Generalized unit commitment calculation method and system considering low-voltage resources and grid constraints

Abstract

The application discloses a kind of generalized unit combination calculation method and system considering low-voltage resource and power grid constraint, belong to electric power system field, method includes obtaining each node net load of each moment corresponding to power grid;Each node net load of each moment corresponding to power grid is used as input data, and each moment voltage and current are converted into safety distance as label, and the neural network established in advance is trained;Objective function and constraint condition of unit combination are constructed, and the neural network trained is used to replace constraint condition;The objective function of unit combination is solved using the neural network of replacement constraint condition, and unit combination calculation is completed.The application uses the neural network trained to replace traditional second-order cone constraint, and the good constraint is replaced into the unit combination of low-voltage level power grid, effectively simplifies the calculation process, greatly speeds up the speed of unit combination solution, significantly improves the calculation efficiency, can more timely provide decision support for power grid operation.

Description

Technical Field

[0001] This invention belongs to the field of power systems, and specifically relates to a generalized unit combination calculation method and system that takes into account low-voltage resources and grid constraints. Background Technology

[0002] Traditional methods for calculating unit combination in low-voltage power grids require taking all constraints into account. Low-voltage power grids, according to both domestic and international standards, refer to AC grids ≤ 1kV. Because low-voltage power grids contain numerous secondary constraints, this significantly reduces calculation speed.

[0003] When optimizing power generation plans, traditional unit combinations typically treat low-voltage grids as equivalent to fixed loads, ignoring their internal power flow constraints and distributed energy back-feeding capabilities. This can lead to coordination mismatches, voltage and current overruns, and even large-scale power outages.

[0004] In recent years, with the high penetration of distributed energy resources, low-voltage power grids have shifted from "passive" to "active," necessitating the explicit consideration of power flow constraints in unit configuration. However, directly employing traditional second-order cone constraints presents challenges due to multi-level, multi-scenario repeated interactions, resulting in time-consuming solutions that grow exponentially with the number of low-voltage grid connections, making it difficult to meet the timeliness requirements of day-ahead and even real-time dispatch. To address this, some studies have attempted to reduce the constraint scale using linearization or clustering methods, but these still require numerous iterations and suffer significant accuracy loss. Other researchers have used neural networks to learn the feasible power flow region offline, but have failed to transform the learning results into explicit linear constraints that can be embedded in mixed-integer linear programming, rendering them unsuitable for direct use in existing unit configuration solvers. Summary of the Invention

[0005] The purpose of this invention is to address the problem in the existing technology that the current generation unit combination cannot quickly consider the secondary constraints of the low-voltage grid after the integration of a high proportion of distributed energy resources. The invention provides a generalized generation unit combination calculation method and system that takes into account both low-voltage resources and grid constraints. By using a trained neural network to replace the traditional second-order cone constraints, the calculation speed of generation unit combination in low-voltage grids is accelerated.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] Firstly, a generalized unit combination calculation method considering low-voltage resources and grid constraints is provided, including:

[0008] Obtain the net load of each node in the power grid at each time point;

[0009] The net load of each node in the power grid at each time point is used as input data, and the voltage and current at each time point are converted into safe distances as labels to train a pre-established neural network.

[0010] Construct the objective function and constraints for unit combination, and replace the constraints with a trained neural network;

[0011] The objective function of unit combination is solved by using a neural network with alternative constraints, thus completing the unit combination calculation.

[0012] As a preferred embodiment, the step of converting voltage and current at various times into safe distances for use as tags is calculated using the following expression:

[0013]

[0014]

[0015] In the formula, It is a set of nodes. yes t Safe distance of voltage at all times yes t The safe distance for current at all times. It is a collection of routes. This is the permissible voltage fluctuation. It is the first i Each node t Voltage amplitude at time 10:00 yes i - j On the line t Current amplitude at time , yes i - j On the line t The maximum current at time t. It is a collection of moments;

[0016] When the safe distance meets the following condition, there will be no over-limit situation in the current and voltage of the power grid:

[0017] .

[0018] As a preferred embodiment, the pre-built neural network includes an input layer, K hidden layers, and an output layer, using ReLU as the activation function. In the formula, For the first k The input of a layered neural network, For the first k Layer weights, For the first k The output of the layer, For the first k Layer bias;

[0019] The Big M method is used to transform the maximization operator into a mixed-integer linear form, resulting in the multilayer perceptron (MLP) expression:

[0020]

[0021] In the formula, This is the input to the first layer of the neural network. It is the input to the neural network. t For a moment, It is a collection of moments; For the first k The output of the layer, For the first k Layer weights, For the first k The input of a layered neural network, For the first k The input to the -1 layer of the neural network, k Indicates a hidden layer. A collection of hidden layers. and As an auxiliary variable, Representing large numbers, This refers to the safe distance for voltage and current.

[0022] As a preferred approach, in the step of constructing the objective function and constraints for unit combination, an on-load tap-changing transformer is used to ensure that the voltage of the bus connecting the transmission network and the low-voltage grid remains at 1.0 pu at any given time, as expressed below:

[0023]

[0024] In the formula, Represents the voltage amplitude at the connection point at different times;

[0025] Meanwhile, the power demand of low-voltage power grids is consistent with the power supplied to them by the transmission grid, conforming to the following expression:

[0026]

[0027] In the formula, It is a collection of routes. i and j Both represent nodes. express t Time of the first i Active load on each node; It refers to the active power supplied by the transmission network to the low-voltage grid at different times. for t Time of the first j Photovoltaic power generation at each node; yest Time of the first i Reactive load on each node This refers to the reactive power generated by the reactive power compensation device at the busbar connection point. t For a moment, It is a collection of moments.

[0028] As a preferred approach, in the step of constructing the objective function and constraints of the unit combination, the objective function of the unit combination is to minimize the sum of the unit start-up and shutdown costs and fuel costs, as expressed below:

[0029]

[0030]

[0031]

[0032]

[0033] In the formula, It is a unit index variable. Indicates the number of thermal power units in the system; It is a time index variable. This represents the number of time periods within a scheduling cycle; This is a Boolean variable representing the unit. In the Start / stop status during a time period This indicates that the unit is in a stopped state. This indicates that the unit is in the powered-on state; This refers to the unit price of coal. It is a generator set In the Fuel consumption during the period and They represent the generating units. Start-up and downtime costs, , , These are the coefficients of the unit cost function. It is a generator set In the Power output during a given time period Indicates the unit In the The square of the output power during the time period, and They represent the generating units. The cost of a single start and stop, This is a Boolean variable representing the unit. In the Startup status of the time period Indicates the unit exist The machine is always in a shutdown state. The device is always powered on. A Boolean variable, representing the unit. In the The shutdown status during a certain period of time. Indicates the unit exist It is always powered on. It is always in a shutdown state.

[0034] As a preferred approach, the constraints in the steps of constructing the objective function and constraints of the unit combination include unit start-up and shutdown constraints, minimum start-up and shutdown time constraints, active power balance constraints, upper and lower limits of conventional unit output constraints, conventional unit ramping constraints, and power grid flow security constraints.

[0035] As a preferred embodiment, the unit start-stop constraint refers to the start-stop state. Startup status Shutdown status The three variables satisfy the following equation:

[0036]

[0037] In the formula, and These are Boolean variables, representing the units respectively. In the t Time period and t -1 indicates the start / stop status; a value of 0 indicates the unit is in a stopped state, and a value of 1 indicates the unit is in a started state.

[0038] The minimum start-up and shutdown time constraint of the unit refers to the start-up and shutdown state variables satisfying the following formula within a continuous number of hours after the unit performs a start-up or shutdown action:

[0039]

[0040]

[0041] In the formula, Indicates the unit In the k Power output during a given time period This represents the number of time periods within a scheduling cycle. Indicates the start-up time of the generator unit. Indicates the duration of the unit shutdown. This indicates iterating through every possible moment when the system might start or stop. Indicates traversal from t From now on, it is necessary to maintain the power on or off for every period of time;

[0042] The expression for the active power balance constraint is as follows:

[0043]

[0044] In the formula, It is a unit index variable. Indicates the number of thermal power units in the system; It is a wind farm index variable. Indicates the number of wind farms in the system. It is a generator set In the Power output during a given time period Indicates wind farm The predicted value, It is a power grid node index variable. Indicates the number of nodes in the system. Represents a node Load forecast;

[0045] The upper and lower limits of output for conventional generating units are expressed as follows:

[0046]

[0047] In the formula, and They are conventional units Maximum and minimum output;

[0048] Conventional unit ramp-up constraints include ramp-up constraints during unit operation and ramp-up constraints during startup and shutdown, expressed as follows:

[0049]

[0050]

[0051] In the formula, and These represent the upward and downward ramp constraints of the unit during operation, respectively. For the ramp-up constraints during the unit's startup process. This refers to the ramp-up constraints during the unit's shutdown process. It is a generator set In the t Power output during the -1 time period;

[0052] The power flow security constraint is expressed as follows:

[0053]

[0054] In the formula: This represents the upper limit of the power flow of the line. For nodes The load on the line Current transfer matrix For conventional units For the line Current transfer matrix For wind turbines For the line The trend transfer matrix.

[0055] Secondly, a generalized unit combination calculation system that takes into account low-voltage resources and grid constraints is provided, including:

[0056] The node net load acquisition module is used to acquire the net load of each node in the power grid at each time point.

[0057] The neural network training module is used to take the net load of each node of the power grid at each time as input data, and convert the voltage and current at each time into safe distance as labels to train the pre-established neural network.

[0058] The constraint substitution module is used to construct the objective function and constraints of the unit combination, and to substitute the constraints using a trained neural network.

[0059] The objective function solving module is used to solve the objective function of unit combination using a neural network with alternative constraints, and to complete the unit combination calculation.

[0060] Thirdly, an electronic device is provided, including a processor and a memory, the processor being configured to execute a computer program stored in the memory to implement the generalized unit combination calculation method taking into account low-voltage resources and grid constraints as described in the first aspect.

[0061] Fourthly, a computer-readable storage medium is provided, the computer-readable storage medium storing at least one instruction, which, when executed by a processor, implements the generalized unit combination calculation method taking into account low-voltage resources and grid constraints as described in the first aspect.

[0062] Fifthly, a computer program product is provided, the computer program product comprising a computer program, which, when executed by a processor, implements the generalized unit combination calculation method taking into account low-voltage resources and grid constraints as described in the first aspect.

[0063] Compared with the prior art, the first aspect of the present invention has at least the following beneficial effects:

[0064] This invention proposes a generalized unit combination calculation method that considers low-voltage resources and grid constraints. It replaces traditional second-order cone constraints with a trained neural network and applies this network to unit combination calculations involving low-voltage grids. Traditional methods require considering all constraints in unit combination calculations, resulting in complex and time-consuming processes. This invention, by replacing some constraints with a neural network, effectively simplifies the calculation process, significantly accelerates the unit combination solution, and substantially improves computational efficiency, enabling more timely decision support for grid operation. Unlike existing unit combination calculation methods that incorporate all constraints, this invention utilizes a neural network to substitute constraints. This computational model breaks through the limitations of traditional methods, providing a new approach and method for unit combination calculations involving low-voltage grids, and contributing to the implementation of 15-minute-level safety verification in areas with high renewable energy penetration rates. The method proposed in this invention is specifically designed for unit combination calculations involving low-voltage grids, precisely solving the computational challenges in this specific scenario. In actual grid operation, unit combination calculations for low-voltage grids are crucial for ensuring stable and economical grid operation. This invention can complete calculations quickly and accurately, providing power grid dispatchers with scientific and reasonable unit combination schemes. It has positive significance for accelerating the application of unit combination calculation problems in practice and helps to improve the operational efficiency and reliability of the entire power grid system. Attached Figure Description

[0065] 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.

[0066] Figure 1 Flowchart of the generalized unit combination calculation method considering low-voltage resources and grid constraints in an embodiment of the present invention;

[0067] Figure 2 A schematic diagram of the generalized unit combination calculation system structure that takes into account low-voltage resources and grid constraints in an embodiment of the present invention. Detailed Implementation

[0068] 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.

[0069] With the increasing penetration of distributed energy resources, power flow constraints of low-voltage grids must be explicitly considered in unit combination. However, traditional unit combination calculation algorithms for low-voltage grids take all constraints into account, which significantly reduces calculation speed due to the numerous secondary constraints present in low-voltage grids. For example, the traditional DistFlow model proposed by Baran and Wu in 1989 is a core mathematical tool for advanced applications such as low-voltage grid reconfiguration, reactive power optimization, and distributed source location. DistFlow establishes a nonlinear equation system between node power balance and branch power loss by treating each branch of the low-voltage grid as an equivalent "π-type" circuit. It uses node injected power, the square of branch current, and the square of node voltage as state variables to form a recursive power flow equation under a radial network structure, thus avoiding the convergence problem of the traditional Newton's method in ill-conditioned low-voltage grids. The DistFlow model's workflow is as follows: First, it generates a node-branch correlation matrix based on the network topology. Then, it calculates branch power losses by moving forward along the feeder from the end to the root node. Subsequently, it updates node voltages by moving back from the root node to the end, iterating until the power imbalance is less than a threshold. Finally, it outputs the voltage amplitude of each node, branch power, and network losses. For unit combination calculations involving low-voltage grids, the DistFlow model introduces many quadratic constraints, which significantly reduces the solution efficiency.

[0070] Please see Figure 1 This invention proposes a generalized unit combination calculation method that takes into account low-voltage resources and grid constraints, mainly including the following steps:

[0071] S1. Obtain the net load of each node in the power grid at each time point;

[0072] S2. Take the net load of each node of the power grid at each time as input data, and convert the voltage and current at each time into safe distance as labels to train the pre-established neural network.

[0073] S3. Construct the objective function and constraints for unit combination, and replace the constraints with a trained neural network;

[0074] S4. Solve the objective function of the unit combination using a neural network with alternative constraints to complete the unit combination calculation.

[0075] Classification-based methods require a sufficient number of samples located on the feasible or infeasible boundary to obtain an ideal solution, but most historical data samples are in normal operating condition and are usually far from the boundary. Therefore, in one possible implementation, step S2 uses a safe distance as the label for the neural network, converting the voltage and current at each time step into a safe distance using the following formula:

[0076]

[0077]

[0078] In the formula, It is a set of nodes. yes t Safe distance of voltage at all times yes t The safe distance for current at all times. It is a collection of routes. This refers to the allowable voltage fluctuation (assuming the allowable voltage fluctuation is [0.95pu, 1.05pu], then...). =0.05pu), It is the first i Each node t Voltage amplitude at time 10:00 yes i - j On the line t Current amplitude at time , yes i - j On the line t The maximum current at time t. It is a collection of moments;

[0079] The power grid's current and voltage will not exceed limits only when the safety distance meets the following condition:

[0080] .

[0081] In one possible implementation, the neural network pre-built in step S2 includes an input layer, K hidden layers, and an output layer, using ReLU as the activation function. In the formula, For the first k The input of a layered neural network, For the first k Layer weights, For the first k The output of the layer, For the first k The layer bias; however, due to the presence of an internal maximization operator, existing solvers struggle to process it directly. Therefore, this embodiment employs the Big M method to transform the maximization operator into a mixed-integer linear form, yielding the multilayer perceptron (MLP) expression:

[0082]

[0083] In the formula, This is the input to the first layer of the neural network. It is the input to the neural network. t For a moment, It is a collection of moments; For the first k The output of the layer, For the first k Layer weights, For the first k The input of a layered neural network, For the first k The input to the -1 layer of the neural network, k Indicates a hidden layer. A collection of hidden layers. and As an auxiliary variable, Representing large numbers, This refers to the safe distance for voltage and current.

[0084] In actual operation, on-load tap-changing transformers have been proven to maintain the grid voltage within a specified range, improve power quality, and reduce the impact of voltage fluctuations on user equipment. In one possible implementation, step S3, when constructing the objective function and constraints for unit combination, uses an on-load tap-changing transformer to ensure that the voltage of the bus connecting the transmission network and the low-voltage grid remains at 1.0 pu at any given time, as expressed below:

[0085]

[0086] In the formula, Represents the voltage amplitude at the connection point at different times;

[0087] Meanwhile, the power demand of low-voltage power grids is consistent with the power supplied to them by the transmission grid, conforming to the following expression:

[0088]

[0089] In the formula, It is a collection of routes. i and j Both represent nodes. express t Time of the first i Active load on each node; It refers to the active power supplied by the transmission network to the low-voltage grid at different times. for t Time of the first j Photovoltaic power generation at each node; yes t Time of the first i Reactive load on each node This refers to the reactive power generated by the reactive power compensation device at the busbar connection point.t For a moment, It is a collection of moments.

[0090] In one possible implementation, the objective function of unit combination is to minimize the sum of unit start-up and shutdown costs and fuel costs, as expressed below:

[0091]

[0092]

[0093]

[0094]

[0095] In the formula, It is a unit index variable. Indicates the number of thermal power units in the system; It is a time index variable. This represents the number of time periods within a scheduling cycle; This is a Boolean variable representing the unit. In the Start / stop status during a time period This indicates that the unit is in a stopped state; This indicates that the unit is in the powered-on state; This refers to the unit price of coal. It is a generator set In the Fuel consumption during the period and They represent the generating units. Start-up and downtime costs, , , These are the coefficients of the unit cost function. It is a generator set In the Power output during a given time period Indicates the unit In the The square of the output power during the time period, and They represent the generating units. The cost of a single start and stop, This is a Boolean variable representing the unit. In the Startup status of the time period Indicates the unit exist The machine is always in a shutdown state. The device is always powered on. A Boolean variable, representing the unit. In the The shutdown status during a certain period of time. Indicates the unit exist It is always powered on. It is always in a shutdown state.

[0096] In one possible implementation, the constraints include unit start-up and shutdown constraints, minimum start-up and shutdown time constraints, active power balance constraints, upper and lower limits of conventional unit output constraints, conventional unit ramping constraints, and power grid flow safety constraints.

[0097] Furthermore, unit start-stop constraints refer to start-stop states. Startup status Shutdown status The three variables satisfy the following equation:

[0098]

[0099] In the formula, and These are Boolean variables, representing the units respectively. In the t Time period and t -1 indicates the start / stop status; a value of 0 indicates the unit is in a stopped state, and a value of 1 indicates the unit is in a started state.

[0100] The minimum start-up and shutdown time constraint of the unit refers to the requirement that the start-up and shutdown state variables of the unit satisfy the following formula within a continuous number of hours after the start-up and shutdown action:

[0101]

[0102]

[0103] In the formula, Indicates the unit In the k Power output during a given time period This represents the number of time periods within a scheduling cycle. Indicates the start-up time of the generator unit. Indicates the duration of the unit shutdown. This indicates iterating through every possible moment when the system might start or stop. Indicates traversal from t From now on, it is necessary to maintain the power on or off for every period of time;

[0104] The expression for the active power balance constraint is as follows:

[0105]

[0106] In the formula, It is a unit index variable. Indicates the number of thermal power units in the system; It is a wind farm index variable. Indicates the number of wind farms in the system. It is a generator set In the Power output during a given time period Indicates wind farm The predicted value, It is a power grid node index variable. Indicates the number of nodes in the system. Represents a node Load forecast;

[0107] The upper and lower limits of output for conventional generating units are expressed as follows:

[0108]

[0109] In the formula, and They are conventional units Maximum and minimum output;

[0110] Conventional unit ramp-up constraints include ramp-up constraints during unit operation and ramp-up constraints during startup and shutdown. Both ramp-up constraints during operation and ramp-up constraints during startup and shutdown are implemented through... Control, the expression is as follows:

[0111]

[0112]

[0113] In the formula, and These represent the upward and downward ramp constraints of the unit during operation, respectively. For the ramp-up constraints during the unit's startup process. This refers to the ramp-up constraints during the unit's shutdown process. It is a generator set In the t Power output during the -1 time period;

[0114] The power flow security constraint is expressed as follows:

[0115]

[0116] In the formula: This represents the upper limit of the power flow of the line. For nodes The load on the line Current transfer matrix For conventional units For the line Current transfer matrix For wind turbines For the line The trend transfer matrix.

[0117] This invention presents a generalized unit combination calculation method that considers low-voltage resources and grid constraints. Addressing the industry pain point that day-ahead unit combination cannot quickly consider the power flow constraints of low-voltage grids after the integration of high-proportion distributed energy (new energy penetration rate ≥ 50%), it proposes a method that completely replaces the traditional quadratic Distflow equation by using "safety distance-MLP (Multi-Layer Sensor) + Big-M (Big M Method) linearization," achieving zero quadratic, zero-iteration, and second-level solution. Furthermore, after field tests on the IEEE 118-node transmission network and 10 IEEE 33-node low-voltage grids, the calculation time was reduced from 417 seconds to 15.7 seconds, a reduction of 96%. The safety boundary is equivalent to the most accurate model, with a generation cost difference of <0.1%, directly meeting the spot market clearing requirement of ≤15 minutes.

[0118] This invention, in its embodiment, targets the power flow security domain of an active distribution network, independently constructing and offline training two fully connected multilayer sensing devices with neural networks. Among them, the voltage security sensing device V... MLP, inputs are active and reactive power injection at nodes, output is the minimum safe distance of each node's voltage amplitude relative to the operating lower limit; Current safety sensor I The MLP (Multi-Level Processing) takes the active and reactive power at the branch head as input and outputs the minimum safe distance of each branch current relative to the thermal stability limit. Simultaneously, a linearization method is proposed, embedding a ReLU neural network into the MILP solver: for V... MLP and I Each ReLU level of the MLP performs Big-M decomposition: introducing binary variables. Indicates the first i The first layer j Whether a neuron is activated and constraints are established. A trained neural network is used to replace the traditional second-order cone constraints, and the replaced constraints are placed into the unit combination with a low-voltage power grid, which has positive significance for accelerating the solution of unit combination calculation problems.

[0119] Please see Figure 2 Another embodiment of the present invention proposes a generalized unit combination calculation system that takes into account low-voltage resources and grid constraints, mainly including:

[0120] The node net load acquisition module 201 is used to acquire the net load of each node in the power grid at each time point.

[0121] The neural network training module 202 is used to take the net load of each node of the power grid at each time as input data, and convert the voltage and current at each time into safe distance as labels to train the pre-established neural network.

[0122] The constraint replacement module 203 is used to construct the objective function and constraints of the unit combination, and replace the constraints with a trained neural network.

[0123] The objective function solving module 204 is used to solve the objective function of the unit combination using a neural network with alternative constraints, and to complete the unit combination calculation.

[0124] Another embodiment of the present invention also proposes an electronic device including a processor and a memory, the processor being used to execute a computer program stored in the memory to implement the generalized unit combination calculation method taking into account low-voltage resources and grid constraints.

[0125] Another embodiment of the present invention also proposes a computer-readable storage medium storing at least one instruction that, when executed by a processor, implements the generalized unit combination calculation method taking into account low-voltage resources and grid constraints.

[0126] Another embodiment of the present invention also proposes a computer program product, the computer program product comprising a computer program, which, when executed by a processor, implements the generalized unit combination calculation method taking into account low-voltage resources and grid constraints.

[0127] 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.

[0128] 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.

[0129] 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.

[0130] 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.

[0131] 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.

[0132] 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 generalized unit combination calculation method considering low-voltage resources and grid constraints, characterized in that, include: Obtain the net load of each node in the power grid at each time point; The net load of each node in the power grid at each time point is used as input data, and the voltage and current at each time point are converted into safe distances as labels to train a pre-established neural network. Construct the objective function and constraints for unit combination, and replace the constraints with a trained neural network; The objective function of unit combination is solved by using a neural network with alternative constraints to complete the unit combination calculation; The step of converting voltage and current at various times into safe distances for use as tags is calculated using the following expression: In the formula, It is a set of nodes. yes t Safe distance of voltage at all times yes t The safe distance for current at all times. It is a collection of routes. This is the permissible voltage fluctuation. It is the first i Each node t Voltage amplitude at time 10:00 yes i - j On the line t Current amplitude at time , yes i - j On the line t The maximum current at time t. It is a collection of moments; When the safe distance meets the following condition, there will be no over-limit situation in the current and voltage of the power grid: 。 2. The generalized unit combination calculation method considering low-voltage resources and grid constraints according to claim 1, characterized in that, The pre-built neural network includes an input layer, K hidden layers, and an output layer, using ReLU as the activation function. In the formula, For the first k The input of a layered neural network, For the first k Layer weights For the first k The output of the layer, For the first k Layer bias; The Big M method is used to transform the maximization operator into a mixed-integer linear form, resulting in the multilayer perceptron (MLP) expression: In the formula, This is the input to the first layer of the neural network. It is the input to the neural network. t For a moment, It is a collection of moments; For the first k The output of the layer, For the first k Layer weights For the first k The input of a layered neural network, For the first k The input to the -1 layer of the neural network, k Indicates a hidden layer. A collection of hidden layers. and As an auxiliary variable, Representing large numbers, This refers to the safe distance for voltage and current.

3. The generalized unit combination calculation method considering low-voltage resources and grid constraints as described in claim 1, characterized in that, In the steps of constructing the objective function and constraints for unit combination, an on-load tap-changing transformer is used to ensure that the voltage of the bus connecting the transmission network and the low-voltage grid remains at 1.0 pu at any given time. Here, the low-voltage grid refers to a grid with AC ≤1kV. The expression is as follows: In the formula, Represents the voltage amplitude at the connection point at different times; Meanwhile, the power demand of low-voltage power grids is consistent with the power supplied to them by the transmission grid, conforming to the following expression: In the formula, It is a collection of routes. i and j Both represent nodes. express t Time of the first i Active load on each node; It refers to the active power supplied by the transmission network to the low-voltage grid at different times. for t Time of the first j Photovoltaic power generation at each node; yes t Time of the first i Reactive load on each node This refers to the reactive power generated by the reactive power compensation device at the busbar connection point. t For a moment, It is a collection of moments.

4. The generalized unit combination calculation method considering low-voltage resources and grid constraints according to claim 1, characterized in that, In the step of constructing the objective function and constraints of the unit combination, the objective function of the unit combination is to minimize the sum of the unit start-up and shutdown costs and fuel costs, as expressed below: In the formula, It is a unit index variable. Indicates the number of thermal power units in the system; It is a time index variable. This represents the number of time periods within a scheduling cycle; This is a Boolean variable representing the unit. In the Start / stop status during a time period This indicates that the unit is in a stopped state. This indicates that the unit is in the powered-on state; This refers to the unit price of coal. It is a generator set In the Fuel consumption during the period and They represent the generating units. Start-up and downtime costs, , , These are the coefficients of the unit cost function. It is a generator set In the Power output during a given time period Indicates the unit In the The square of the output power during the time period, and They represent the generating units. The cost of a single start and stop, This is a Boolean variable representing the unit. In the Startup status of the time period Indicates the unit exist The machine is always in a shutdown state. The device is always powered on. This is a Boolean variable, representing the unit. In the The shutdown status during a certain period of time. Indicates the unit exist It is always powered on. It is always in a shutdown state.

5. The generalized unit combination calculation method considering low-voltage resources and grid constraints according to claim 1, characterized in that, In the steps of constructing the objective function and constraints of the unit combination, the constraints include unit start-up and shutdown constraints, unit minimum start-up and shutdown time constraints, active power balance constraints, upper and lower limits of conventional unit output constraints, conventional unit ramping constraints, and power grid flow security constraints.

6. The generalized unit combination calculation method considering low-voltage resources and grid constraints according to claim 5, characterized in that, The unit start-stop constraints refer to the start-stop status. Startup status Shutdown status The three variables satisfy the following equation: In the formula, and These are Boolean variables, representing the units respectively. In the t Time period and t -1 indicates the start / stop status; a value of 0 indicates the unit is in a stopped state, and a value of 1 indicates the unit is in a started state. The minimum start-up and shutdown time constraint of the unit refers to the start-up and shutdown state variables satisfying the following formula within a continuous number of hours after the unit performs a start-up or shutdown action: In the formula, Indicates the unit In the k Power output during a given time period This represents the number of time periods within a scheduling cycle. Indicates the start-up time of the generator unit. Indicates the duration of the unit shutdown. This indicates iterating through every possible moment when the system might start or stop. Indicates traversal from t From now on, it is necessary to maintain the power on or off for every period of time; The expression for the active power balance constraint is as follows: In the formula, It is a unit index variable. Indicates the number of thermal power units in the system; It is a wind farm index variable. Indicates the number of wind farms in the system. It is a generator set In the Power output during a given time period Indicates wind farm The predicted value, It is a power grid node index variable. Indicates the number of nodes in the system. Represents a node Load forecast; The upper and lower limits of output for conventional generating units are expressed as follows: In the formula, and They are conventional units Maximum and minimum output; Conventional unit ramp-up constraints include ramp-up constraints during unit operation and ramp-up constraints during startup and shutdown, expressed as follows: In the formula, and These represent the upward and downward ramp constraints of the unit during operation, respectively. For the ramp-up constraints during the unit's startup process. This refers to the ramp-up constraints during the unit's shutdown process. It is a generator set In the t Output power during the -1 time period; The power flow security constraint is expressed as follows: In the formula: This is the upper limit of the power flow of the line. For nodes The load on the line Current transfer matrix For conventional units For the line Current transfer matrix For wind turbines For the line The trend transfer matrix.

7. A generalized unit combination calculation system considering low-voltage resources and grid constraints, characterized in that, include: The node net load acquisition module is used to acquire the net load of each node in the power grid at each time point. The neural network training module is used to take the net load of each node of the power grid at each time as input data, and convert the voltage and current at each time into safe distance as labels to train the pre-established neural network. The constraint substitution module is used to construct the objective function and constraints of the unit combination, and to substitute the constraints using a trained neural network. The objective function solving module is used to solve the objective function of the unit combination using a neural network with alternative constraints, and to complete the unit combination calculation. The step of converting voltage and current at various times into safe distances for use as tags is calculated using the following expression: In the formula, It is a set of nodes. yes t Safe distance of voltage at all times yes t The safe distance for current at all times. It is a collection of routes. This is the permissible voltage fluctuation. It is the first i Each node t Voltage amplitude at time 10:00 yes i - j On the line t Current amplitude at time , yes i - j On the line t The maximum current at time t. It is a collection of moments; When the safe distance meets the following condition, there will be no over-limit situation in the current and voltage of the power grid: 。 8. An electronic device, characterized in that, It includes a processor and a memory, the processor being used to execute a computer program stored in the memory to implement the generalized unit combination calculation method that takes into account low-voltage resources and grid constraints as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction, which, when executed by a processor, implements the generalized unit combination calculation method that takes into account low-voltage resources and grid constraints as described in any one of claims 1 to 6.

10. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the generalized unit combination calculation method as described in any one of claims 1 to 6, taking into account low-voltage resources and grid constraints.

Images