system for safety-constrained units of this invention can be customized and modified according to requirements, enhancing the technical support capabilities of the integer programming optimization engine.

[0131] Example 3

[0132] Another embodiment of the present invention also proposes 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 steps of the hybrid integer programming joint optimization method for the security-constrained unit.

[0133] Example 4

[0134] Another embodiment of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the hybrid integer programming joint optimization method for the security-constrained unit.

[0135] The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable storage medium can include any entity or device capable of carrying the computer program code, a medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory, a random access memory, an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals. For ease of explanation, the above content only shows the parts related to the embodiments of the present invention; for specific technical details not disclosed, please refer to the method section of the embodiments of the present invention. This computer-readable storage medium is non-transitory and can be stored in storage devices formed by various electronic devices, enabling the execution process described in the method of the embodiments of the present invention.

[0136] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0137] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0138] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0139] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A hybrid integer programming joint optimization method for safety-constrained units, characterized in that, include: The pre-constructed safety-constrained unit combination optimization model is processed to obtain a mixed-integer linear programming model. Based on the two-stage multivariate branching method, multiple variables in the mixed-integer linear programming model are fixed according to prior rules; Artificial intelligence methods are used to obtain solution information during the branch and bound process of the multivariate branching method, and to learn the values ​​of the optimal solution and the patterns of related parameters, thereby predicting the values ​​of some variables in the optimal solution. In the step of fixing multiple variables in the mixed-integer linear programming model based on the two-stage multivariate branching method according to prior rules, when the variables When the probability of taking 1 is high, take When the variable When the probability of taking 0 is high, take Use machine learning methods to fit the corresponding model parameters. After that, with Use a threshold value to ensure that the predicted variable value is greater than a certain threshold. Take 1, which is less than the threshold. When the value is 0, the first stage of fixing the variable is completed, and the expression is as follows: ,when and fixed , ,in , Requirements parameters dimensionality ; In the step of acquiring solution information using artificial intelligence methods during the branch and bound process of the multivariate branch method, when the problem reaches an intermediate threshold, an MVB cut plane is added in the second stage to expand the feasible region of the problem. Specifically, this includes: Branching based on infeasibility: Branch upwards using the following formula: and in , These are the two cutting planes corresponding to the upward branches. These are slack variables for the upward branch; Branch downwards using the following formula: and in , These are the two cutting planes corresponding to the downward branches. These are slack variables for the downward branch; Solve the original problem and add the cutting planes in the following order: a) , ； b) , and , ; c) 、 。 2. The hybrid integer programming joint optimization method for safety-constrained units according to claim 1, characterized in that, In the step of model processing the pre-built safety-constrained unit combination optimization model, for the parameters determined by the demand parameters... A defined safety-constrained unit combination optimization model, with a given abstract function. ,exist and ,in, The set of optimal solutions. For set The mapping function, for The mapping function, , All are optimal solution sets A subset of such that: Based on the safety constraints of the unit characteristics, and with the demand parameters The variables vary within a certain interval, and the partial variables of the optimal solution are obtained by solving for these variables. The parameter remains unchanged, meaning it is always either 1 or 0. Therefore, we assume the requirement parameter... There is a correlation between the optimal solution and the requirement parameters. To predict the values ​​of some variables.

3. The hybrid integer programming joint optimization method for safety-constrained units according to claim 2, characterized in that, In the passage of demand parameters In the step of predicting the values ​​of some variables, the property of probability is introduced. The prediction is for the probability that a certain variable will take the value 1. The calculation expression is as follows: In the formula, It is a sigmoid function, and These are model parameters in the real number field, i.e. Using abstract functions To distinguish different Model parameters , where abstract functions Defined as: , Requirements parameters Dimensions For the optimal solution set The dimension of the variable is determined by using artificial intelligence methods to fit this parameter and predict the probability that the variable takes the value 1.

4. The hybrid integer programming joint optimization method for safety-constrained units according to claim 3, characterized in that, The partial variables of the optimal solution Using model parameters Calculate according to the formula , ,in, , These are the model parameters and demand parameters determined based on the required unit combination model.

5. A hybrid integer programming joint optimization system for safety-constrained units, characterized in that, include: The model processing module is used to process the pre-built safety-constrained unit combination optimization model to obtain a mixed-integer linear programming model. The variable fixing module is used to fix multiple variables in a mixed integer linear programming model based on a priori rules using a two-stage multivariate branching method. The optimal solution prediction module is used to acquire solution information during the branch and bound process of the multivariate branch method using artificial intelligence methods, and to learn the value of the optimal solution and the pattern of related parameters, thereby predicting the value of some variables of the optimal solution. The variable fixing module when the variable When the probability of taking 1 is high, take When the variable When the probability of taking 0 is high, take Use machine learning methods to fit the corresponding model parameters. After that, with Use a threshold value to ensure that the predicted variable value is greater than a certain threshold. Take 1, which is less than the threshold. When the value is 0, the first stage of fixing the variable is completed, and the expression is as follows: ,when and fixed , ,in , Requirements parameters dimensionality ; In the branch and bound process of the multivariate branch method, when the optimal solution prediction module reaches an intermediate threshold, it adds an MVB cutting plane in the second stage and expands the feasible region of the problem. Specifically, this includes: Branching based on infeasibility: Branch upwards using the following formula: and in , These are the two cutting planes corresponding to the upward branches. These are slack variables for the upward branch; Branch downwards using the following formula: and in , These are the two cutting planes corresponding to the downward branches. These are slack variables for the downward branch; Solve the original problem and add the cutting planes in the following order: a) , ； b) , and , ; c) 、 。 6. The hybrid integer programming joint optimization system for safety-constrained units according to claim 5, characterized in that, The model processing module processes parameters based on demand parameters. A defined safety-constrained unit combination optimization model, with a given abstract function. ,exist and ,in, The set of optimal solutions. For set The mapping function, for The mapping function, , All are optimal solution sets A subset of such that: Based on the safety constraints of the unit characteristics, and with the demand parameters The variables vary within a certain interval, and the partial variables of the optimal solution are obtained by solving for these variables. The parameter remains unchanged, meaning it is always either 1 or 0. Therefore, we assume the requirement parameter... There is a correlation between the optimal solution and the requirement parameters. To predict the values ​​of some variables.

7. The hybrid integer programming joint optimization system for safety-constrained units according to claim 6, characterized in that, The model processing module uses demand parameters. When predicting the values ​​of some variables, we introduce the properties of probability, predicting the probability that a certain variable will take the value 1. The calculation expression is as follows: In the formula, It is a sigmoid function, and These are model parameters in the real number field, i.e. Using abstract functions To distinguish different Model parameters , where abstract functions Defined as: , Requirements parameters Dimensions For the optimal solution set The dimension of the variable is determined by using artificial intelligence methods to fit this parameter and predict the probability that the variable takes the value 1.

8. The hybrid integer programming joint optimization system for safety-constrained units according to claim 7, characterized in that, The model processing module uses model parameters. Calculate the partial variables of the optimal solution using the following formula : In the formula, , , These are the model parameters and demand parameters determined based on the required unit combination model.

9. 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 steps of the hybrid integer programming joint optimization method for the safety-constrained unit as described in any one of claims 1 to 4.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, it implements the steps of the hybrid integer programming joint optimization method for the safety-constrained unit as described in any one of claims 1 to 4.

Images