Large power grid reactive power optimization method and device, terminal and medium

Abstract

The invention discloses a large power grid reactive power optimization method and device, a terminal and a medium, and relates to the technical field of large power grid reactive power optimization. According to the scheme provided by the invention, through improving a water circulation optimization algorithm, the inertia weight is linearly decreased along with iteration so as to switch exploration and development strategies, and the dynamic balance global and local search of distance factors is combined so as to improve the reactive power optimization efficiency. And a directional updating mechanism is introduced, so that the convergence performance of the algorithm is effectively improved, the problem of slow later convergence of a classical water circulation optimization algorithm is solved, the convergence speed and the global search capability of the algorithm are improved, the engineering practicability of the algorithm in reactive power optimization of the power grid is improved, and the purpose of improving the reactive power optimization effect of the large power grid is achieved.

Description

Technical Field

[0001] This application relates to the field of reactive power optimization technology for large power grids, and in particular to a method, device, terminal and medium for reactive power optimization of large power grids. Background Technology

[0002] Reactive power optimization is an important means to ensure the safe, stable, and economical operation of power systems. Its core is to minimize network losses or active power losses by adjusting control variables such as generator reactive power output, reactive power compensation device switching, and transformer tap positions, while satisfying various system operating constraints. This is a complex nonlinear, multi-constraint, mixed-integer programming problem.

[0003] In recent years, swarm intelligence algorithms have been gradually introduced into the field of reactive power optimization in large power grids to overcome the limitations of traditional mathematical optimization methods in handling high-dimensional and complex problems. Among them, the water cycle optimization algorithm, as a metaheuristic algorithm that simulates the natural water cycle process, has strong global search capabilities. This algorithm maps the solution space of the optimization problem to the Earth's surface, with "ocean" representing the current global optimum and "river" representing a locally better solution. By simulating the dynamic process of water flowing from river to ocean, it gradually guides the search towards a more optimal region.

[0004] However, in practical applications, general water cycle optimization algorithms still suffer from insufficient convergence accuracy and constraint handling capabilities when dealing with high-dimensional and complex constraint problems such as reactive power optimization of large power grids, which limits their further improvement in optimization performance in engineering applications. Summary of the Invention

[0005] This application provides a method, device, terminal, and medium for optimizing reactive power in large power grids, which addresses the technical problem of insufficient optimization performance of existing water cycle optimization algorithms when facing reactive power optimization projects in large power grids.

[0006] To address the aforementioned technical problems, the first aspect of this application provides a method for reactive power optimization in large power grids, comprising:

[0007] Obtain the power grid parameters and water cycle optimization algorithm parameters of the large power grid;

[0008] Based on the power grid parameters and the water cycle optimization algorithm parameters, several water flows are initialized, where each water flow corresponds to a set of decision variables;

[0009] The penalty value for each water flow is calculated using the large power grid optimization penalty function, and the current optimal solution is initialized based on the penalty value. The large power grid optimization penalty function is constructed based on the large power grid optimization objective function combined with the constraint penalty term.

[0010] The position information of the water flow is updated according to the preset improved flow direction update formula, wherein the improved flow direction update formula is used to converge the position of the water flow to a direction closer to the current optimal solution based on the dynamic inertia weight parameter and the update coefficient.

[0011] Based on the updated water flow and the newly generated water flow through the evaporation-rainfall treatment mechanism, the penalty value for each water flow is calculated using the large power grid optimization penalty function, and the current optimal solution is updated based on the penalty value.

[0012] If the current number of iterations has not reached the maximum number of iterations, the position information of the water flow is updated according to the preset improved flow direction update formula until the number of iterations reaches the maximum number of iterations. Then, the current optimal solution is output so as to determine the reactive power optimization scheme of the power grid based on the current optimal solution.

[0013] Preferably, the improved flow direction update formula is:

[0014]

[0015]

[0016] In the formula, Let m be the value of the water flow in the t-th iteration; Let be the optimal water flow value in the t-th iteration. Let be the dynamic inertia weight parameter for the t-th iteration. Let m be the distance factor of the water flow m in the t-th iteration. Let be the update coefficient of water flow m in the t-th iteration.

[0017] Preferably, the formula for calculating the update coefficient is:

[0018]

[0019] In the formula, express and Covariance between and They represent variance and The variance.

[0020] Preferably, the formula for calculating the dynamic inertia weight parameter is:

[0021]

[0022] In the formula, T is the maximum number of iterations set. and These are the upper and lower limits of the dynamic inertia weight parameter, respectively.

[0023] Preferably, the evaporation-rainfall treatment mechanism includes:

[0024] The optimal and high-quality water flows in each iteration are stored in a preset cache space. Based on the water flows in the cache space, a new water flow is generated using a preset evaporation-rainfall processing formula. The high-quality water flow is any water flow whose distance factor from the optimal water flow is less than a preset distance threshold.

[0025] Preferably, the evaporation-rainfall treatment formula is:

[0026]

[0027] In the formula, This is a new water flow formed through an evaporation-rainfall treatment mechanism; The water flow stored in the buffer space; It follows a standard normal distribution; and These represent vectors consisting of the maximum and minimum values ​​of all decision variables, respectively. This represents the disturbance coefficient.

[0028] Preferably, initializing several water flows based on the power grid parameters and the water cycle optimization algorithm parameters includes:

[0029] Based on the power grid parameters and the water cycle optimization algorithm parameters, several water flows are initialized using a Logistic chaotic mapping method.

[0030] The second aspect of this application provides a reactive power optimization device for a large power grid, comprising:

[0031] The parameter acquisition unit is used to acquire the power grid parameters and water cycle optimization algorithm parameters of the large power grid.

[0032] The water flow initialization unit is used to initialize several water flows according to the power grid parameters and the water circulation optimization algorithm parameters, wherein each water flow corresponds to a set of decision variables;

[0033] The optimal solution initialization unit is used to calculate the penalty value of each water flow through the large power grid optimization penalty function, so as to initialize the current optimal solution according to the penalty value. The large power grid optimization penalty function is constructed based on the large power grid optimization objective function combined with the constraint penalty term.

[0034] The water flow position update unit is used to update the position information of the water flow according to a preset improved flow direction update formula, wherein the improved flow direction update formula is used to converge the water flow position to a direction closer to the current optimal solution according to the dynamic inertia weight parameter and the update coefficient.

[0035] The optimal solution update unit is used to calculate the penalty value of each water flow based on the water flow after the update position and the newly generated water flow through the evaporation-rainfall treatment mechanism, through the large power grid optimization penalty function, so as to update the current optimal solution based on the penalty value;

[0036] The cyclic control unit is used to update the position information of the water flow according to the preset improved flow direction update formula if the current number of cyclic iterations has not reached the maximum number of iterations, until the number of cyclic iterations reaches the maximum number of iterations, and then output the current optimal solution to determine the reactive power optimization scheme of the power grid based on the current optimal solution.

[0037] A third aspect of this application provides a large power grid reactive power optimization terminal, comprising: a memory and a processor;

[0038] The memory is used to store program code, which corresponds to the large power grid reactive power optimization method provided in the first aspect of this application.

[0039] The processor is used to read and execute the program code to implement the large power grid reactive power optimization method.

[0040] The fourth aspect of this application provides a computer-readable storage medium storing program code, which is read and executed by a processor to implement the large power grid reactive power optimization method provided in the first aspect of this application.

[0041] As can be seen from the above technical solutions, this application has the following advantages:

[0042] The solution provided in this application improves the water cycle optimization algorithm by linearly decreasing the inertia weight with iteration to switch between exploration and development strategies. It combines the distance factor to dynamically balance global and local searches and introduces a directional update mechanism, which effectively improves the convergence performance of the algorithm. This solves the problem of slow convergence in the later stages of the classic water cycle optimization algorithm, improves the convergence speed and global search capability of the algorithm, and enhances the engineering practicality of the algorithm in reactive power optimization of power grids. In this way, it achieves the goal of improving the reactive power optimization effect of large power grids. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, 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.

[0044] Figure 1 This is a flowchart illustrating an embodiment of a reactive power optimization method for a large power grid provided in this application.

[0045] Figure 2 This is a schematic diagram of the architecture of an embodiment of a large power grid reactive power optimization device provided in this application.

[0046] Figure 3 This is a schematic diagram of the architecture of a large power grid reactive power optimization terminal embodiment provided in this application. Detailed Implementation

[0047] While existing water cycle optimization algorithms possess good global search potential, practical engineering applications have revealed significant limitations when dealing with large-scale power grid reactive power optimization problems. These limitations include a sharp decline in convergence speed in the later stages of iteration, a tendency to get trapped in local optima leading to insufficient convergence accuracy, and weak ability to handle complex power grid operational constraints. Furthermore, these algorithms struggle to effectively balance the dynamic demands of global exploration and local development, resulting in optimization results that fail to fully meet the high-precision requirements of real-time power grid operation. These shortcomings severely restrict the algorithm's engineering applicability in power grid reactive power optimization, especially in high-dimensional, heavily constrained environments where slow convergence in the later stages becomes a key bottleneck hindering performance improvement.

[0048] This application provides a method, device, terminal, and medium for optimizing reactive power in large power grids, which addresses the technical problem of insufficient optimization performance of existing water cycle optimization algorithms when facing reactive power optimization projects in large power grids.

[0049] To make the inventive objectives, features, and advantages of this application more apparent and understandable, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of this application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0050] First, a detailed description of an embodiment of a reactive power optimization method for a large power grid provided in this application is as follows:

[0051] Please see Figure 1The main steps of the reactive power optimization method for a large power grid proposed in this application include:

[0052] Step 101: Obtain the power grid parameters and water cycle optimization algorithm parameters of the large power grid.

[0053] Step 102: Initialize several water flows based on the power grid parameters and the water circulation optimization algorithm parameters.

[0054] Step 103: Optimize the penalty function using the large power grid, calculate the penalty value for each water flow, and initialize the current optimal solution based on the penalty value;

[0055] Among them, the large power grid optimization penalty function is constructed based on the large power grid optimization objective function combined with the constraint penalty term.

[0056] Step 104: Update the water flow position information according to the preset improved flow direction update formula. The improved flow direction update formula is used to converge the water flow position to a direction closer to the current optimal solution based on the dynamic inertia weight parameter and the update coefficient.

[0057] Step 105: Based on the updated water flow and the newly generated water flow through the evaporation-rainfall processing mechanism, calculate the penalty value for each water flow using the large power grid optimization penalty function, and update the current optimal solution based on the penalty value.

[0058] Step 106: Determine if the number of iterations has reached the maximum number of iterations. If not, return to step 104. If yes, output the current optimal solution to determine the reactive power optimization scheme for the large power grid based on the current optimal solution.

[0059] It should be noted that power grid parameters can be understood as data describing the operating status of a large power grid, such as node voltage, line impedance, and load demand. These are mainly used to define the constraints and objective function of the optimization problem. Water cycle optimization algorithm parameters, on the other hand, include configuration parameters related to the algorithm's behavior, such as the water flow rate, initial distribution range, and iterative control parameters. Their role is to regulate the algorithm's search behavior.

[0060] More specifically, the design of the penalty function for large power grid optimization can be implemented using a weighted summation method. For example, the penalty value can be the weighted sum of the objective function value and the degree of constraint violation. The aim is to consider both the objective optimization effect and the constraint satisfaction during the optimization process. The implementation of the improved flow direction update formula can be designed based on fixed rules, such as using linear interpolation or gradient descent to adjust the water flow position to gradually approach the current optimal solution. Furthermore, the evaporation-rainfall handling mechanism can be implemented by simulating random perturbations in natural phenomena, such as introducing random noise or regenerating part of the water flow, to avoid the algorithm getting trapped in local optima.

[0061] Specifically, the working principle of this application embodiment is as follows:

[0062] First, by acquiring the power grid parameters and water cycle optimization algorithm parameters, necessary input data is provided for subsequent steps. The power grid parameters define the variable constraints and optimization objectives, while the water cycle optimization algorithm parameters control the search behavior. Further, several water flows are initialized based on these parameters, each corresponding to a set of decision variables. This problem-characteristic-based initialization method ensures a high degree of matching between the initial water flow distribution and the solution space characteristics, avoiding uneven distribution caused by random initialization, thereby improving the diversity and quality of the initial solutions.

[0063] Subsequently, the penalty value for each water flow is calculated using the large power grid optimization penalty function, and this value is used to initialize the current optimal solution. This penalty function combines the large power grid optimization objective function and constraint penalty terms, enabling simultaneous quantification of the objective optimization effect and the degree of constraint violation. This allows for comprehensive consideration of feasibility and economy when initializing the current optimal solution, thereby enhancing the ability to handle complex operational constraints.

[0064] Next, the water flow position information is updated according to a pre-defined improved flow direction update formula. This formula introduces a dynamic inertia weight parameter and an update coefficient to dynamically adjust the water flow direction. Specifically, the dynamic inertia weight parameter decreases linearly with each iteration, allowing the algorithm to prioritize expanding the search range for global exploration in early iterations, while focusing on local fine-tuning in later iterations. Simultaneously, by combining a distance factor to dynamically balance the search breadth and depth, the update coefficient guides the water flow to converge directionally to the optimal solution, effectively addressing the problem of slow convergence in later stages of classic algorithms.

[0065] Furthermore, based on the updated water flow and the newly generated water flow through the evaporation-rainfall processing mechanism, the penalty value for each water flow is recalculated using the large power grid optimization penalty function, and the current optimal solution is updated accordingly. The evaporation-rainfall processing mechanism avoids information loss by preserving high-quality solutions and employs an adaptive perturbation intensity decay strategy, making the early perturbations larger to expand the search breadth and the later perturbations smaller to improve the optimization accuracy, thereby enhancing the algorithm's ability to escape local optima.

[0066] Finally, it is determined whether the maximum number of iterations has been reached based on the current iteration count. If not, the system returns to update the water flow position information according to the improved flow direction update formula; if the maximum number of iterations has been reached, the current optimal solution is output and converted into an executable reactive power optimization scheme.

[0067] Thus, through the coordinated operation of the above steps, an efficient solution to the reactive power optimization problem of large power grids is achieved, solving the technical problems of slow convergence, insufficient convergence accuracy, and insufficient constraint handling capability of the classical water cycle optimization algorithm in the later stage.

[0068] Furthermore, the expression for the improved flow direction update formula mentioned in the above embodiments can be found here:

[0069]

[0070] In the formula, Let m be the value of the water flow in the t-th iteration; Let be the optimal water flow value in the t-th iteration. Let be the dynamic inertia weight parameter for the t-th iteration. Let m be the distance factor of the water flow in the t-th iteration, and its expression can be: , Let be the update coefficient of water flow m in the t-th iteration.

[0071] It should be noted that the improved flow direction update formula proposed in this embodiment operates by achieving an efficient optimization process through the organic combination of several key elements. Firstly, the dynamic inertia weight parameter... As the iteration process gradually decreases, this pattern allows the algorithm to maintain a large search range in the early stages to discover potential high-quality solutions, while narrowing the search range in the later stages for refined optimization. Secondly, the distance factor... The search range is dynamically adjusted based on the actual distance between the water flow and the current optimal solution. When the distance is large, the search range is expanded to enhance global exploration capabilities, while when the distance is small, the search range is narrowed to focus on local optimization. Furthermore, the coefficients are updated. By quantifying the update direction between the water flow and the optimal solution, we ensure that each iteration converges more accurately to the optimal value. These elements work together to form a complete optimization mechanism that can adaptively adjust the search strategy, thereby effectively solving the problems of decreased convergence speed and insufficient optimization accuracy.

[0072] The above technical solutions not only achieve the goal of dynamically adjusting the search strategy, but also significantly improve the performance of the algorithm in dealing with high-dimensional and complex problems such as reactive power optimization of large power grids.

[0073] Furthermore, the calculation formulas for the dynamic inertia weight parameter and update parameter mentioned in the improved flow direction update formula are as follows:

[0074] Formula for calculating dynamic inertia weight parameters:

[0075]

[0076] In the formula, T is the maximum number of iterations set. and These are the upper and lower limits of the dynamic inertia weight parameter, respectively.

[0077] It should be noted that the dynamic inertia weight parameter is a key factor used to adjust the update step size of the water flow position during the optimization process, and it achieves adaptive adjustment through linear decrease. In practical applications, this parameter can be normalized based on the ratio of the current iteration number to the maximum iteration number. The purpose is to ensure that the algorithm has appropriate search capabilities at different stages. The maximum iteration number serves as a normalization benchmark, effectively synchronizing the adjustment process of the inertia weight. The upper and lower limits can be set by pre-setting fixed values ​​or dynamically determined according to the specific problem. The purpose is to limit the reasonable range of the weights, thereby ensuring the stability and robustness of the algorithm.

[0078] Specifically, the above technical solution achieves a dynamic balance between the algorithm's global exploration and local exploitation capabilities through a linearly decreasing dynamic inertia weight parameter. In the early stages of the algorithm, due to the small number of iterations, the inertia weight is close to its upper limit, and a larger search step size helps expand the search range, preventing premature entrapment in local optima. As the number of iterations increases, the inertia weight gradually decreases. In the later stages of the algorithm, when it approaches its lower limit, a smaller search step size can improve local exploitation accuracy, thereby accelerating the convergence process. This design not only solves the problem of slow convergence in the later stages caused by the inertia weight not being dynamically adjusted with iterations, but also further improves the overall performance of the algorithm by combining it with the water flow position update mechanism. Furthermore, by combining the dynamic inertia weight parameter with the improved flow direction update formula, the water flow can be guided more effectively to converge towards a better region, thereby improving the optimization effect.

[0079] Update coefficient calculation formula:

[0080]

[0081] In the formula, express and Covariance between and They represent variance and The variance.

[0082] It should be noted that in practical applications, covariance is a statistic used to measure the degree of linear correlation between two variables, and it can be implemented using either the sample covariance formula or the population covariance formula. Variance, on the other hand, is an indicator of the dispersion of a single variable, and it can be calculated using either sample variance or population variance. The purpose of this design is to improve the convergence efficiency of the algorithm in the later stages of iteration by dynamically adjusting the updated coefficients through the introduction of a ratio of covariance to variance.

[0083] Specifically, the formula for updating coefficients is essentially based on the statistical correlation between the current water flow and the globally optimal water flow (ocean). It uses "covariance" to capture the direction of the correlation between the current solution and the optimal solution, and uses "variance" to adjust the stability of the update step size, providing "directional weights" for water flow updates. This ensures the water flow always adjusts towards the optimal solution, shortening the optimization path and improving convergence speed. Its core value lies in improving the algorithm's convergence speed and optimization accuracy without sacrificing global exploration capabilities, providing an efficient solution update mechanism for high-dimensional, complex constraint problems such as reactive power optimization in large power grids. This mechanism, dynamically adjusted based on real-time solution distribution, enables the algorithm to intelligently balance global search and local development capabilities. Especially in the high-dimensional and complex scenarios of reactive power optimization in large power grids, this design ensures that the water flow update process closely matches the characteristics of the solution space, thereby improving the algorithm's ability to escape local optima while maintaining the stability and accuracy of later convergence.

[0084] Furthermore, this application also proposes an evaporation-rainfall processing mechanism that includes: storing the optimal water flow and high-quality water flow in each iteration into a preset buffer space, and generating new water flow based on the water flow in the buffer space through a preset evaporation-rainfall processing formula, wherein the high-quality water flow is any water flow whose distance factor from the optimal water flow is less than a preset distance threshold.

[0085] It should be noted that the cache space mentioned in this embodiment refers to an area for temporary data storage, which can be implemented using memory, hard disk, or other devices with storage capabilities. Its purpose is to preserve high-quality solutions generated during the iteration process and prevent the loss of excellent solutions due to randomness. In practical applications, the distance factor is an indicator that measures the proximity between water flows. It can be calculated using Euclidean distance, Manhattan distance, etc. Its purpose is to dynamically filter out high-quality solutions that are close to the optimal solution, ensuring that the search process focuses on potential optimization regions.

[0086] Specifically, this scheme improves the stability and efficiency of the algorithm during the search process by reconstructing the core process of the evaporation-rainfall processing mechanism. First, the optimal and high-quality water flows from each iteration are stored in a pre-defined cache space. This ensures that high-quality solutions are continuously retained throughout the iterations, avoiding the loss of elite solutions caused by the random generation of new solutions in traditional mechanisms, thus maintaining the continuity of the search process and the accumulation effect of high-quality solutions. Second, new water flows are generated based on the water flows in the cache space, utilizing the stored high-quality solutions as the basis for generation. This makes new solutions more likely to be located in the potential optimization region, reducing the randomness of blind searches. This ensures that the newly generated water flows retain the core characteristics of high-quality solutions while effectively exploring the solution space. Simultaneously, combined with a dynamic filtering mechanism using distance factors, the generation process of new water flows can adaptively focus on the high-quality solution set, avoiding the introduction of inferior solutions that interfere with the search direction. Furthermore, a high-quality water flow is defined as any water flow whose distance factor from the optimal water flow is less than a preset distance threshold. This screening strategy dynamically measures the proximity of the water flow to the optimal solution through the distance factor, and the threshold setting ensures that only high-quality solutions are retained. This allows the search process to expand the exploration range to discover new areas in the early stage of iteration, and to finely adjust it in the later stage to improve the convergence accuracy. This effectively enhances the algorithm's ability to escape local optima and optimizes the balance between global and local searches.

[0087] Based on this, the above-mentioned evaporation-rainfall processing mechanism is combined with the overall framework of the large power grid reactive power optimization method. Through the design of elite retention and adaptive perturbation intensity decay with iteration, not only is the global search capability of the algorithm improved, but also the fineness of local development is enhanced. This better solves the technical problems of decreased convergence speed and easy getting trapped in local optima in high-dimensional complex constraint problems.

[0088] More specifically, this embodiment also proposes an evaporation-rainfall treatment formula as follows:

[0089]

[0090] In the formula, This is a new water flow formed through an evaporation-rainfall treatment mechanism; The water flow stored in the cache space; It follows a standard normal distribution; and These represent vectors consisting of the maximum and minimum values ​​of all decision variables, respectively. The disturbance coefficients can be calculated using an adaptive approach to update the disturbance strategy.

[0091] The expression for the adaptive perturbation coefficient update strategy can be found here:

[0092] in, This is the initial standard deviation of the disturbance, which can generally be taken as 0.4. For natural index, This is the attenuation coefficient.

[0093] Specifically, this technical solution implements the perturbation mechanism by defining a precise mathematical expression. It utilizes historical high-quality solutions stored in the cache space as the basis for perturbation, avoiding fluctuations in solution quality caused by random perturbations while preserving the core characteristics of elite solutions. The random perturbation introduced by the standard normal distribution helps to uniformly explore new regions in the solution space, enhancing the algorithm's ability to escape local optima. Perturbation scaling is based on a vector composed of the maximum and minimum values ​​of the decision variables, ensuring that the perturbation amplitude matches the actual scale of the solution space, keeping the newly generated water flow within the feasible region. By dynamically adjusting the perturbation intensity through the perturbation coefficient, adaptive control is achieved, with large perturbations in the early stages to expand the search range and small perturbations in the later stages for refined optimization.

[0094] Furthermore, this application proposes a step of initializing several water flows based on power grid parameters and water cycle optimization algorithm parameters, including: initializing several water flows, i.e., column vectors composed of decision variables, through Logistic chaotic mapping based on power grid parameters and water cycle optimization algorithm parameters.

[0095] In practical applications, the Logistic chaotic mapping method refers to a mathematical mapping approach based on nonlinear dynamic systems. The purpose of this method is to leverage the ergodicity and randomness of chaotic sequences to generate a uniformly distributed set of points covering the entire solution space, thereby effectively avoiding the clustering of solution points caused by traditional random initialization. The adaptive adjustment mechanism for variable dimensions can be understood as a technique for dynamically adjusting mapping parameters. This can be achieved by setting an adjustment function related to the dimension of the decision variables, aiming to adapt the initialization process to optimization problems of different scales and prevent locally dense or sparse uneven distributions in high-dimensional problems.

[0096] More specifically, the Logistic chaotic mapping expression mentioned in this embodiment can be seen in the following example:

[0097]

[0098] in, Let k be the value of the k-th variable in the column vector of decision variables. and These are the upper and lower bounds of the k-th variable, respectively. n is the sequence number of the initial water flow. and A chaotic sequence that satisfies:

[0099]

[0100] in, The chaos coefficient is preferably set to 5.0 to ensure chaotic characteristics. The initial value is Take 0.3691.

[0101] Through the above technical solutions, the diversity and uniformity of the initial solutions are significantly improved, providing a better starting point for subsequent optimization iterations and effectively enhancing the algorithm's global search capability and convergence efficiency.

[0102] The above is a detailed description of the technical logic of an embodiment of a large power grid reactive power optimization method provided in this application. Below is a verification example obtained using the above-mentioned large power grid reactive power optimization method based on power grid data from a provincial power grid, as detailed below:

[0103] The provincial power grid has a load capacity exceeding 160,000 MW, and the installed capacity of various power sources is shown in Table 1 below. The number of nodes at different voltage levels in this power grid exceeds 20,000, making the optimization scale large and the solution quite difficult.

[0104] Table 1. Installed capacity of various power sources in a provincial power grid

[0105]

[0106] The optimization results of the improved water cycle optimization method proposed in this embodiment and its comparison with other optimization methods are shown in Table 2 below:

[0107] Table 2 Comparison of results between the improved water cycle optimization algorithm and other optimization algorithms

[0108]

[0109] The results in the table show that, compared with classical water cycle optimization algorithms, genetic algorithms, particle swarm optimization algorithms, reinforcement learning algorithms, and interior-point methods, the improved water cycle optimization algorithm proposed in this application achieves the lowest network loss power, indicating that it is closest to the global optimum and has the best optimization effect. This also indirectly proves that the improvement proposed in this application to the classical water cycle optimization algorithm is effective. Furthermore, the improved water cycle optimization algorithm requires the least optimization time among these methods, exhibiting the highest optimization efficiency, and is therefore more suitable for reactive power optimization problems in large power grids.

[0110] The above is a detailed description of an embodiment of a reactive power optimization method for a large power grid provided in this application. The following are detailed descriptions of related embodiments of a reactive power optimization device, terminal, and storage medium for a large power grid provided in this application.

[0111] Please see Figure 2 The embodiment of a large power grid reactive power optimization device proposed in this application mainly comprises:

[0112] The parameter acquisition unit 201 is used to acquire the power grid parameters and water cycle optimization algorithm parameters of the large power grid;

[0113] The water flow initialization unit 202 is used to initialize several water flows according to the power grid parameters and the water cycle optimization algorithm parameters, wherein each water flow corresponds to a set of decision variables;

[0114] The optimal solution initialization unit 203 is used to calculate the penalty value of each water flow through the large power grid optimization penalty function, so as to initialize the current optimal solution according to the penalty value. The large power grid optimization penalty function is constructed based on the large power grid optimization objective function combined with the constraint penalty term.

[0115] The water flow position update unit 204 is used to update the water flow position information according to the preset improved flow direction update formula. The improved flow direction update formula is used to converge the water flow position to a direction closer to the current optimal solution according to the dynamic inertia weight parameter and the update coefficient.

[0116] The optimal solution update unit 205 is used to calculate the penalty value of each water flow based on the water flow after the update position and the water flow newly generated through the evaporation-rainfall treatment mechanism, through the large power grid optimization penalty function, so as to update the current optimal solution based on the penalty value;

[0117] The cyclic control unit 206 is used to update the water flow position information according to the preset improved flow direction update formula if the current number of cyclic iterations has not reached the maximum number of iterations, until the number of cyclic iterations reaches the maximum number of iterations, and then output the current optimal solution to determine the reactive power optimization scheme of the power grid based on the current optimal solution.

[0118] This embodiment addresses the slow convergence and insufficient convergence accuracy issues of classic water cycle optimization algorithms by combining dynamic inertia weight parameters with update coefficients and introducing an evaporation-rainfall processing mechanism. This improves both global search capability and local optimization capability. Specifically, the dynamic inertia weight parameters decrease linearly with iteration, achieving a smooth transition from global exploration to refined local optimization. The update coefficients guide the water flow towards the optimal solution through covariance calculation, further enhancing the algorithm's convergence performance. Furthermore, the evaporation-rainfall processing mechanism, through an adaptive perturbation intensity attenuation strategy, expands the search breadth in the early stages while improving optimization accuracy in the later stages, effectively preventing the algorithm from getting trapped in local optima and ensuring the efficiency and reliability of the optimization results.

[0119] Furthermore, such as Figure 3As shown in the figure, this application embodiment also discloses a large power grid reactive power optimization terminal, including: a memory and a processor; the memory is used to store program code, which corresponds to the large power grid reactive power optimization method; the processor is used to read and execute the program code to implement the large power grid reactive power optimization method.

[0120] In this context, "memory" refers to the hardware device used to store program code, which can be implemented using flash memory, hard disk, or solid-state drive. Its function is to provide the processor with a set of executable instructions, ensuring that the logic of the control method can be fully invoked. "Processor" refers to the arithmetic unit used to execute the program code, which can be implemented using a microprocessor, digital signal processor, or application-specific integrated circuit. Its function is to calculate and output the current optimal solution in real time by running the program code, thereby achieving dynamic adjustment of reactive power optimization for the large power grid. "Program code" refers to the set of computer instructions containing the steps of the transient stability enhancement control method for grid-connected converters. Its function is to translate the control algorithm into executable machine instructions, ensuring that the terminal can automatically complete the optimization of reactive power optimization schemes for the large power grid.

[0121] In some implementations, the memory may employ non-volatile storage media, such as EEPROM or NOR flash memory, to ensure that program code is retained after power loss. The processor may integrate a multi-core architecture, such as using an ARM Cortex-M series microcontroller, to enhance parallel computing capabilities.

[0122] Furthermore, this application also proposes an embodiment of a computer-readable storage medium, wherein program code is stored therein, the program code being read and executed by a processor to improve the reactive power optimization performance of a large power grid and enhance the engineering applicability of the water cycle optimization algorithm in power grid reactive power optimization.

[0123] Computer-readable storage media refers to physical carriers capable of long-term storage of program code, such as solid-state drives, flash memory chips, or optical discs. Its function is to provide repeatedly readable code storage support for the execution of control methods. Program code refers to a set of code containing computer instructions, which can be written in C, Python, or assembly language. Its function is to transform the logic of the control method into a sequence of instructions that can be parsed and executed by the processor. The processor is an integrated circuit with computation and control functions, which can be implemented using a central processing unit, digital signal processor, or microcontroller. Its function is to read and execute the program code in the storage medium to complete the optimization operation of the reactive power optimization scheme for a large power grid based on an improved water cycle optimization algorithm.

[0124] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the terminals, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0125] In the several embodiments provided in this application, it should be understood that the disclosed terminals, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between devices or units, and may be electrical, mechanical, or other forms.

[0126] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0127] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0128] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0129] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0130] If the integrated unit is implemented as 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. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0131] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A reactive power optimization method for a large power grid, characterized in that, include: Obtain the power grid parameters and water cycle optimization algorithm parameters of the large power grid; Based on the power grid parameters and the water cycle optimization algorithm parameters, several water flows are initialized, where each water flow corresponds to a set of decision variables; The penalty value for each water flow is calculated using the large power grid optimization penalty function, and the current optimal solution is initialized based on the penalty value. The large power grid optimization penalty function is constructed based on the large power grid optimization objective function combined with the constraint penalty term. The position information of the water flow is updated according to the preset improved flow direction update formula, wherein the improved flow direction update formula is used to converge the position of the water flow to a direction closer to the current optimal solution based on the dynamic inertia weight parameter and the update coefficient. Based on the updated water flow and the newly generated water flow through the evaporation-rainfall treatment mechanism, the penalty value for each water flow is calculated using the large power grid optimization penalty function, and the current optimal solution is updated based on the penalty value. If the current number of iterations has not reached the maximum number of iterations, the position information of the water flow is updated according to the preset improved flow direction update formula until the number of iterations reaches the maximum number of iterations. Then, the current optimal solution is output so as to determine the reactive power optimization scheme of the power grid based on the current optimal solution.

2. The reactive power optimization method for a large power grid according to claim 1, characterized in that, The improved flow direction update formula is as follows: In the formula, Let m be the value of the water flow in the t-th iteration; Let be the optimal water flow value in the t-th iteration. Let be the dynamic inertia weight parameter for the t-th iteration. Let m be the distance factor of the water flow m in the t-th iteration. Let be the update coefficient of water flow m in the t-th iteration.

3. The reactive power optimization method for a large power grid according to claim 2, characterized in that, The formula for calculating the update coefficient is: In the formula, express and Covariance between and They represent variance and The variance.

4. The reactive power optimization method for a large power grid according to claim 2, characterized in that, The formula for calculating the dynamic inertia weight parameter is: In the formula, T is the maximum number of iterations set. and These are the upper and lower limits of the dynamic inertia weight parameter, respectively.

5. The reactive power optimization method for a large power grid according to claim 1, characterized in that, The evaporation-rainfall treatment mechanism includes: The optimal and high-quality water flows in each iteration are stored in a preset cache space. Based on the water flows in the cache space, a new water flow is generated using a preset evaporation-rainfall processing formula. The high-quality water flow is any water flow whose distance factor from the optimal water flow is less than a preset distance threshold.

6. The reactive power optimization method for a large power grid according to claim 5, characterized in that, The evaporation-rainfall treatment formula is as follows: In the formula, This is a new water flow formed through an evaporation-rainfall treatment mechanism; The water flow stored in the buffer space; It follows a standard normal distribution; and These represent vectors consisting of the maximum and minimum values ​​of all decision variables, respectively. This represents the disturbance coefficient.

7. The reactive power optimization method for a large power grid according to claim 1, characterized in that, The initialization of several water flows based on the power grid parameters and the water cycle optimization algorithm parameters includes: Based on the power grid parameters and the water cycle optimization algorithm parameters, several water flows are initialized using a Logistic chaotic mapping method.

8. A reactive power optimization device for a large power grid, characterized in that, include: The parameter acquisition unit is used to acquire the power grid parameters and water cycle optimization algorithm parameters of the large power grid. The water flow initialization unit is used to initialize several water flows according to the power grid parameters and the water circulation optimization algorithm parameters, wherein each water flow corresponds to a set of decision variables; The optimal solution initialization unit is used to calculate the penalty value of each water flow through the large power grid optimization penalty function, so as to initialize the current optimal solution according to the penalty value. The large power grid optimization penalty function is constructed based on the large power grid optimization objective function combined with the constraint penalty term. The water flow position update unit is used to update the position information of the water flow according to a preset improved flow direction update formula, wherein the improved flow direction update formula is used to converge the water flow position to a direction closer to the current optimal solution according to the dynamic inertia weight parameter and the update coefficient. The optimal solution update unit is used to calculate the penalty value of each water flow based on the water flow after the update position and the newly generated water flow through the evaporation-rainfall treatment mechanism, through the large power grid optimization penalty function, so as to update the current optimal solution based on the penalty value; The cyclic control unit is used to update the position information of the water flow according to the preset improved flow direction update formula if the current number of cyclic iterations has not reached the maximum number of iterations, until the number of cyclic iterations reaches the maximum number of iterations, and then output the current optimal solution to determine the reactive power optimization scheme of the power grid based on the current optimal solution.

9. A reactive power optimization terminal for a large power grid, characterized in that, include: Memory and processor; The memory is used to store program code, which corresponds to the large power grid reactive power optimization method as described in any one of claims 1 to 7; The processor is used to read and execute the program code to implement the large power grid reactive power optimization method.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium contains program code that is read and executed by a processor to implement the large power grid reactive power optimization method as described in any one of claims 1 to 7.

Images