Day-ahead reactive power optimization method for traction power supply system and related device

Abstract

The invention discloses a day-ahead reactive power optimization method for a traction power supply system and a related device, and belongs to the technical field of rail transit and power system dispatching automation. The method comprises the following steps: determining decision variables according to topological structures of a rail transit traction power supply network and a power distribution network and the number of actually deployed OLTC and CB; constructing a traction power supply system reactive power optimization model based on a particle swarm algorithm; collecting day-ahead prediction data of a load and a plurality of distributed photovoltaic systems as model input; improving population initialization and inertia weight of a particle swarm algorithm, and solving the traction power supply system reactive power optimization model based on the particle swarm algorithm to obtain an action strategy of each reactive power compensation device; and determining the action time of each reactive compensation device by adopting a Fisher optimal segmentation method, and finally obtaining a reactive optimization result. According to the invention, by scientifically dividing time periods and enhancing the search efficiency of the algorithm, efficient reactive power optimization in the day-ahead time dimension is achieved on the premise of ensuring stable operation of equipment.

Description

Technical Field

[0001] The present invention belongs to the technical field of rail transit and power system dispatching automation, and relates to a method for optimizing reactive power of a traction power supply system and a related device. Background Art

[0002] In the rail transit traction power supply system, reactive power optimization is a key link to ensure the stable operation of trains and improve power supply efficiency. With the rapid development of urban rail transit systems and the widespread application of power electronics equipment, the reactive power demand of traction power supply systems is increasing. Reactive power is not only crucial to maintaining the voltage stability of the power supply network, but also directly affects the energy efficiency and service life of traction equipment. Efficient reactive power optimization strategies can significantly reduce power loss, improve the economy and reliability of the power supply system, and thus better meet the growing operational needs of rail transit. Therefore, the development of advanced reactive power optimization technology suitable for rail transit traction power supply systems is of great significance to improving the overall system performance.

[0003] Traditional reactive power optimization methods, such as linear programming and nonlinear programming, have obvious limitations in the application of rail transit traction power supply systems. These methods often fail to fully consider the special operating constraints of traction power supply equipment, especially the use restrictions of OLTC (On-Load Tap Changer) and CB (Capacitor Bank) under frequent adjustments, which can easily lead to frequent equipment operation, aggravate equipment wear, and increase the risk of failure. At the same time, the traditional methods have high computational complexity and are difficult to adapt to the rapid changes in rail transit loads and complex operating scenarios, which limits their application effect in real-time control and dynamic optimization.

[0004] Therefore, the current reactive power optimization strategy still faces significant challenges in the actual application of rail transit traction power supply system. It is not only difficult to effectively control the action frequency of OLTC and CB, but also easy to fall into the local optimal solution, resulting in poor optimization results. At the same time, the traditional strategy has obvious shortcomings in calculation speed and global optimization ability, making it difficult to achieve efficient and accurate reactive power control in the complex traction power supply network. Summary of the invention

[0005] The purpose of the present invention is to provide a method and related devices for optimizing the reactive power of a traction power supply system a day ahead, so as to solve the technical problems that the reactive power optimization strategies in the prior art have high calculation complexity, low accuracy, and are difficult to adapt to the rapid changes in rail transit load and complex operating scenarios.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a method for optimizing reactive power of a traction power supply system on the day before, comprising the following steps: Define decision variables based on the topological structures of the rail transit traction power supply network and the distribution network, as well as the actual number of OLTCs and CBs deployed. Construct a reactive power optimization model for the traction power supply system based on the particle swarm optimization algorithm with the decision variables. Collect the day-ahead prediction data of the load and multiple distributed photovoltaic systems, and input them into the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm. Improve the population initialization and inertia weight of the particle swarm optimization algorithm, and solve the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm to obtain the action strategies of each reactive power compensation device. Regard the action strategies of each reactive power compensation device as a sample sequence for ordered clustering, and use the Fisher optimal segmentation method to determine the action moments of each reactive power compensation device, and finally obtain the reactive power optimization result.

[0007] Furthermore, the step of defining decision variables based on the topological structures of the rail transit traction power supply network and the distribution network, as well as the actual number of OLTCs and CBs deployed, specifically includes: According to the topological structures of the rail transit traction power supply network and the distribution network, analyze the connection methods of the nodes, lines and equipment in the topological structure; determine the actual number of OLTCs and CBs deployed in the distribution network. Based on the number of OLTCs and CBs, define decision variables and clarify the total number of decision variables. Set the value ranges and constraint conditions for each decision variable.

[0008] Furthermore, the step of constructing a reactive power optimization model for the traction power supply system based on the particle swarm optimization algorithm with the decision variables, specifically includes: Establish the distribution network power flow constraint, the distribution network operation safety constraint, and the OLTC and CB operation constraints. Based on the distribution network operation reliability, define the voltage deviation variance index; based on the distribution network operation economy, define the expected network loss index and the reactive power compensation device switching cost index; construct an objective function according to the voltage deviation variance index, the expected network loss index and the reactive power compensation device switching cost index to obtain the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm.

[0009] Furthermore, the expression of the distribution network power flow constraint is:

[0010] In the formula, is the active power injected at node ; is the reactive power injected at node ; is node voltage; is the conductance between node and node ; is the susceptance between node and node ; is the phase angle difference between node and node ; is the number of branches; The expression of the operation safety constraint of the distribution network is:

[0011] In the formula, is the maximum allowable voltage; is the minimum allowable voltage; is the upper limit value of the branch current; is the branch current; is the voltage of node ; The expression of the operation constraints of the OLTC and CB is:

[0012]

[0013] In the formula, is the reactive power of the CB in node at time t; is the compensation power of each group of CBs; is the switching group number of the CB at time t; is the upper limit value of the CB at time t; is the lower limit value of the CB at time t; is the actual transformation ratio of the OLTC in branch at time t; is the upper limit value of the OLTC in branch at time t; is the lower limit value of the OLTC in branch at time t; is the tap position of the OLTC at time t; is the adjustment step of the OLTC transformation ratio.

[0014] Furthermore, the voltage variance index is:

[0015] In the formula, is the total number of nodes, is the total number of time periods, is the voltage of node at time The voltage of is the rated voltage of node ; is the maximum allowable voltage; is the minimum allowable voltage; The expected network loss index is:

[0016] In the formula, is the number of branches; is the active power of the branch where node is located at time is the reactive power of the branch where node is located at time is the resistance of the branch where node is located at time is the voltage at the head of the branch where node is located at time is the total number of time periods; The switching cost index of the reactive power compensation device is:

[0017] In the formula, is the single action cost of the OLTC; is the single action cost of the CB, is the number of actions of the OLTC; is the number of actions of the CB.

[0018] Furthermore, the steps of improving the population initialization and inertia weight of the particle swarm algorithm specifically include: Performing Bernoulli shift chaos initialization on the population of the particle swarm algorithm through the Bernoulli shift chaos equation; the Bernoulli shift chaos equation is:

[0019] In the formula, is the chaos sequence generated between 0 and 1; is the next chaos sequence generated according to is a constant; Suppose the value range of a certain variable is , then the corresponding initial value generated is:

[0020] In the formula, is the initial value of the variable; is the lower limit of the variable's value; is the upper limit of the variable's value; Based on the symmetric Sigmoid curve, the inertia weight of the particle swarm algorithm is dynamically adjusted. At the th iteration, the inertia weight is:

[0021] In the formula, is the minimum weight; is the maximum weight; is the current iteration number; is the maximum iteration number.

[0022] Furthermore, the steps of using the Fisher optimal segmentation method to determine the action time of each reactive power compensation device specifically include: Define the optimal segmentation scheme; divide the sample sequence of n ordered clusters into m parts, and the segmentation points are represented by , then the segmentation result is expressed as:

[0023] Among all the segmentation schemes there is an optimal segmentation scheme that makes the objective function value the smallest ; Define the objective function ; The goal of optimal segmentation is to minimize the sum of the within-class scatter sums of each part. The objective function of any segmentation scheme is expressed as:

[0024] In the formula, represents the within-class scatter sum of the cth part, The calculation formula of

[0025] In the formula, represents the value of the qth sample; represents the segmentation point of the (c + 1)th part; represents the segmentation point of the cth part; represents the sample mean of the cth part, The calculation formula of

[0026] In the formula, Indicates the segmentation point of the (c + 1)-th part; The Fisher optimal segmentation method is adopted, and the recursive method is combined to solve the objective function. The recursive formula adopted by the recursive method is:

[0027] In the formula, Indicates the segmentation point The objective function value of the optimal (m - 1)-segmentation of all samples before Indicates the segmentation point The sum of squared deviations of the samples after to the n-th sample; Indicates the objective function value of dividing the first (q - 1) samples into (m - 1) segments; Indicates the sum of squared deviations of the q-th sample to the n-th sample; Indicates the m-th optimal segmentation point; According to the recursive formula, first calculate the optimal two-segmentation of the first b samples , and the specific calculation formula is:

[0028] In the formula, the value range of b is ; Successively solve the optimal three-segmentation of the b samples, the optimal four-segmentation , …, the optimal (m - 1)-segmentation ; On the basis of the optimal (m - 1)-segmentation scheme, solve the optimal m-segmentation scheme of the first n sample sequences according to the recursive formula , and determine the corresponding m-th optimal segmentation point Based on the m-th optimal segmentation point , solve the (m - 1)-th optimal segmentation point to make it satisfy:

[0029] Finally, solve all the optimal segmentation points to obtain the action moments of each reactive power compensation device.

[0030] In a second aspect, the present invention provides a traction power supply system day-ahead reactive power optimization system, including: A decision variable determination module, configured to clarify the decision variables according to the topological structures of the rail transit traction power supply network and the distribution network, and the actual deployed numbers of OLTCs and CBs; A model construction module, configured to construct a reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm based on the decision variables; A data acquisition module, configured to collect the day-ahead prediction data of the load and multiple distributed photovoltaic systems, and input the data into the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm; A model solving module, configured to improve the population initialization and inertia weight of the particle swarm optimization algorithm, and solve the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm to obtain the action strategies of each reactive power compensation device; A reactive power optimization module, configured to regard the action strategies of each reactive power compensation device as an ordered clustering sample sequence, determine the action moments of each reactive power compensation device by using the Fisher optimal segmentation method, and finally obtain the reactive power optimization result.

[0031] In a third aspect, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the day-ahead reactive power optimization method for a traction power supply system are implemented.

[0032] In a fourth aspect, the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the steps of the day-ahead reactive power optimization method for a traction power supply system are implemented.

[0033] Compared with the prior art, the present invention has the following beneficial effects: The present invention discloses a day-ahead reactive power optimization method and related devices for a traction power supply system. By reasonably dividing time periods and restricting the operation times of OLTC and CB, the frequent adjustment of equipment is effectively reduced, the service life of the equipment is extended, and at the same time, the optimization accuracy and calculation efficiency are improved. The improved particle swarm optimization algorithm not only enhances the global search ability but also accelerates the convergence speed and avoids the local optimum problem by introducing the chaotic population initialization and inertia weight adjustment methods. Combining the precise scheduling of the action moments of reactive power compensation devices by the Fisher optimal segmentation method further improves the reliability and safety of the traction power supply system. In the practical application of the rail transit traction power supply system, the present invention can effectively improve the real-time performance and flexibility of reactive power optimization, reduce the action frequency of equipment by adjusting optimization parameters, improve voltage stability and economy, and is of great significance for promoting the technological progress and sustainable development of the industry. Description of the Drawings

[0034] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0035] Figure 1 is the flowchart of the method of the present invention; Figure 2 is the schematic diagram of the system of the present invention; Figure 3 is the overall flowchart of the day-ahead reactive power optimization method for a traction power supply system according to an embodiment of the present invention; Figure 4 is the model network topology diagram of the embodiment of the present invention; Figure 5 is the schematic diagram of the structure of the computer device of the present invention. Detailed implementation manners

[0036] The present invention will be described in detail below with reference to the drawings and in conjunction with embodiments. It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments may be combined with each other.

[0037] The following detailed descriptions are all exemplary descriptions, aiming to provide further detailed descriptions of the present invention. Unless otherwise specified, all technical terms adopted by the present invention have the same meaning as commonly understood by those of ordinary skill in the art to which the present application belongs. The terms used in the present invention are only for the purpose of describing specific implementation manners, and are not intended to limit the exemplary embodiments according to the present invention.

[0038] Refer to Figure 1 and Figure 3 , an embodiment of the present invention discloses a day-ahead reactive power optimization method for a traction power supply system, including the following steps: S1. Based on the topological structures of the rail transit traction power supply network and the distribution network, as well as the actual deployed numbers of OLTCs and CBs, clarify the decision variables; S101. According to the topological structures of the rail transit traction power supply network and the distribution network, analyze the connection modes of the nodes, lines, and devices in the topological structure; determine the actual deployed numbers of OLTCs and CBs in the distribution network.

[0039] S102. Based on the numbers and types of devices, define decision variables, such as the regulation gears of each OLTC, the switching groups of CBs, etc., and clarify the total number of decision variables.

[0040] S103. Set reasonable value ranges and constraint conditions for each decision variable to ensure the effectiveness of the subsequent optimization process.

[0041] S2. Refer to Figure 4 , and based on the decision variables, construct a reactive power optimization model for the traction power supply system based on the particle swarm optimization algorithm; S201. Establish the distribution network power flow constraint, the distribution network operation safety constraint, and the operation constraints of OLTCs and CBs; (1) Distribution network power flow constraint:

[0042] Wherein, is the active power injected at node ; is the reactive power injected at node ; is the voltage of node ; is the conductance between node and node ; is the susceptance between node and node ; is the phase angle difference between node and node ; is the number of branches.

[0043] (2) Operating safety constraints of the distribution network:

[0044] Wherein, is the maximum allowable voltage; is the minimum allowable voltage; is the upper limit value of the branch current; is the branch current; is the voltage of node ;

[0045] (3) Operating constraints of OLTC and CB:

[0046]

[0047] Wherein, is the reactive power of CB at node at time t; is the compensation power of each group of CB; is the switching group number of CB at time t; is the upper limit value of CB at time t; is the lower limit value of CB at time t; is the actual transformation ratio of OLTC in branch at time t; is the upper limit value of OLTC in branch at time t; is the lower limit value of OLTC in branch at time t; is the tap position of OLTC at time t; is the adjustment step of the OLTC transformation ratio.

[0048] S202. Construct an objective function based on the voltage deviation variance index, the expected network loss index, and the reactive power compensation equipment switching cost index to obtain a reactive power optimization model for the traction power supply system based on the particle swarm algorithm.

[0049] (1) Define the voltage deviation variance index based on the operation reliability of the distribution network , and the formula is:

[0050] In the formula, is the total number of nodes, is the total number of time periods, is the voltage of node at time node rated voltage; is the maximum allowable voltage; is the minimum allowable voltage.

[0051] (2) Define the expected network loss index and the reactive power compensation equipment switching cost index based on the operation economy of the distribution network, and the formulas are respectively:

[0052] In the formula, is the number of branches, is the active power of the branch where node is located at time is the reactive power of the branch where node is located at time is the resistance of the branch where node is located at time is the voltage at the head end of the branch where node is located at time is the total number of time periods.

[0053]

[0054] In the formula, is the single action cost of the OLTC; is the single action cost of the CB, is the action times of the OLTC; is the action times of the CB.

[0055] S3. Collect the day-ahead prediction data of the load and multiple distributed photovoltaic systems, and input them into the reactive power optimization model of the traction power supply system based on the particle swarm algorithm; Clean and preprocess the collected data to ensure the accuracy and consistency of the data. Organize the data into a time series format for subsequent model input.

[0056] S4. Improve the population initialization and inertia weight of the particle swarm algorithm, and solve the reactive power optimization model of the traction power supply system based on the particle swarm algorithm to obtain the action strategies of each reactive power compensation device; The day-ahead reactive power / voltage optimization control model belongs to a typical mixed-integer nonlinear programming (MINLP) problem with high complexity. The traditional particle swarm algorithm is prone to falling into local optimal solutions during the solution process, and has a high computational complexity, making it difficult to meet the accuracy requirements of the actual power grid. Therefore, the population initialization and inertia weight of the particle swarm algorithm are correspondingly improved.

[0057] (1) Perform Bernoulli shift chaotic initialization on the population of the particle swarm algorithm through the Bernoulli shift chaotic equation; for the nonlinear programming problem with multiple constraints, multiple dimensions, and multiple peaks, there is a low convergence accuracy, and randomly generated and uneven initial populations will cause the particle swarm to be unable to comprehensively search the solution space and may fall into local optimality and exhibit the "premature" phenomenon. Using chaotic mapping for initialization, compared with the typical Logistic chaotic system, the Bernoulli shift chaotic system has a more uniform distribution on [0,1]. The Bernoulli shift chaotic equation is:

[0058] In the formula, is the chaotic sequence generated between 0 and 1; is the next chaotic sequence generated according to ; is a constant.

[0059] Suppose the value range of a certain variable is , then the corresponding initial value is

[0060] In the formula, is the initial value of the variable; is the lower limit of the variable's value; is the upper limit of the variable's value.

[0061] (2)Dynamic adjustment of inertia weight based on symmetric Sigmoid curve. When solving practical optimization problems, it is often desired to first perform global search to quickly converge the search space to a certain region, and then perform local fine search to obtain a high-precision solution. Therefore, an adaptive adjustment strategy is proposed, that is, as the iteration progresses, the value is correspondingly reduced. The symmetric Sigmoid curve is used to dynamically adjust the inertia weight . The inertia weight at the th iteration is

[0062] where, is the minimum weight; is the maximum weight; is the current iteration number; is the maximum iteration number.

[0063] S5. Regarding the action strategies of each reactive power compensation device as an ordered clustering sample sequence, the Fisher optimal segmentation method is used to determine the action moments of each reactive power compensation device, and finally the reactive power optimization result is obtained.

[0064] S501. Define the optimal segmentation scheme. Divide the n ordered clustering sample sequences into m parts, and the segmentation points are represented by . Then the segmentation result can be expressed as:

[0065] Among all the segmentation schemes , there must be a segmentation scheme that minimizes the objective function value.

[0066] S502. Define the objective function .

[0067] The goal of optimal segmentation is to minimize the sum of the within-class scatter matrices of each part. The objective function of any segmentation scheme can be expressed as:

[0068] where, represents the within-class scatter matrix of the c-th part, and the calculation formula is:

[0069] where, represents the value of the q-th sample; represents the segmentation point of the c + 1-th part; represents the segmentation point of the c-th part; represents the sample mean of the c-th part, The calculation formula of

[0070] In the formula, represents the segmentation point of the c-th part; represents the segmentation point of the (c + 1)-th part; represents the value of the q-th sample; S503. The Fisher optimal segmentation method is used to solve the objective function in combination with the recursive method.

[0071] The Fisher optimal segmentation method uses the sum of squared deviations to represent the degree of difference between samples of the same category. By recursive calculation, the optimal segmentation point is determined to minimize the difference between samples of the same category and maximize the difference between samples of different categories. Therefore, the key to the recursive method lies in the solution of the segmentation point solution.

[0072] It can be seen from the definition of the objective function that on the basis of the optimal m - 1 segmentation scheme of the sample sequence adding the last segment can obtain the optimal m segmentation scheme , and its recursive formula is as follows:

[0073] In the formula, represents the segmentation point The objective function value of the optimal m - 1 segmentation of all samples before represents the segmentation point The sum of squared deviations of the samples after to the n-th sample; represents the objective function value of dividing the first q - 1 samples into m - 1 segments; represents the m-th optimal segmentation point; According to the recursive formula, first calculate the optimal two-segmentation of the first b samples,

[0074] In the formula, the value range of b is .

[0075] Similarly, solve the optimal three-segmentation of the b samples, the optimal four-segmentation in turn, and so on, until the optimal m - 1 segmentation .

[0076] Based on the optimal m - 1 segmentation scheme, solve the optimal m - segmentation scheme for the first n sample sequences according to the recursive formula , and determine the corresponding m - th optimal segmentation point .

[0077] Based on step , solve the (m - 1) - th optimal segmentation point , so that it satisfies:

[0078] By analogy, all optimal segmentation points can be solved to obtain the action moments of each reactive power compensation device.

[0079] In dynamic reactive power optimization, the action sequences of OLTC and capacitor banks can both be regarded as an ordered clustering sample sequence. The maximum number of actions allowed by the power grid for discrete reactive power compensation devices is the optimal number of segmentations, and the action moments are the optimal segmentation points. The mean value of each class is used as the action value for all time periods within the class. In addition, since the action values of reactive power compensation devices are discrete integers, the integer value closest to the mean is taken as the actual action scheme, and finally the reactive power optimization result is obtained.

[0080] Refer to Figure 2 , an embodiment of the present invention discloses a traction power supply system day - ahead reactive power optimization system, including a decision variable determination module, a model construction module, a data acquisition module, a model solution module, and a reactive power optimization module.

[0081] Among them, the decision variable determination module is used to clarify the decision variables based on the topological structure of the rail transit traction power supply network, the distribution network, and the actual number of deployed OLTC and CB; the model construction module is used to construct a reactive power optimization model of the traction power supply system based on particle swarm optimization algorithm based on the decision variables; the data acquisition module is used to collect the day - ahead prediction data of the load and multiple distributed photovoltaic systems and input them into the reactive power optimization model of the traction power supply system based on particle swarm optimization algorithm; the model solution module is used to improve the population initialization and inertia weight of the particle swarm optimization algorithm and solve the reactive power optimization model of the traction power supply system based on particle swarm optimization algorithm to obtain the action strategies of each reactive power compensation device; the reactive power optimization module is used to regard the action strategies of each reactive power compensation device as an ordered clustering sample sequence, and use the Fisher optimal segmentation method to determine the action moments of each reactive power compensation device, and finally obtain the reactive power optimization result.

[0082] In one embodiment of the present invention, refer to Figure 5, a computer device is provided. The computer device includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the day-ahead reactive power optimization method for the traction power supply system.

[0083] The present invention also provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is the memory device in the computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, and this storage space stores the operating system of the terminal. And, one or more instructions suitable for being loaded and executed by the processor are also stored in this storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be high-speed RAM (Random Access Memory), or non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the day-ahead reactive power optimization method for the traction power supply system in the above embodiments.

[0084] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. 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. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM (Compact Disc Read-Only Memory), optical memory, etc.) that contain computer-usable program code.

[0085] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce means for implementing the functions specified in Figure 1 one or more of the processes Figure 1 or blocks.

[0086] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in Figure 1 one or more of the processes Figure 1 or blocks.

[0087] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are performed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one or more of the processes Figure 1 or blocks.

[0088] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific implementation manners of the present invention, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A method for optimizing the daily reactive power of a traction power supply system, characterized in that, It includes the following steps: Based on the topological structures of the rail transit traction power supply network and the distribution network, as well as the actual number of OLTCs and CBs deployed, determine the decision variables; Based on the decision variables, construct a reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm; Collect the day-ahead prediction data of the load and multiple distributed photovoltaic systems, and input them into the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm; Improve the population initialization and inertia weight of the particle swarm optimization algorithm, and solve the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm to obtain the action strategies of each reactive power compensation device; Regard the action strategies of each reactive power compensation device as a sample sequence for ordered clustering, and use the Fisher optimal segmentation method to determine the action moments of each reactive power compensation device, and finally obtain the reactive power optimization result.

2. The day-ahead reactive power optimization method for a traction power supply system according to claim 1, wherein The step of determining the decision variables based on the topological structures of the rail transit traction power supply network and the distribution network, as well as the actual number of OLTCs and CBs deployed, specifically includes: According to the topological structures of the rail transit traction power supply network and the distribution network, analyze the connection methods of the nodes, lines and equipment in the topological structure; determine the actual number of OLTCs and CBs deployed in the distribution network; Based on the number of OLTCs and CBs, define the decision variables and clarify the total number of decision variables; Set the value ranges and constraint conditions for each decision variable.

3. A method for optimizing the daily reactive power of a traction power supply system according to claim 1, characterized in that The step of constructing a reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm based on the decision variables, specifically includes: Establish the distribution network power flow constraint, the distribution network operation safety constraint, and the OLTC and CB operation constraints; Based on the distribution network operation reliability, define the voltage deviation variance index; based on the distribution network operation economy, define the expected network loss index and the reactive power compensation device switching cost index; construct the objective function according to the voltage deviation variance index, the expected network loss index and the reactive power compensation device switching cost index to obtain the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm.

4. A method for optimizing the reactive power of a traction power supply system on a daily basis according to claim 3, characterized in that, The expression of the distribution network power flow constraint is: Wherein, is the active power injected at node ; is the reactive power injected at node ; is the voltage of node ; is the conductance between node and node ; is the susceptance between node and node ; is the phase angle difference between node and node ; is the number of branches; The expression of the distribution network operation safety constraint is: Wherein, is the maximum allowable voltage; is the minimum allowable voltage; is the upper limit value of the branch current; is the branch current; is the node voltage; The expression of the OLTC and CB operation constraints is: Wherein, is the reactive power of CB in node at time t; is the compensation power of each group of CB; is the switching group number of CB at time t; is the upper limit value of CB at time t; is the lower limit value of CB at time t; is the actual transformation ratio of OLTC in branch at time t; is the upper limit value of OLTC in branch at time t; is the lower limit value of OLTC in branch at time t; is the tap position of OLTC at time t; is the adjustment step of the OLTC transformation ratio.

5. A method for optimizing the reactive power of a traction power supply system on a daily basis according to claim 3, characterized in that, The voltage deviation variance index is as follows: Wherein, is the total number of nodes, is the total number of time periods, is the voltage of node at time is the rated voltage of node ; is the maximum allowable voltage; is the minimum allowable voltage; The expected network loss index is as follows: In the formula, is the number of branches; is the active power of the branch where the node at time is located; is the reactive power of the branch where the node at time is located, is the resistance of the branch where the node at time is located; is the voltage at the head end of the branch where the node at time is located; is the total number of time periods; The switching cost index of the reactive power compensation device is as follows: wherein, is the single - action cost of the OLTC; is the single - action cost of the CB, is the number of actions of the OLTC; is the number of actions of the CB.

6. The day-ahead reactive power optimization method for a traction power supply system according to claim 1, wherein The step of improving the population initialization and inertia weight of the particle swarm optimization algorithm, specifically includes: Perform Bernoulli shift chaotic initialization on the population of the particle swarm optimization algorithm through the Bernoulli shift chaotic equation; the Bernoulli shift chaotic equation is: wherein, is the generated chaotic sequence between 0 and 1; is the next chaotic sequence generated according to ; is a constant; Let the value range of a certain variable be , then the corresponding initial value is: In the formula, is the initial value of the variable; is the lower limit of the variable's value; is the upper limit of the variable's value; Dynamic adjustment of the inertia weight of the particle swarm optimization algorithm based on a symmetric sigmoid curve. The inertia weight at the th iteration is as follows: Wherein, is the minimum weight; is the maximum weight; is the current iteration number; is the maximum iteration number.

7. A method for optimizing the daily reactive power of a traction power supply system according to claim 1, characterized in that, The step of using the Fisher optimal segmentation method to determine the action moments of each reactive power compensation device, specifically includes: Define the optimal segmentation scheme; for the sample sequence of n ordered clusters be segmented into m parts, and the segmentation points are represented by , then the segmentation result is expressed as: All segmentation schemes There exists an optimal segmentation scheme that minimizes the objective function value ; Define the objective function ; The goal of the optimal segmentation is to minimize the sum of the sum of squared deviations of each part. The objective function of any segmentation scheme is expressed as: In the formula, represents the sum of squared deviations of the c-th part, The calculation formula of is: In the formula, represents the value of the q-th sample; represents the segmentation point of the (c + 1)-th part; represents the segmentation point of the c-th part; represents the sample mean of the c-th part, and its calculation formula is: In the formula, represents the segmentation point of the (c + 1)-th part; Use the Fisher optimal segmentation method to solve the objective function in combination with the recursive method, and the recursive formula used by the recursive method is: In the formula, represents the segmentation point and the objective function value of the optimal m - 1 segmentation of all samples before it; represents the segmentation point and the sum of squared deviations of the samples after it to the nth sample; represents the objective function value of dividing the first q - 1 samples into m - 1 segments; represents the sum of squared deviations of the qth sample to the nth sample; represents the mth optimal segmentation point; First, calculate the optimal binary partition of the first b samples according to the recursive formula , and the specific calculation formula is as follows: wherein, the value range of b is ; Successively solve the optimal three-way partitions of b samples and the optimal four-way partitions and so on, up to the optimal (m - 1)-way partitions ; Based on the optimal m-1 segmentation scheme, solve the optimal m segmentation scheme for the first n sample sequences according to the recurrence formula and determine the corresponding m-th optimal segmentation point ; Based on the m-th optimal splitting point , solve for the (m - 1)-th optimal splitting point such that it satisfies: Finally, solve all the optimal segmentation points to obtain the action moments of each reactive power compensation device.

8. A day-ahead reactive power optimization system for a traction power supply system, characterized in that, It includes: A decision variable determination module, which is used to determine the decision variables based on the topological structures of the rail transit traction power supply network and the distribution network, as well as the actual number of OLTCs and CBs deployed; A model construction module, which is used to construct a reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm based on the decision variables; A data acquisition module, configured to collect the day-ahead prediction data of the load and multiple distributed photovoltaic systems, and input the data into the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm; A model solving module, configured to improve the population initialization and inertia weight of the particle swarm optimization algorithm, and solve the reactive power optimization model of the traction power supply system based on the particle swarm optimization algorithm to obtain the action strategies of each reactive power compensation device; A reactive power optimization module, configured to regard the action strategies of each reactive power compensation device as a sample sequence for ordered clustering, and use the Fisher optimal segmentation method to determine the action moments of each reactive power compensation device, and finally obtain the reactive power optimization result.

9. A computer 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, the steps of a day-ahead reactive power optimization method for a traction power supply system according to any one of claims 1-7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the steps of a day-ahead reactive power optimization method for a traction power supply system according to any one of claims 1-7 are implemented.

Images