Power system optimization method and device and electronic equipment

Abstract

The invention discloses a power system optimization method and device and electronic equipment. The method comprises the following steps: collecting power data of a target power system; constructing a target function of the target power system optimization problem and a constraint condition corresponding to the target function; decomposing the target power system optimization problem into a multi-scene optimization problem by adopting a Lagrangian relaxation algorithm, and respectively solving the solutions of the multi-scene optimization problem; and iteratively optimizing the solution of the multi-scene optimization problem by adopting a progressive hedging algorithm to obtain a target solution of the target function, the target solution being used for representing a power generation strategy of the new energy power generation system and a storage strategy of the battery storage system under the condition of ensuring the power demand. The technical problem that the scheduling decision of the hybrid power system connected to the new energy power generation system is unreliable is solved.

Description

Technical Field

[0001] The present invention relates to a field, and in particular to a method, a device and an electronic device for optimizing a power system. Background Art

[0002] As more and more renewable energy power generation systems are added to the power system, the complexity and uncertainty of the power system increase, resulting in unreliable decisions such as fixed power supply commitments and battery charging and discharging decisions made by the power system during the operation of the power grid, thus affecting the dispatching efficiency and stability of the power grid operation.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present invention provide a power system optimization method, device and electronic equipment to at least solve the technical problem of unreliable scheduling decision of a hybrid power system connected to a new energy power generation system.

[0005] According to one aspect of an embodiment of the present invention, there is provided a method for optimizing a power system, comprising: collecting power data of a target power system, wherein the target power system comprises a new energy power generation system and a battery storage system for storing the power generated by the new energy power generation system; constructing an objective function of the target power system optimization problem and constraints corresponding to the objective function, wherein the target power system optimization problem is to optimize the new energy power generation system and the battery storage system while ensuring power demand; using a Lagrangian relaxation algorithm to decompose the target power system optimization problem into a multi-scenario optimization problem, and respectively solving solutions to the multi-scenario optimization problem; using a progressive hedging algorithm to iteratively optimize the solution of the multi-scenario optimization problem to obtain a target solution of the objective function, wherein the target solution is used to characterize the power generation strategy of the new energy power generation system and the storage strategy of the battery storage system while ensuring power demand.

[0006] Optionally, constructing the objective function of the target power system optimization problem includes: obtaining the unit price of electricity sales revenue at the target time, the power delivered by the target power system at the target time, the depreciation cost of charging the battery storage system, the depreciation cost of discharging the battery storage system, the maximum charging power of the battery storage system, and the maximum discharging power of the battery storage system; based on the unit price of electricity sales revenue, the power delivered by the target power system at the target time, The depreciation cost of charging the battery storage system, the depreciation cost of discharging the battery storage system cost , the maximum charging power of the battery storage system, the maximum discharging power of the battery storage system, and the objective function of the target power system optimization problem are constructed.

[0007] Optionally, the constraint conditions include: a joint opportunity constraint, a constraint that the battery storage system cannot be charged and discharged simultaneously, a charging power constraint, a discharging power constraint, and an energy storage capacity constraint of the battery storage system.

[0008] Optionally, the Lagrangian relaxation algorithm is used to decompose the target power system optimization problem into multi-scenario optimization problems, and solutions to the multi-scenario optimization problems are respectively solved, including: determining the constraints after relaxation using the Lagrangian relaxation algorithm, the activation scenario number constraints, and the Lagrangian multipliers corresponding to the activation scenarios; based on the constraints after relaxation using the Lagrangian relaxation algorithm, the activation scenario number constraints, and the Lagrangian multipliers corresponding to the activation scenarios, the target power system optimization problem is decomposed into multi-scenario optimization problems, and solutions to the multi-scenario optimization problems are respectively solved.

[0009] Optionally, the use of a progressive hedging algorithm to iteratively optimize the solution to the multi-scenario optimization problem to obtain a target solution to the objective function includes: weighted averaging the solution to the multi-scenario optimization problem to obtain a weighted average; and iteratively optimizing the multi-scenario optimization problem based on the weighted average to obtain the target solution to the objective function.

[0010] Optionally, the iterative optimization of the multi-scenario optimization problem based on the weighted average value to obtain the target solution of the objective function includes: in the process of iterative optimization of the multi-scenario optimization problem based on the weighted average value, determining the difference between the solution corresponding to the previous iteration and the solution corresponding to the current iteration; when the difference is less than a predetermined difference threshold, terminating the iteration to obtain the target solution of the objective function.

[0011] According to another aspect of the present invention, there is provided an electric power system optimization device, comprising: a collection module for collecting electric power data of a target electric power system, wherein the target electric power system comprises a new energy power generation system and a battery storage system for storing the power generated by the new energy power generation system; a construction module for constructing an objective function of the target electric power system optimization problem and constraints corresponding to the objective function, wherein the target electric power system optimization problem is to optimize the new energy power generation system and the battery storage system while ensuring electric power demand; a decomposition module for decomposing the target electric power system optimization problem into multi-scenario optimization problems by adopting a Lagrangian relaxation algorithm, and respectively solving solutions to the multi-scenario optimization problems; an iteration module for iteratively optimizing the solutions to the multi-scenario optimization problems by adopting a progressive hedging algorithm to obtain an objective solution of the objective function, wherein the objective solution is used to characterize the power generation strategy of the new energy power generation system and the storage strategy of the battery storage system while ensuring electric power demand.

[0012] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned power system optimization methods.

[0013] According to another aspect of the present invention, there is provided an electronic device, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes any one of the above-mentioned power system optimization methods when running.

[0014] According to yet another aspect of the present invention, there is provided a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the steps of any one of the methods for optimizing a power system are implemented.

[0015] In an embodiment of the present invention, the Lagrangian relaxation algorithm is used to decompose the optimization problem of the power system into optimization sub-problems under multiple scenarios, and the progressive hedging algorithm is used to perform weighted averaging on the solutions of multiple optimization sub-problems, and the solutions of the sub-problems are updated so that all solutions tend to be consistent, thereby achieving the purpose of optimizing the dispatching decision of the power system, thereby achieving the technical effect that the power system can match the power supply and demand through flexible dispatch, and further solving the technical problem of unreliable dispatching decisions of hybrid power systems connected to new energy power generation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0017] Figure 1 is a flow chart of a power system optimization method according to an embodiment of the present invention;

[0018] Figure 2 It is a structural block diagram of a power system optimization device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0020] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0021] First, some nouns or terms that appear in the description of the embodiments of the present application are subject to the following explanations:

[0022] Joint Chance Constraints (JCCs): JCCs are an advanced method for dealing with random variables and uncertainty in optimization models. They allow setting a certain level of probability assurance on the relationship between decision variables and random variables in the face of uncertainty and randomness. Unlike independent chance constraints, which only restrict the probability of a single uncertain event, joint chance constraints constrain the probability of multiple random variables satisfying a series of constraints at the same time.

[0023] Lagrangian Relaxation: It is an effective strategy in optimization theory for solving complex optimization problems, especially those with difficult-to-handle constraints. It is based on the principle of the Lagrange Multiplier Method, but is more flexible and practical, especially when solving large-scale, multi-stage optimization problems. In the Lagrangian relaxation technique, some hard constraints (i.e., constraints that must be strictly satisfied) originally included in the optimization problem are "relaxed" and become part of the objective function. By introducing Lagrange multipliers to multiply these constraint terms, and then adding the product to the objective function, a Lagrangian dual problem is created. This method allows constraints to no longer be strict boundary restrictions, but to become conditions that can be violated by paying a certain "cost" or "penalty".

[0024] Progressive Hedging Algorithm (PHA): is an iterative algorithm for solving stochastic optimization problems, especially optimization problems involving multi-stage decisions and multiple random scenarios. The algorithm was originally proposed by Rockafellar and Wets in 1991 to solve complex optimization problems introduced by random uncertainty. The core idea of ​​PHA is to gradually refine and adjust the solutions under each scenario to eventually reach a globally consistent optimal solution. It is particularly suitable for dealing with decision-making problems with complex coupled constraints, where decisions need to be adjusted at multiple stages according to different random scenarios.

[0025] Big M Method: It is a commonly used technique in mathematical programming, especially linear programming, to deal with constraints containing "if-then" logical relationships, or to convert logical constraints into mathematical constraints so that they can be solved in the optimization model. The core of this method is to introduce a sufficiently large positive number M (called "big M") and use it to penalize or control the behavior of certain variables in the model to ensure that the model complies with specific logical rules. In the Big M method, the value of M usually needs to be large enough to ensure that when a certain condition in the model is not met, it can force another related variable to be in a specific state, and when the condition is met, the role of M can be ignored and does not interfere with the natural solution process of the model.

[0026] Photovoltaic power stations are systems that use solar panels and inverters to convert light energy into electrical energy. Due to the uncertainty and non-all-weather availability of solar energy, this system faces challenges in power supply. In order to improve the economic benefits of the system, photovoltaic power generation systems are usually combined with large-scale energy storage equipment to form a hybrid energy system. This system can release energy during peak demand and is widely used in modern microgrids, especially as independent systems on remote islands. Similarly, other new energy power generation systems can also be combined with battery storage systems to form a hybrid energy system.

[0027] However, the relevant technologies are still insufficient in ensuring the reliability of power supply in hybrid energy systems. Due to the unstable supply of renewable energy, even with the use of new energy power generation systems, other power generation systems and battery storage systems, fixed power supply commitments in the power market are still challenging. In addition, in order to solve the power system optimization problem, an optimization model is established, but the optimization model containing joint probability constraints is computationally complex and belongs to an NP-hard problem, and the feasible region of the optimization model is a non-convex region, which increases the difficulty of solving the problem.

[0028] In the power system, the first stage decisions need to be made without full knowledge, such as commitments in the power day-ahead market. Subsequently, decisions such as battery charging and discharging are made in the second stage based on the actual output of new energy such as solar energy. This two-stage or multi-stage decision-making makes the optimization model more complex, but it is essential to ensure high reliability and adaptability of the power system. Although the performance and economic benefits of the system can be improved to a certain extent through technological progress and optimization model design, how to reduce the consumption of computing resources while ensuring high reliability is still a key issue that needs to be solved in technological development.

[0029] Therefore, in order to improve the reliability and adaptability of the power system, a power system optimization method is provided. By adopting the Lagrangian relaxation algorithm to relax the power system optimization problem, and using the progressive hedging algorithm to iteratively solve the relaxed power system optimization problem, an accurate and effective power system scheduling strategy can be obtained. By controlling the charging and discharging of the new energy power generation system and the charging and discharging of the battery storage system, the reliability of the power system is improved.

[0030] According to an embodiment of the present invention, an embodiment of a power system optimization method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0031] Figure 1 is a flow chart of a power system optimization method according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0032] Step S102, collecting power data of a target power system, wherein the target power system includes a new energy power generation system and a battery storage system for storing power generated by the new energy power generation system.

[0033] As an optional embodiment, the execution subject of the method of this embodiment can be a terminal or server for power system optimization. When applied to a terminal, it can easily realize power system optimization. When applied to a server, it can call on the rich computing resources of the server and optimize the power system more accurately. Among them, the types of the above-mentioned terminals can be various, for example, it can be a mobile terminal with certain computing capabilities, or it can be a fixed computer device with recognition capabilities, etc. The types of the above-mentioned servers can also be various, for example, it can be a local server or a virtual cloud server. According to the computing power, it can be a single computer device or a computer cluster integrated with multiple computer devices.

[0034] As an optional embodiment, when collecting power data of the target power system, the above-mentioned power system includes a new energy power generation system and a battery storage system for storing the power generated by the above-mentioned new energy power generation system. The above-mentioned power data may include various types of data, such as power generation data, load data, price data, network data, energy storage data, weather and environmental data, equipment performance data, user behavior data, etc.

[0035] Among them, power generation data can include the power generation of each energy type (such as solar energy, wind energy, hydropower, nuclear energy, fossil fuels, etc.), power generation cost, power generation efficiency, as well as the operating status and maintenance records of power generation equipment; load data is power demand data, that is, the power consumption of power consumers at different time points, which can include peak load, valley load, daily load curve and annual load curve, etc., which can be used to predict and manage the balance of power supply and demand; price data is the electricity price information in the power market, which can include real-time electricity price, day-ahead market electricity price, futures market electricity price, etc.; network data is the operating data of the power transmission and distribution network, which can include the current, voltage, power flow of the line, The operating status of the substation, etc.; energy storage data can include the charging and discharging status, storage capacity, charging and discharging efficiency and life cycle of energy storage equipment (such as batteries, flywheels, pumped storage, etc.), which can optimize power dispatch and improve system flexibility; weather and environmental data for power systems that rely on renewable energy can include weather data (such as wind speed, light intensity, temperature, etc.) and environmental data (such as humidity, atmospheric pressure, etc.); equipment performance data can be performance parameters of power generation equipment, transmission equipment and distribution equipment, such as efficiency, failure rate and maintenance cycle, etc.; user behavior data can include user power consumption patterns, demand response capabilities, behavioral changes under dynamic electricity prices and other data. By collecting the above power data, we can fully understand the operating status information of the power system, thereby guiding decision-making and optimizing the overall performance of the power system.

[0036] Step S104, constructing the objective function of the above-mentioned target power system optimization problem and the constraints corresponding to the above-mentioned objective function, wherein the above-mentioned target power system optimization problem is to optimize the above-mentioned new energy power generation system and the above-mentioned battery storage system while ensuring the power demand.

[0037] As an optional embodiment, when constructing the objective function of the above-mentioned target power system optimization problem, a variety of methods can be used. For example, cost minimization can be used as the objective function to minimize the operating cost of the power system, or revenue maximization can be used as the objective function to maximize the total revenue of the power system, so that the objective function is the most direct and easy to obtain. For another example, other considerations can be added when determining the objective function, such as reliability optimization, and the power system can be kept reliable under various conditions by adding penalty terms to the objective function. Specifically, the unit price of electricity sales revenue at the target time, the power delivered by the above-mentioned target power system at the above-mentioned target time, the depreciation cost of charging the above-mentioned battery storage system, the depreciation cost of discharging the above-mentioned battery storage system, the maximum charging power of the above-mentioned battery storage system, and the maximum discharge power of the above-mentioned battery storage system can be obtained; based on the above-mentioned acquired parameters, the above-mentioned objective function of the above-mentioned target power system optimization problem is constructed. By comprehensively considering the above parameters to construct the objective function, the decision-making of the power system can be data-driven, the reliability and sustainability of the power system can be improved, the conditions of supply and demand balance, balance of revenue and expenditure can be fully met, and the new energy power generation system and battery storage system can be guided to operate efficiently and economically in a complex and changing environment, while reducing risks and meeting social needs.

[0038] As an optional embodiment, when constructing the constraint conditions corresponding to the above objective function, multiple types of constraints can be used. For example, joint opportunity constraints, the above battery storage system cannot be charged and discharged at the same time, charging power constraints, discharging power constraints, and energy storage capacity constraints of the above battery storage system. Among them, the joint opportunity constraint is a probabilistic constraint that can ensure that the probability of multiple constraints being met at the same time in an uncertain environment reaches at least a predetermined threshold, that is, the dispatch of the power system can ensure that the power supply meets the demand, even in the case of fluctuations in renewable energy production or changes in power demand; the battery storage system cannot be charged and discharged at the same time, and the charging power constraint and the discharging power constraint ensure that the battery can only be in one of the charging or discharging states at any given time point, meeting the physical characteristics and operating mechanism of the battery; the energy storage capacity constraint of the battery storage system defines the upper and lower limits of the battery storage energy, ensuring that the battery is within a safe and effective operating range. By using the above constraints to solve the optimization problem of the power system, the power system can find the optimal dispatching strategy under the conditions of meeting power demand, maintaining system reliability, protecting battery hardware, and being economically feasible.

[0039] Step S106, using the Lagrangian relaxation algorithm, decompose the above target power system optimization problem into multi-scenario optimization problems, and solve the solutions of the above multi-scenario optimization problems respectively.

[0040] As an optional embodiment, when using the Lagrangian relaxation algorithm to decompose the above-mentioned target power system optimization problem into multi-scenario optimization problems and solve the solutions of the above-mentioned multi-scenario optimization problems respectively, the constraints after relaxation using the Lagrangian relaxation algorithm, the constraints on the number of activated scenarios, and the Lagrangian multipliers corresponding to the activated scenarios can be determined; based on the constraints after relaxation using the Lagrangian relaxation algorithm, the constraints on the number of activated scenarios, and the Lagrangian multipliers corresponding to the above-mentioned activated scenarios, the above-mentioned target power system optimization problem is decomposed into multi-scenario optimization problems, and the solutions of the above-mentioned multi-scenario optimization problems are solved respectively.

[0041] Since the above-mentioned target power system optimization problem involves random and uncertain conditions in multiple scenarios, the constraints will be coupled across scenarios, making the optimization problem too large to solve. Therefore, the large M method can be used. By setting the large M constant, the original constraints can be relaxed and ensure that the constraints are met even if the scenario is not activated. Among them, the above-mentioned scenario is a specific state or situation that may occur in the future. It is used to describe a set of specific values ​​that uncertainty parameters (such as solar energy production, wind power generation, power demand, etc.) may take within a given time period. Each scenario represents a possible future environment. The above-mentioned activated scenario is a set of scenarios selected from all possible scenarios for specific calculations and decisions when solving the optimization problem. These scenarios represent different states of different operating conditions, market conditions, weather conditions or other uncertain factors, and play an important role in ensuring the optimal dispatch of the power system in the face of uncertainty.

[0042] By using the Lagrangian relaxation algorithm, the above-mentioned target power system optimization problem is decomposed into multi-scenario optimization problems, and the solutions of the above multi-scenario optimization problems are solved separately, which can simplify complex optimization problems. The scale and complexity of each scenario optimization problem are usually much smaller than the original problem, making the originally difficult global problem manageable, greatly reducing the demand for computing resources and solution time. At the same time, the multi-scenario optimization problem obtained by decomposition can simulate the different states of uncertain variables faced by power system dispatch, so as to evaluate and optimize the system's operating strategy under various possible conditions, which can improve the robustness and adaptability of the power system, so that power dispatch can better cope with the fluctuations and changes that may be encountered in future actual operations.

[0043] Step S108, using a progressive hedging algorithm to iteratively optimize the solution of the above-mentioned multi-scenario optimization problem to obtain a target solution of the above-mentioned objective function, wherein the above-mentioned target solution is used to characterize the power generation strategy of the above-mentioned new energy power generation system and the storage strategy of the above-mentioned battery storage system while ensuring the power demand.

[0044] As an optional embodiment, when the solution of the multi-scenario optimization problem is iteratively optimized by the progressive hedging algorithm to obtain the target solution of the objective function, the solution of the multi-scenario optimization problem can be weighted averaged to obtain a weighted average value; based on the weighted average value, the multi-scenario optimization problem is iteratively optimized to obtain the target solution of the objective function. A penalty term can also be introduced in the process of iteratively optimizing the solution of the multi-scenario optimization problem by the progressive hedging algorithm. The penalty term is related to the difference of the scenario solution and is used to ensure that the solutions between the scenarios gradually converge, which can accelerate the convergence speed and improve the quality of the solution. By combining the progressive hedging algorithm with the Lagrangian relaxation algorithm, first relaxing the coupling constraints, and then iteratively optimizing the solution of each scenario with the progressive hedging algorithm, large-scale random optimization problems can be effectively handled. Specifically, by calculating the weighted average of the solutions under all scenarios, the differences between the solutions of different scenarios can be gradually reduced, and their convergence to a common and robust solution can be promoted, ensuring that the power system can adopt a strategy that performs well in all expected scenarios when facing uncertainty. Through iterative optimization, the algorithm can gradually approach the global optimal solution instead of being limited to the local optimal solution, and at the same time improve the stability of the solution.

[0045] As an optional embodiment, when the multi-scenario optimization problem is iteratively optimized based on the weighted average value to obtain the target solution of the objective function, multiple methods can be used. For example, the difference between the solution corresponding to the previous iteration and the solution corresponding to the current iteration can be determined in the process of iterative optimization of the multi-scenario optimization problem based on the weighted average value; when the difference is less than the predetermined difference threshold, the iteration is terminated to obtain the target solution of the objective function. For another example, when the number of iterative optimizations executed reaches the predetermined maximum number of iterations, the iteration is terminated to obtain the target solution of the objective function. The method of using the predetermined maximum number of iterations as the termination condition is simpler and easier to implement, and can avoid wireless loops, but may miss higher quality solutions, that is, the target solution obtained is not the optimal solution to the optimization problem, and the solution obtained in this way cannot guarantee its robustness. The method of using the predetermined difference threshold as the termination condition can ensure that the solution found is a solution that performs similarly in all scenarios, representing a robust decision-making strategy that can better adapt to the uncertainty distribution of power system optimization problems and the characteristics of the objective function.

[0046] Among them, the above-mentioned predetermined difference threshold can be obtained in a variety of ways. For example, the threshold can be set based on the experience of similar problems in the past; the difference threshold can be set according to the degree of acceptance of the power system for risk. The more critical the dispatch decision made by the power system is, the smaller the difference threshold can be set; a reasonable difference threshold can also be determined by analyzing the convergence behavior and speed of the algorithm. For example, an initial threshold can be set, and then the threshold can be dynamically adjusted according to the downward trend of the difference in the scenario solution during the algorithm iteration process, so that each iteration optimization is an effective iteration, and the speed at which the algorithm obtains the target solution is greatly increased. In addition, when obtaining the above-mentioned predetermined difference threshold, the inherent characteristics of the problem and the nature of the data distribution can also be considered. For example, for scenario data with high uncertainty in distribution, a higher threshold can be set to ensure the convergence of the algorithm; conversely, for data with a more concentrated distribution, a lower threshold can be used. By comprehensively considering the characteristics of the problem, computational efficiency and decision preferences, the difference threshold can be reasonably set to ensure that the algorithm reaches the optimal solution or an acceptable solution within the limited resources.

[0047] Through the above steps, the Lagrangian relaxation algorithm can decompose the original power system optimization problem into a multi-scenario optimization problem, simplify the calculation, and enhance the adaptability of the power system for each scenario. Then, the progressive hedging algorithm iteratively optimizes multiple solutions to the multi-scenario optimization problem, coordinates multiple independent scenarios, eliminates the differences between multiple solutions, and finally provides a coordinated solution that takes all uncertain scenarios into account, ensuring the reliability and adaptability of power system scheduling in the face of uncertain scenarios. By combining the Lagrangian relaxation algorithm and the progressive hedging algorithm, the challenges brought by large-scale and uncertainty in power system scheduling can be effectively solved, the efficiency and accuracy of the solution can be improved, high-dimensional random scenarios can be processed, and the optimal scheduling strategy can be implemented in the power system under complex conditions such as fluctuations in renewable energy and changes in power demand.

[0048] In combination with the above embodiment and the optional embodiment, an optional implementation is provided. In this optional implementation, a stochastic optimization model for scheduling a hybrid solar cell energy storage system is proposed, and the method includes the following processing.

[0049] S1, data acquisition and processing.

[0050] Collect all necessary data to be used, including historical weather data, solar power generation data records, power consumption patterns, battery performance parameters, etc. These data provide the necessary input and verification basis for the model. Then clean and preprocess the collected data, including processing missing values, outliers and data formatting, to ensure data quality so that it can be accurately interpreted and applied by the model.

[0051] S2, establish a mathematical model.

[0052] A mathematical model is established to optimize the management of battery storage and solar power generation in the power system. The model aims to maximize the economic benefits of the power grid while ensuring the reliability of meeting power demand. The model achieves this goal by integrating the interaction between solar power generation, battery storage and demand. The mathematical expression of the model objective function is as follows:

[0053]

[0054] In the formula, R t is the unit price of electricity sales revenue at time t; y t is the power committed to be delivered to the grid at time t; C c , C d denote the depreciation cost of battery charging and discharging respectively; are the charging and discharging powers in scenario w and time t respectively; Δ represents the scheduling step; E is the calculation expectation.

[0055] The constraints include a joint opportunity constraint, which requires that the joint probability of meeting the promised power through the available solar energy and battery discharge is above a preset threshold ε over the entire time range; if the solar energy output exceeds the requirement, solar power may be abandoned, and this optional implementation does not penalize solar power abandonment. The mathematical expression of the joint opportunity constraint is as follows:

[0056]

[0057] In the formula, is the available solar power at time t in scenario ω; ε represents the tolerable failure probability, which is used to ensure the reliability of the system.

[0058] In order to express the energy conversion during the charging or discharging process of the energy storage battery, the mathematical expression of the energy change in the energy storage battery involving efficiency is as follows:

[0059]

[0060] In the formula, is the battery storage energy at time t in scenario ω; η is the battery charge and discharge efficiency.

[0061] The energy storage charging and discharging power limit should ensure that the charging power and discharging power are not positive at the same time, that is, ensure that the storage cannot be charged and discharged at the same time. The mathematical expression of the charging power constraint is as follows:

[0062]

[0063] Where P is the maximum charging power of the battery, It is a binary state variable. The variable is 0, which means the energy storage battery is charged, and 1, which means the energy storage battery is discharged.

[0064] The mathematical expression of energy storage discharge power constraint is as follows:

[0065]

[0066] Where Q is the maximum discharge power of the battery.

[0067] Energy storage capacity constraints ensure that the energy stored in the energy storage battery is between the minimum and maximum constraint limits to prevent the energy storage battery from being overcharged or over-discharged.

[0068]

[0069] In the formula, X, are the minimum and maximum limits of battery energy respectively.

[0070] S3, Lagrangian dual process.

[0071] Due to the random and uncertain conditions involving multiple scenarios, constraints will be coupled across scenarios, which increases the scale of the problem and complicates the solution. The Lagrangian dual method relaxes these cross-scenario coupling constraints by introducing Lagrangian multipliers. In order to provide a choice of activation scenarios for the Lagrangian dual process, the big M method is first used to relax the original constraints. The mathematical expression of the relaxed constraint conditions is:

[0072]

[0073] In the formula, is a 0-1 state variable, indicating the activation state of the scene ω, and M is a large constant, which ensures that the constraint condition must be established when the scene is not activated, otherwise the constraint condition will fail.

[0074] Activate the scene quantity constraint, its mathematical expression is as follows:

[0075]

[0076] Where N e For all scenes that can be activated The number of scenarios is Ω, which is a strategy for dealing with the computational complexity that may be brought by different scenarios.

[0077] Among them, the activation scenario is a set of scenarios selected from all possible scenarios for specific calculations and decision-making when solving optimization problems. These scenarios represent different operating conditions, market conditions, weather conditions or other uncertain factors, which are crucial for evaluating the performance of the system under various conditions and formulating corresponding strategies.

[0078] In Lagrangian dual theory, the constraints of the original problem are transformed into part of the objective function by introducing one or more Lagrangian multipliers. These multipliers correspond to the constraints in the original problem, and the purpose is to find a dual solution that provides a lower or upper bound on the feasible solution of the original problem. For the power system optimization model, the mathematical expression of the Lagrangian dual function L(λ) is as follows:

[0079]

[0080] In the formula, The term represents the expected value of the total expected revenue of the power system minus the charging and discharging costs within time T; Where λ is the Lagrange multiplier, corresponding to the scenario coupling constraint, which aims to penalize scenarios that exceed the allowed number of activations to ensure that the number of activations of scenarios in the model does not exceed the set threshold.

[0081] When solving, first select an appropriate λ as the initial value; then solve the above maximization problem for a given λ to determine the possible maximum value of the objective function; then update λ based on the solution to the dual problem to better approximate the optimal solution to the original problem; repeat solving the dual problem and updating λ until the convergence condition is met or the maximum number of iterations is reached. Table 1 is a flow chart of the Lagrangian dual process, as shown in Table 1.

[0082] Table 1 Lagrangian dual process flow chart

[0083]

[0084] This method can accelerate the solution process of the power system optimization problem constructed in the present invention by decomposing and simplifying the problem, especially when it can be decomposed into multiple relatively independent sub-problems. When used, it is necessary to design an iterative process and adjust appropriate parameters to achieve the best effect.

[0085] S4, establish a progressive hedging algorithm.

[0086] The progressive hedging algorithm is an algorithm specifically used to handle random optimization, and is suitable for situations where the scenario parameters involved in this optional implementation are multiple and interrelated. By decomposing the large problem into several small problems, each corresponding to a different random scenario, the optimization problem is solved independently in each scenario. In an iterative manner, the progressive hedging algorithm gradually adjusts the solution under each scenario so that the solution of each scenario converges in a globally consistent direction. In addition, the scenario coupling problem that may not be completely solved in the above-mentioned Lagrangian dual process can be effectively managed through the iterative update mechanism of the progressive hedging algorithm. Therefore, the progressive hedging algorithm can be used to solve the problem of the model proposed in this optional implementation, and its mathematical expression is as follows:

[0087]

[0088] In the formula, is the average value of power commitment in all scenarios; ρ t is the penalty factor related to the power commitment deviation.

[0089] The progressive hedging algorithm provides an effective solution strategy when dealing with power system optimization problems, especially when multiple random scenarios are obtained through the Lagrangian dual process. First, the algorithm requires an initial solution for each scenario, which can be based on past data, simplified heuristic methods, or previous analytical solutions. This step is crucial because it provides a solid starting point for the algorithm and ensures that each scenario has a practical solution to start the iterative process.

[0090] Each scenario is then optimized independently. This step uses appropriate mathematical programming techniques to solve the optimization problem for each scenario independently without considering the solutions for other scenarios. The advantage of this approach is that it reduces the overall computational complexity of the problem, and the problems for each scenario can be processed in parallel, greatly speeding up the overall computation.

[0091] Next, based on the independent solutions of each scenario, the progressive hedging algorithm updates the current solution of each scenario by calculating the weighted average of all scenario solutions. This step is the core of the algorithm, which continuously adjusts the solution to reduce the differences between different scenarios, thereby driving the overall solution to converge to a global consistent solution. The above method can be applied to power system optimization problems because it can ensure the consistency of solutions under different weather conditions or different market reactions, thereby improving the stability and reliability of the entire system.

[0092] In addition, the appropriate adjustment of penalty parameters plays a key role in the algorithm process. Adjusting the penalty parameters according to the deviation size of the scenario solution can effectively balance the adjustment speed and accuracy of the solution. This adjustment ensures the flexibility and efficiency of the algorithm in practical applications, especially in the power market environment with large amounts of data and changing conditions.

[0093] Finally, by continuously iterating until the change of the solution is lower than the predetermined threshold, the algorithm is ensured to converge to a stable optimal solution. This not only ensures the optimal performance of the power system under various operating conditions, but also ensures the improvement of dispatch efficiency and dispatch quality in actual operation. Table 2 is the flow chart of the progressive hedging algorithm, as shown in Table 2.

[0094] Table 2 Progressive hedging algorithm flow chart

[0095]

[0096] S5, model solution.

[0097] First, appropriate historical data, including electricity demand, weather conditions, and market prices, need to be collected to build a variety of test scenarios. Then, the above algorithm is run under these scenarios to monitor its computational efficiency, convergence speed, and the accuracy of the optimization results. The performance of the algorithm is optimized by adjusting algorithm parameters such as penalty factors and number of iterations.

[0098] In order to improve the reliability of power system dispatch, this optional implementation adopts an optimization model based on joint opportunity constraints. The model sets a predetermined risk tolerance threshold (ε) to ensure that the system meets the power supply commitment with a high probability under uncertain solar energy output and power demand fluctuations. Specifically, this model applies joint opportunity constraints to evaluate and manage uncertainties from renewable energy in the day-ahead market dispatch of the power system, and optimizes these constraints to ensure the continuity and reliability of power supply. This is particularly important for modern power systems that rely on renewable energy, especially highly volatile energy such as solar and wind energy. By reasonably setting the risk tolerance threshold, the power system can operate within an acceptable risk range, thereby ensuring that the power supply commitment can be met even under extreme conditions.

[0099] At the same time, this optional implementation also adopts an optimization framework that combines Lagrangian relaxation technology and progressive hedging algorithm to solve complex power system scheduling problems with joint opportunity constraints. First, the overall optimization problem is decomposed into multiple independent sub-problems through Lagrangian relaxation technology, and each sub-problem corresponds to a specific operating scenario in the power system. Then, the progressive hedging algorithm is used to perform systematic iterative optimization on these sub-problems, and the solutions of each sub-problem are adjusted through algorithm iteration, gradually guiding the solutions of these sub-problems to approach the global optimal solution. By effectively coordinating the solutions of each sub-problem, the efficiency and accuracy of the solution can be improved, which is suitable for dealing with large-scale uncertain power system optimization problems. By decomposing and optimizing the problem processing, large-scale and high-dimensional optimization problems become easy to manage and solve, while maintaining the quality of the solution, allowing the power system to make more accurate decisions in a short time, optimize resource allocation, and reduce operating costs.

[0100] Furthermore, the model proposed in this optional implementation is a general optimization model that is only applicable to a single type of power storage system, and can also adapt to complex microgrid systems containing multiple batteries and multiple photovoltaic units. The model is able to integrate and optimize the operating strategies of multiple storage units and power generation units, and optimize the distribution and utilization of power in the system by accurately calculating the power input, output and storage status of each unit. The model includes a highly flexible scheduling algorithm that can dynamically adjust the charging and discharging plan and the power generation plan based on real-time data, customize the optimization strategy according to specific needs and existing resources, ensure the optimal match between power supply and demand, and maximize economic benefits and energy efficiency.

[0101] Therefore, this optional implementation provides a practical and efficient solution to optimize the operation of the power system through advanced mathematical modeling and advanced algorithm design, especially in the face of the increasing use of renewable energy and the changing power market. The application of these technologies is expected to help the power industry better cope with future challenges and enhance the sustainable development capabilities of the entire industry.

[0102] According to an embodiment of the present invention, a power system optimization device is provided. Figure 2 is a structural block diagram of a power system optimization device according to an embodiment of the present invention. Figure 2 As shown, the device includes: a collection module 202, a construction module 204, a decomposition module 206, and an iteration module 208. The device is described below.

[0103] The acquisition module 202 is used to collect power data of the target power system, wherein the target power system includes a new energy power generation system and a battery storage system for storing the power generated by the new energy power generation system; the construction module 204 is used to construct the objective function of the optimization problem of the target power system and the constraints corresponding to the objective function, wherein the optimization problem of the target power system is to optimize the new energy power generation system and the battery storage system while ensuring the power demand; the decomposition module 206 is used to use the Lagrangian relaxation algorithm to decompose the optimization problem of the target power system into multi-scenario optimization problems, and solve the solutions of the multi-scenario optimization problems respectively; the iteration module 208 is used to iteratively optimize the solutions of the multi-scenario optimization problems by using the progressive hedging algorithm to obtain the target solution of the objective function, wherein the target solution is used to characterize the power generation strategy of the new energy power generation system and the storage strategy of the battery storage system while ensuring the power demand.

[0104] It should be noted here that the above-mentioned acquisition module 202, construction module 204, decomposition module 206, and iteration module 208 correspond to steps S102 to S108 in the embodiment, and the instances and application scenarios implemented by the multiple modules are the same as the corresponding steps, but are not limited to the contents disclosed in the above-mentioned embodiment.

[0105] As an optional embodiment, the construction module 204 includes: an acquisition unit and a construction unit. The acquisition unit is used to acquire the unit price of electricity sales revenue at the target time, the power delivered by the target power system at the target time, the depreciation cost of charging the battery storage system, the depreciation cost of discharging the battery storage system, the maximum charging power of the battery storage system, and the maximum discharging power of the battery storage system; the construction unit is connected to the acquisition unit and is used to construct the objective function of the target power system optimization problem based on the unit price of electricity sales revenue, the power delivered by the target power system at the target time, the depreciation cost of charging the battery storage system, the depreciation cost of discharging the battery storage system, the maximum charging power of the battery storage system, and the maximum discharging power of the battery storage system.

[0106] As an optional embodiment, the above-mentioned constraints include: joint opportunity constraints, constraints that the above-mentioned battery storage system cannot be charged and discharged at the same time, charging power constraints, discharging power constraints, and energy storage capacity constraints of the above-mentioned battery storage system.

[0107] As an optional embodiment, the decomposition module 206 includes: a first determination unit and a decomposition unit. The first determination unit is used to determine the constraints after relaxation using the Lagrangian relaxation algorithm, the activation scenario quantity constraints, and the Lagrangian multipliers corresponding to the activation scenarios; the decomposition unit is connected to the first determination unit and is used to decompose the target power system optimization problem into multi-scenario optimization problems based on the constraints after relaxation using the Lagrangian relaxation algorithm, the activation scenario quantity constraints, and the Lagrangian multipliers corresponding to the activation scenarios, and respectively solve the multi-scenario optimization problems.

[0108] As an optional embodiment, the iteration module 208 includes: an averaging unit and an iteration unit. The averaging unit is used to perform weighted averaging on the solutions of the multi-scenario optimization problem to obtain a weighted average value; the iteration unit is connected to the averaging unit and is used to perform iterative optimization on the multi-scenario optimization problem based on the weighted average value to obtain the target solution of the target function.

[0109] As an optional embodiment, the iteration module 208 further includes: a second determination unit and a termination unit. The second determination unit is used to determine the difference between the solution corresponding to the previous iteration and the solution corresponding to the current iteration during the iterative optimization of the multi-scenario optimization problem based on the weighted average value; the termination unit is connected to the second determination unit and is used to terminate the iteration when the difference is less than a predetermined difference threshold, so as to obtain the target solution of the target function.

[0110] According to an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium includes a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned power system optimization methods.

[0111] According to an embodiment of the present invention, an electronic device is provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes any one of the above-mentioned power system optimization methods when running.

[0112] According to an embodiment of the present invention, a computer program product is provided, including a computer program, wherein when the computer program is executed by a processor, the steps of any one of the above methods are implemented.

[0113] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0114] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0115] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0116] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0117] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0118] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0119] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for optimizing a power system, characterized in that: include: Collecting power data of a target power system, wherein the target power system includes a new energy power generation system and a battery storage system for storing power generated by the new energy power generation system; Constructing an objective function of the target power system optimization problem and constraints corresponding to the objective function, wherein the target power system optimization problem is to optimize the new energy power generation system and the battery storage system while ensuring power demand; Using a Lagrangian relaxation algorithm, the target power system optimization problem is decomposed into multi-scenario optimization problems, and solutions to the multi-scenario optimization problems are obtained respectively; The solution of the multi-scenario optimization problem is iteratively optimized using a progressive hedging algorithm to obtain a target solution of the objective function, wherein the target solution is used to characterize the power generation strategy of the new energy power generation system and the storage strategy of the battery storage system while ensuring power demand.

2. The method according to claim 1, characterized in that The objective function of constructing the target power system optimization problem includes: Obtaining the unit price of electricity sales revenue at the target time, the power delivered by the target power system at the target time, the depreciation cost of charging the battery storage system, the depreciation cost of discharging the battery storage system, the maximum charging power of the battery storage system, and the maximum discharging power of the battery storage system; Based on the unit price of electricity sales revenue, the power delivered by the target power system at the target time, the depreciation cost of charging the battery storage system, the depreciation cost of discharging the battery storage system, the maximum charging power of the battery storage system, and the maximum discharging power of the battery storage system, the objective function of the target power system optimization problem is constructed.

3. The method according to claim 1, characterized in that The constraints include: a joint opportunity constraint, a constraint that the battery storage system cannot be charged and discharged simultaneously, a charging power constraint, a discharging power constraint, and an energy storage capacity constraint of the battery storage system.

4. The method according to claim 1, characterized in that: The Lagrangian relaxation algorithm is used to decompose the target power system optimization problem into multi-scenario optimization problems, and solutions to the multi-scenario optimization problems are obtained respectively, including: Determine the relaxed constraint conditions using the Lagrangian relaxation algorithm, the number of activated scenarios, and the Lagrangian multipliers corresponding to the activated scenarios; Based on the relaxed constraints using the Lagrangian relaxation algorithm, the activation scenario quantity constraint, and the Lagrangian multipliers corresponding to the activation scenarios, the target power system optimization problem is decomposed into multi-scenario optimization problems, and solutions to the multi-scenario optimization problems are respectively solved.

5. The method according to claim 1, characterized in that The step of iteratively optimizing the solution of the multi-scenario optimization problem using the progressive hedging algorithm to obtain the target solution of the objective function includes: Performing weighted averaging on the solutions to the multi-scenario optimization problem to obtain a weighted average value; Based on the weighted average, the multi-scenario optimization problem is iteratively optimized to obtain the target solution of the objective function.

6. The method according to any one of claims 1 to 5, characterized in that The iterative optimization of the multi-scenario optimization problem based on the weighted average value to obtain the target solution of the objective function includes: Based on the weighted average, during the iterative optimization of the multi-scenario optimization problem, determining a difference between a solution corresponding to a previous iteration and a solution corresponding to the current iteration; When the difference is less than a predetermined difference threshold, the iteration is terminated to obtain the target solution of the target function.

7. A power system optimization device, characterized in that: include: A collection module, used for collecting power data of a target power system, wherein the target power system includes a new energy power generation system and a battery storage system for storing power generated by the new energy power generation system; A construction module, used to construct an objective function of the target power system optimization problem and constraints corresponding to the objective function, wherein the target power system optimization problem is to optimize the new energy power generation system and the battery storage system while ensuring power demand; A decomposition module, used to decompose the target power system optimization problem into multi-scenario optimization problems by using a Lagrangian relaxation algorithm, and respectively solve the multi-scenario optimization problems; An iteration module is used to iteratively optimize the solution of the multi-scenario optimization problem using a progressive hedging algorithm to obtain a target solution of the objective function, wherein the target solution is used to characterize the power generation strategy of the new energy power generation system and the storage strategy of the battery storage system while ensuring power demand.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is executed, the device where the computer-readable storage medium is located is controlled to execute the power system optimization method according to any one of claims 1 to 6.

9. An electronic device, characterized in that: include: A memory storing an executable program; A processor is used to run the program, wherein the program, when running, executes the power system optimization method described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images