Power system dispatching method and device based on uncertainty and computer equipment

Abstract

The application relates to a power system scheduling method and device based on uncertainty and a computer device. The method comprises the following steps: acquiring real-time operation data of a power system, and determining an operation state of the power system according to the real-time operation data; determining at least one preselected scheduling model from at least one pre-trained candidate scheduling model according to a state type of the operation state; determining a respective matching degree of the operation state and the at least one preselected scheduling model according to the real-time operation data and the at least one preselected scheduling model; determining a target scheduling model from the at least one preselected scheduling model according to the respective matching degrees of the operation state and the at least one preselected scheduling model and the real-time operation data; generating a scheduling strategy for the power system by the target scheduling model based on the real-time operation data, and scheduling the power system according to the scheduling strategy, so as to improve the pertinence and effectiveness of the scheduling strategy and ensure the scheduling accuracy of the power system.

Description

Technical Field

[0001] This application relates to the field of power dispatching technology, and in particular to a power system dispatching method, apparatus, computer equipment, computer-readable storage medium, and computer program product based on uncertainty. Background Technology

[0002] With the increasing complexity and dynamism of power systems, their operation and dispatch face a growing number of uncertainties, primarily stemming from load fluctuations, intermittent fluctuations in renewable energy sources, and equipment failures. To better address these uncertainties, optimized power system dispatch requires the introduction of various uncertainty modeling methods to improve the economic efficiency, reliability, and sustainability of the power system.

[0003] Currently, there are various models for uncertainty in power systems. Different models have their own advantages and disadvantages in different application scenarios. However, the accuracy of power system scheduling using existing uncertainty modeling methods is relatively low. Summary of the Invention

[0004] Therefore, it is necessary to provide an uncertainty-based power system dispatching method, apparatus, computer equipment, computer-readable storage medium, and computer program product that can improve the dispatching accuracy of power systems, addressing the aforementioned technical problems.

[0005] In a first aspect, this application provides a power system dispatching method based on uncertainty, the method comprising:

[0006] Acquire real-time operating data of the power system and determine the operating status of the power system based on the real-time operating data;

[0007] Based on the state type of the operating state, at least one pre-selected scheduling model is determined from at least one pre-trained candidate scheduling model; the candidate scheduling model is trained based on the uncertainty parameters of the power system and historical operating data, and the uncertainty parameters are used to characterize the fluctuation of the power system under the corresponding operating state of each state type;

[0008] Based on the real-time operation data and the at least one pre-selected scheduling model, determine the matching degree between the operation state and each of the at least one pre-selected scheduling model;

[0009] Based on the matching degree between the operating status and each of the at least one pre-selected scheduling models and the real-time operating data, a target scheduling model is determined from the at least one pre-selected scheduling model;

[0010] Based on the real-time operational data, a scheduling strategy for the power system is generated through the target scheduling model, and the power system is scheduled according to the scheduling strategy.

[0011] In one embodiment, determining the matching degree between the operating state and each of the at least one pre-selected scheduling model based on the real-time operating data and the at least one pre-selected scheduling model includes:

[0012] For each of the pre-selected scheduling models, the model performance parameters of the pre-selected scheduling model are determined based on the real-time running data; the model performance parameters are used to characterize the degree of matching between the pre-selected scheduling model and the corresponding state type.

[0013] The weighting coefficients of the model performance parameters are determined based on their relative importance when generating scheduling strategies for the pre-selected scheduling model.

[0014] The matching degree between the pre-selected scheduling model and the running state is determined based on the model performance parameters and the corresponding weight coefficients.

[0015] In one embodiment, the model performance parameters include at least one of the computational complexity, model accuracy, model stability parameters, and model adaptability parameters of the pre-selected scheduling model; determining the model performance parameters of the pre-selected scheduling model based on the real-time running data includes:

[0016] Based on the real-time operational data, determine at least one of the following parameters for the pre-selected scheduling model: computational complexity, model accuracy, model stability, and model adaptability.

[0017] In one embodiment, determining at least one of the computational complexity, model accuracy, model stability parameters, and model adaptability parameters of the pre-selected scheduling model based on the real-time runtime data includes at least one of the following:

[0018] Based on the data scale of the real-time running data and the model structure of the pre-selected scheduling model, the time complexity in terms of running time and the space complexity in terms of storage space of the pre-selected scheduling model are determined respectively, and the computational complexity is determined based on the time complexity and the space complexity.

[0019] Determine the predicted running data of the pre-selected scheduling model at the running time of the real-time running data, and determine the model accuracy based on the difference between the real-time running data and the predicted running data;

[0020] Based on the real-time operating data, the pre-selected strategies for the power system under different fluctuation conditions are determined using the pre-selected scheduling model, and the model stability parameters are determined according to the fluctuation range of each pre-selected strategy.

[0021] Based on the real-time operation data, the adaptability index of the pre-selected scheduling model under the operation state is determined, and the model adaptability parameters are determined according to the adaptability index and the index weight corresponding to each adaptability index.

[0022] In one embodiment, determining the target scheduling model from the at least one pre-selected scheduling models based on the matching degree between the operating state and each of the at least one pre-selected scheduling models and the real-time operating data includes:

[0023] Based on the matching degree between the operating status and the target pre-selected scheduling model and the real-time operating data, determine the adaptation parameters between the target pre-selected scheduling model and the operating status;

[0024] If the adaptation parameters meet the model parameter update conditions, the model parameters of the pre-selected scheduling model are iteratively updated until the termination conditions are met, and the global optimization parameters are obtained.

[0025] The target scheduling model is determined from the at least one pre-selected scheduling model based on the global optimization parameters corresponding to each of the at least one pre-selected scheduling model.

[0026] In one embodiment, the method further includes:

[0027] Obtain the scheduling execution result of the power system based on the scheduling strategy, and determine the scheduling difference between the scheduling execution result and the expected scheduling result;

[0028] If the scheduling differences satisfy the model update conditions, the operating data of the power system after the scheduling is executed are determined.

[0029] The real-time operating data of the power system is determined based on the operating data after the scheduling execution, and the steps of obtaining the real-time operating data of the power system and determining the operating status of the power system based on the real-time operating data are returned, or the steps of determining the matching degree between the operating status and the at least one pre-selected scheduling model based on the real-time operating data and the at least one pre-selected scheduling model are returned.

[0030] Secondly, this application also provides an uncertainty-based power system dispatching device, the device comprising:

[0031] The status determination module is used to acquire real-time operating data of the power system and determine the operating status of the power system based on the real-time operating data.

[0032] A model selection model is used to determine at least one pre-selected scheduling model from at least one pre-trained candidate scheduling model based on the state type of the operating state; the candidate scheduling model is trained based on the uncertainty parameters of the power system and historical operating data, and the uncertainty parameters are used to characterize the fluctuation of the power system under the corresponding operating state of each state type;

[0033] The matching degree determination module is used to determine the matching degree between the running state and the at least one pre-selected scheduling model based on the real-time running data and the at least one pre-selected scheduling model.

[0034] The model matching module is used to determine the target scheduling model from the at least one pre-selected scheduling model based on the matching degree between the running status and the at least one pre-selected scheduling model and the real-time running data.

[0035] The operation and scheduling module is used to generate a scheduling strategy for the power system based on the real-time operation data and the target scheduling model, and to schedule the power system according to the scheduling strategy.

[0036] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described above.

[0037] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0038] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described above.

[0039] The aforementioned uncertainty-based power system dispatching method, apparatus, computer equipment, computer-readable storage medium, and computer program product acquire real-time operating data of the power system and determine the operating state of the power system based on the real-time operating data; determine at least one pre-selected dispatching model from at least one pre-trained candidate dispatching model according to the state type of the operating state; the candidate dispatching models are trained based on the uncertainty parameters of the power system and historical operating data, and the uncertainty parameters are used to characterize the fluctuation of the power system under the corresponding operating state of each state type; determine the matching degree between the operating state and at least one pre-selected dispatching model according to the real-time operating data and at least one pre-selected dispatching model; and determine the appropriate dispatching model from at least one pre-selected dispatching model based on the matching degree between the operating state and at least one pre-selected dispatching model and the real-time operating data. The target scheduling model generates scheduling strategies for the power system based on real-time operational data, and then schedules the power system according to these strategies. It dynamically determines the operating state of the power system based on real-time operational data and dynamically evaluates the matching degree between pre-selected scheduling models and the power system under the current operating state. Based on the matching degree, it selects a target scheduling model from the pre-selected models that matches the current operating state. Finally, it generates corresponding scheduling strategies through the target scheduling model to schedule the power system. This ensures that the selected target scheduling model accurately adapts to the actual operating state of the power system and dynamically generates corresponding scheduling strategies based on real-time operational data. It effectively addresses uncertainties such as load fluctuations and the intermittency of renewable energy, thereby improving the targeting and effectiveness of the scheduling strategy and ensuring the accuracy of power system scheduling. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a diagram illustrating the application environment of an uncertainty-based power system dispatching method in one embodiment.

[0042] Figure 2 This is a flowchart illustrating an uncertainty-based power system dispatching method in one embodiment;

[0043] Figure 3 This is a flowchart illustrating the steps for determining the matching degree in one embodiment;

[0044] Figure 4 This is a structural block diagram of a power system dispatching device based on uncertainty in one embodiment;

[0045] Figure 5 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0047] The uncertainty-based power system dispatching method provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be integrated onto server 104 or placed on a cloud or other network server. Server 104 acquires real-time operating data of the power system and determines the operating status of the power system based on the real-time operating data. Server 104 then determines at least one pre-selected scheduling model from at least one pre-trained candidate scheduling model based on the status type of the operating status. This candidate scheduling model is trained based on the uncertainty parameters of the power system and historical operating data; the uncertainty parameters characterize the fluctuations of the power system under the corresponding operating status type. Subsequently, server 104 determines the matching degree between the operating status and at least one pre-selected scheduling model based on the real-time operating data and at least one pre-selected scheduling model, and determines a target scheduling model from at least one pre-selected scheduling model based on the matching degree between the operating status and at least one pre-selected scheduling model and the real-time operating data. Finally, server 104 generates a scheduling strategy for the power system based on the real-time operating data and the target scheduling model, and schedules the power system according to the scheduling strategy.

[0048] The terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle systems, and projection devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted displays. Head-mounted displays can be virtual reality (VR) devices, augmented reality (AR) devices, and smart glasses. The server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0049] In one exemplary embodiment, such as Figure 2As shown, an uncertainty-based power system dispatching method is provided, which can be applied to... Figure 1 Taking server 104 as an example, it can be understood that this method can also be applied to... Figure 1 The terminal 102 in this embodiment can also be applied to a system including terminal 102 and server 104, and is implemented through the interaction between terminal 102 and server 104. The method in this embodiment includes the following steps 202 to 210. Wherein:

[0050] Step 202: Obtain real-time operating data of the power system and determine the operating status of the power system based on the real-time operating data.

[0051] Real-time operational data refers to data collected from the power system at the current moment that reflects the actual operating status of the power system. Real-time operational parameters may include, but are not limited to, electrical parameters, equipment status parameters, load parameters, and renewable energy parameters. For example, electrical parameters may include voltage, current, frequency, and power (active power and reactive power); equipment status parameters may include the status parameters of equipment such as generators, transformers, and transmission lines, such as the switching status (on or off) and temperature; load parameters may include the electricity demand data of various users in the power system, such as the load size and load distribution at different times; and renewable energy parameters may include the power generation of renewable energy sources participating in power generation in the power system.

[0052] Operating status refers to the overall working condition of the power system at a given moment, determined by real-time operating data. Operating status can include normal operating status, abnormal operating status, and fault operating status. For example, normal operating status means that all parameters in the power system are within normal ranges, equipment operates stably, and the power demand of users can be met. At this time, parameters such as voltage and frequency remain near their rated values, and there are no abnormal alarms from generation and transmission equipment. Abnormal operating status refers to phenomena in the system that deviate from normal conditions but have not yet reached the level of a fault. For example, some lines may experience slight overloads, or the temperature of some equipment may be slightly higher than normal but still within a safe and tolerable range. Fault operating status refers to a significant fault occurring in the power system, such as a line short circuit or equipment damage, causing system parameters to deviate significantly from normal values ​​and preventing normal operation.

[0053] For example, the server can acquire real-time operating data of the power system from data acquisition devices on various devices within the power system. These data acquisition devices may include various instrument transformers for collecting current and voltage, transmitters for collecting power, electricity meters for collecting current, frequency meters for collecting frequency, and various detection devices for collecting status data from power generation equipment, substation equipment, and transmission lines. The server can also determine the operating status of the power system based on whether the real-time operating data is within the normal range.

[0054] Step 204: Determine at least one pre-selected scheduling model from at least one pre-trained candidate scheduling model based on the state type of the running state.

[0055] Candidate scheduling models are models trained based on the uncertainty parameters and historical operating data of the power system. They can be used to generate scheduling strategies for the power system under different operating conditions. Uncertainty parameters characterize the fluctuations of the power system under its respective operating conditions. For example, in load forecasting, due to the randomness and uncertainty of user electricity consumption behavior, load data will have a certain range of fluctuations; in this case, the fluctuation of load data can be described as an uncertainty parameter. Similarly, in the forecasting of renewable energy generation (such as wind power and photovoltaic power), since renewable energy generation usually directly depends on fluctuations in natural conditions (such as wind speed, solar intensity, cloud cover, etc.), the generation of renewable energy is uncertain; in this case, the fluctuation of renewable energy generation can also be described as an uncertainty parameter. Uncertainty parameters help models better adapt to changes in power system operation and improve the robustness of scheduling strategies. Historical operating data can include various information about the power system under different operating conditions in the past, such as load curves, unit output / start-up and shutdown, generation plans, transmission schemes, load changes, electrical parameters, etc. In practice, historical operating data can be obtained through the power system's SCADA system (Supervisory Control And Data Acquisition), enabling candidate scheduling models to learn from historical operating data and determine scheduling strategies under different operating conditions, thereby providing a reference for current power dispatching decisions.

[0056] A pre-selected dispatch model refers to a model selected from at least one candidate dispatch model based on the operating state type of the power system. Operating state types can include normal, abnormal, and fault states, and within normal and abnormal states, various uncertainties in the power system can be included. For example, when the power system is in normal operation, candidate dispatch models that perform well under normal load conditions can be selected as pre-selected dispatch models; while when the system is in an abnormal or fault state, models capable of handling the corresponding abnormal or fault conditions can be selected. Similarly, in operating states with large load fluctuations, models capable of handling load fluctuations can be selected, and in operating states with large fluctuations in renewable energy generation, models capable of handling renewable energy fluctuations can be selected.

[0057] In some exemplary embodiments, candidate scheduling models can be constructed using at least one of the following methods: probabilistic modeling methods (such as Monte Carlo simulation, probabilistic constraint optimization), fuzzy set theory modeling methods (such as fuzzy logic control), rough set theory modeling methods, and artificial intelligence modeling methods (such as neural networks, genetic algorithms, particle swarm optimization, etc.). Candidate scheduling models obtained using different modeling methods can address scheduling decisions for power systems under different operating states.

[0058] Probabilistic modeling methods can describe uncertainties in power systems by introducing random variables. For example, in a power system, if the power generation of a certain load or renewable energy source is a random variable, then the goal of probabilistic modeling methods is to solve for the optimal dispatch scheme, i.e.:

[0059] (1)

[0060] in, The scheduling cost function; Let be the probability density function of the uncertain variable; This is an uncertain space; by simulating multiple operating scenarios through random sampling, the scheduling cost of each scenario can be determined, and then the overall expected cost can be obtained through statistical analysis.

[0061] Fuzzy set theory modeling methods can be used to handle situations with fuzziness and uncertainty. For example, consider the load demand in a power system. If it is a fuzzy set, then fuzzy rules can be defined. To make inferences:

[0062] (2)

[0063] in, For fuzzy rules; For load demand Fuzzy linguistic variables, such as "high load" and "low load"; For control variables, such as the generator's output power; For control variables Fuzzy linguistic variables; that is, based on fuzzy rules. ,when Belongs to set At that time, Belongs to set Thus, after defuzzification through fuzzy reasoning, a specific scheduling scheme can be obtained.

[0064] Rough set theory can be used to handle problems with incomplete information. For example, in a power system, suppose the power system has a set of states... Incomplete or ambiguous For the state of the power system, Let be the number of states. When using rough set theory for uncertainty modeling, the objective of rough set theory can be to provide decision support through conceptual approximation, i.e.:

[0065] (3)

[0066] in, By establishing equivalence relationships between state concepts and handling uncertainties in data through approximate classification and reduction, decisions on determining scheduling schemes can be made.

[0067] Artificial intelligence modeling methods can be used for load forecasting or generation forecasting. For example, in a power system, when using neural networks for load forecasting or generation forecasting, it is assumed that historical data on load and renewable energy generation are... , For historical data, The amount of historical data. Neural networks predict future load or renewable energy generation by learning the non-linear relationships between historical data.

[0068] (4)

[0069] in, For the predicted The load at any given time or the amount of electricity generated by renewable energy sources. This is the mapping function for the neural network.

[0070] In practice, the multiple candidate scheduling models obtained from training can be stored in a pre-created model library. For each operating state of the power system, there is at least one candidate scheduling model in the model library that can handle the uncertainty in that operating state. Each candidate scheduling model in the model library corresponds to an operating state, and the two can be determined by a defined mapping relationship or label information, so that the pre-selected scheduling model can be matched from the model library by the determined operating state.

[0071] For example, the server can determine the candidate scheduling model that can cope with the current state type based on the mapping relationship between the state type of the power system operation and the candidate scheduling models in the model library, so as to obtain the pre-selected scheduling model.

[0072] In other embodiments, the operating status of the power system may also be affected by climate or weather factors, which in turn affect power dispatch. For example, under extreme weather conditions (such as rainstorms, typhoons, etc.), the power system needs to adjust the power generation plan and transmission scheme according to meteorological data to ensure the stability and security of power supply. Therefore, when determining the pre-selected dispatch model, meteorological data can also be integrated for comprehensive determination. For example, weights can be set according to the correlation between the state type and meteorological data, or a correlation coefficient can be determined to match it with the candidate dispatch models in the model library, thereby obtaining the pre-selected dispatch model.

[0073] Step 206: Based on real-time operating data and at least one pre-selected scheduling model, determine the matching degree between the operating status and each of the at least one pre-selected scheduling model.

[0074] The matching degree refers to a parameter used to measure the degree of fit between the operating state of the power system and various pre-selected scheduling models. The higher the matching degree, the more suitable the pre-selected scheduling model is for the current operating state. In practice, the degree of fit between the operating state and the pre-selected scheduling model can be scored to obtain the matching degree between the operating state and the pre-selected scheduling model.

[0075] For example, the server can score the degree of fit between the running state and the pre-selected scheduling model based on real-time running data and the corresponding pre-selected scheduling model, and determine the degree of matching between the running state and at least one pre-selected scheduling model based on the score.

[0076] Step 208: Determine the target scheduling model from at least one pre-selected scheduling model based on the matching degree between the running status and at least one pre-selected scheduling model and the real-time running data.

[0077] The target scheduling model refers to the scheduling model that is suitable for the current operating state of the power system, determined from the pre-selected scheduling models based on the matching degree between the operating state and at least one pre-selected scheduling model and real-time operating data. The determined target scheduling model can effectively meet the current operating needs of the power system and optimize system performance (such as improving power supply reliability and reducing power generation costs). For example, when real-time operating data shows that the system load fluctuates greatly, a pre-selected scheduling model with strong adaptability to load changes may be selected as the target scheduling model.

[0078] For example, the server can employ an optimization algorithm to solve and optimize the matching degree between the running state and at least one pre-selected scheduling model, as well as the real-time running data, thereby determining the target scheduling model from at least one pre-selected scheduling model. The optimization algorithm can employ one or more methods, including but not limited to particle swarm optimization, ant colony optimization, genetic algorithm, and gradient descent.

[0079] Step 210: Based on real-time operational data, generate a scheduling strategy for the power system through the target scheduling model, and schedule the power system according to the scheduling strategy.

[0080] In this context, dispatching strategies refer to specific dispatching schemes used for the power system. These strategies are generated based on real-time operational data through a target dispatching model and may include, but are not limited to, generation plans, transmission schemes, and reserve arrangements. Generation plans determine the output of each power plant, including generator start-up and shutdown schedules and power allocation, to meet system load demands. Transmission schemes plan power transmission paths and capacity to ensure the safe and efficient delivery of electricity from power plants to users. Reserve arrangements determine the system's reserve capacity to cope with potential emergencies such as generator failures or sudden load increases.

[0081] Power system dispatching refers to the control and regulation of power generation, transmission, and distribution within a power system according to dispatching strategies, in order to achieve safe, stable, and economical operation of the power system. In practice, dispatching can be achieved either by a server generating corresponding control commands based on the dispatching strategy, which then control the power system's equipment through automated and control systems to regulate power generation output and perform switching operations on transmission lines; or by dispatchers making manual decisions and operations based on the dispatching strategy and actual conditions.

[0082] For example, the server can input real-time operational data into the target scheduling model, and the target scheduling model can generate a scheduling strategy for the power system based on the real-time operational data. The server can also generate control commands for each power device in the power system according to the scheduling strategy, and apply the control commands to the power devices in the power system to control the power devices to operate according to the control commands, thereby realizing the scheduling of the power system.

[0083] In the aforementioned uncertainty-based power system dispatching method, real-time operating data of the power system is acquired, and the operating state of the power system is determined based on this data. At least one pre-selected dispatching model is selected from at least one pre-trained candidate dispatching model based on the state type of the operating state. The candidate dispatching models are trained based on the uncertainty parameters of the power system and historical operating data; the uncertainty parameters characterize the fluctuations of the power system under the corresponding operating states. The matching degree between the operating state and each of the at least one pre-selected dispatching model is determined based on the real-time operating data and the at least one pre-selected dispatching model. A target dispatching model is then selected from the at least one pre-selected dispatching model based on the matching degree between the operating state and each of the at least one pre-selected dispatching model, along with the real-time operating data. Based on the real-time operating data... Data is used to generate scheduling strategies for the power system through a target scheduling model, and the power system is scheduled according to the scheduling strategies. The operating status of the power system is dynamically determined based on real-time operating data, and the matching degree between the pre-selected scheduling model and the power system under the current operating status is dynamically evaluated. Based on the matching degree, a target scheduling model that matches the current operating status is selected from the pre-selected scheduling models. Finally, the corresponding scheduling strategy is generated through the target scheduling model to schedule the power system. This ensures that the selected target scheduling model accurately adapts to the actual operating status of the power system, and the corresponding scheduling strategy is dynamically generated based on real-time operating data. This effectively copes with uncertainties such as load fluctuations and the intermittency of renewable energy, thereby improving the pertinence and effectiveness of the scheduling strategy and ensuring the accuracy of power system scheduling.

[0084] In some embodiments, such as Figure 3 As shown, step 206 includes steps 302 to 306. Wherein:

[0085] Step 302: For each pre-selected scheduling model, determine the model performance parameters of the pre-selected scheduling model based on real-time running data.

[0086] Among them, model performance parameters refer to indicators used to quantitatively describe the degree of matching between the pre-selected scheduling model and the corresponding state type under real-time running data, so as to reflect the performance of the pre-selected scheduling model in terms of accuracy, stability, adaptability, etc. Model performance parameters can serve as an important basis for evaluating the merits of the pre-selected scheduling model, so as to further select a scheduling model that is more suitable for the current running state from at least one pre-selected scheduling model.

[0087] For example, the server can determine the matching index between the pre-selected scheduling model and the corresponding state type based on real-time running data and different methods for each pre-selected scheduling model, and quantify each index to determine the model performance parameters of the pre-selected scheduling model.

[0088] Step 304: Determine the weighting coefficients of the model performance parameters based on their relative importance in generating the scheduling strategy for the selected pre-selected scheduling model.

[0089] Relative importance refers to the relative magnitude of the influence of each model performance parameter on the generated scheduling strategy when evaluating pre-selected scheduling models. The importance of different model performance parameters may vary under different operating conditions and requirements. Relative importance can be determined through expert experience, historical data analysis, and experimental verification. For example, in some cases, the accuracy of the scheduling strategy may be more important than its adaptability; while in others, stability may be the primary consideration. Therefore, assigning different weight coefficients to different model performance parameters can increase the accuracy of matching determination.

[0090] Weighting coefficients are numerical values ​​assigned to each model performance parameter based on its relative importance. They quantify the role of each parameter in determining the matching degree. The magnitude of the weighting coefficient reflects the proportion of each performance parameter in the overall matching degree determination, adjusting the influence of each parameter when comprehensively calculating the matching degree between the pre-selected scheduling model and the running state. This ensures that the matching degree calculation results more accurately reflect the actual performance of the model. In practice, weighting coefficients are non-negative, and the sum of the weighting coefficients of all model performance parameters is typically 1 to ensure the reasonableness and comparability of the evaluation results.

[0091] For example, the server can determine the relative importance of each model performance parameter when generating the scheduling policy for the pre-selected scheduling model by analyzing historical data, and then determine the weight coefficient of each model performance parameter according to the relative importance of each model performance parameter.

[0092] Step 306: Determine the matching degree between the pre-selected scheduling model and the running state based on the model performance parameters and the corresponding weight coefficients.

[0093] The matching degree comprehensively considers model performance parameters and their corresponding weighting coefficients to quantitatively evaluate the suitability between the pre-selected scheduling model and the current operating state of the power system. A higher matching degree indicates that the pre-selected scheduling model is more suitable for the current operating state. When determining the matching degree based on each model's performance parameters and weighting coefficients, a weighted average method is typically used. This involves multiplying the value of each model's performance parameter by its corresponding weighting coefficient and then summing all the products to obtain the matching degree value. As the basis for selecting the final scheduling model, the matching degree enables the server to choose the model suitable for the current operating state from multiple pre-selected scheduling models, thereby generating a scheduling strategy that matches the current operating state of the power system.

[0094] For example, the server can use a weighted average method to multiply the value of each model performance parameter by its corresponding weight coefficient, and then add all the products to determine the degree of matching between the pre-selected scheduling model and the running state.

[0095] In some optional embodiments, the model performance parameters include at least one of the computational complexity, model accuracy, model stability parameters, and model fitness parameters for the pre-selected scheduling model.

[0096] Computational complexity refers to the computational resources and time required by a pre-selection scheduling model to process real-time operational data. Computational complexity is typically related to the model's algorithm structure, data scale, and the complexity of its computational steps. As a metric for measuring the efficiency of a pre-selection scheduling model in processing real-time operational data, computational complexity is crucial in power system dispatching, which requires rapid response to system changes. Therefore, the lower the computational complexity of the pre-selection scheduling model, the better it can meet the demands of real-time dispatching. In practical implementation, the computational complexity of the pre-selection scheduling model can be evaluated by analyzing its algorithmic complexity (such as time complexity and space complexity) or its actual runtime computation time and memory usage.

[0097] Model accuracy refers to the degree of agreement between the scheduling strategy generated by the pre-selected scheduling model and the actual operating requirements of the power system. It reflects the predictive ability of the pre-selected scheduling model to predict the operating state of the power system and the effectiveness of the scheduling strategy. As an indicator of model accuracy, a high-accuracy model can generate scheduling strategies that are closer to actual needs, thereby improving the operating efficiency and stability of the power system. In practice, model accuracy can be evaluated by comparing the scheduling strategy generated by the model with actual operating data (such as load forecasting errors, deviations between generation plans and actual generation).

[0098] Model stability parameters refer to the ability of a pre-selected dispatch model to maintain stable performance when faced with noise, outliers, or changes in system state in real-time operational data. They reflect the robustness of the pre-selected dispatch model to uncertainty. As an indicator of the stability of a pre-selected dispatch model, model stability is crucial because operational data often contains noise and uncertainty. Therefore, in power systems, a stable pre-selected dispatch model is better able to maintain its performance under various conditions and avoid significant changes in dispatch strategies due to data fluctuations. In practice, model stability can be evaluated by repeatedly analyzing the model's performance under different noise levels, outliers, or changes in system state (such as the fluctuation range of prediction errors and the stability of dispatch strategies).

[0099] Model adaptability parameters refer to the ability of a pre-selected dispatch model to adjust its dispatch strategy to adapt to changes in the operating state of the power system. These parameters reflect the flexibility and adjustability of the pre-selected dispatch model. Since the operating state of the power system is constantly changing, the model adaptability parameters need to possess good adaptability to generate effective dispatch strategies under different conditions. In practice, model adaptability can be evaluated by analyzing the model's performance under different operating states (such as the speed of dispatch strategy adjustment and the degree of adaptation to new states).

[0100] Taking the model performance parameters, including the computational complexity, model accuracy, model stability, and model fitness parameters of the pre-selected scheduling model, as an example, weight coefficients are determined for the computational complexity, model accuracy, model stability, and model fitness parameters respectively. Then, the computational complexity, model accuracy, model stability, and model fitness parameters are multiplied by their respective weight coefficients, and all products are added together to obtain the matching degree value.

[0101] In an exemplary embodiment, the matching degree can be expressed as:

[0102] (5)

[0103] in, For the first The degree of matching between the pre-selected scheduling model and the running state; The first The computational complexity, model accuracy, model stability parameters, and model adaptability parameters of each pre-selected scheduling model; These are the weight coefficients corresponding to computational complexity, model accuracy, model stability parameters, and model fitness parameters, respectively.

[0104] In this embodiment, by determining the model performance parameters of the pre-selected scheduling model, and assigning weight coefficients to each model performance parameter according to the relative importance of the model performance parameters in the generation of scheduling strategies when the pre-selected scheduling model executes the scheduling strategy, and determining the matching degree based on the model performance parameters and their corresponding weight coefficients, the matching degree between the pre-selected scheduling model and the running state can be scientifically and reasonably quantified. This is beneficial for accurately and quickly selecting a scheduling model suitable for the current running state in subsequent steps.

[0105] In some embodiments, the model performance parameters for the pre-selected scheduling model are determined based on real-time runtime data, including:

[0106] Based on real-time operational data, determine at least one of the following parameters for the pre-selected scheduling model: computational complexity, model accuracy, model stability, and model adaptability.

[0107] For example, taking model performance parameters including computational complexity, model accuracy, model stability, and model adaptability as examples, the server can input real-time running data into each of the targeted pre-selected scheduling models to determine the computational complexity, model accuracy, model stability, and model adaptability parameters for each pre-selected scheduling model. In specific implementation, the server can calculate the computational complexity of the model by analyzing its algorithm structure and actual runtime computational resource consumption. The server can evaluate the model's accuracy by comparing the scheduling strategy generated by the pre-selected scheduling model with the actual running data and calculating indicators such as prediction error and deviation. The server can evaluate the model's stability by running the pre-selected scheduling model multiple times under different noise levels, outliers, or system state changes and observing its performance, calculating indicators such as the fluctuation range of prediction error. The server can evaluate the model's adaptability by running the pre-selected scheduling model under different operating conditions and observing the adjustment speed and adaptability of its scheduling strategy.

[0108] In this embodiment, by accurately evaluating the computational complexity, model accuracy, model stability, and model adaptability of the selected pre-selected scheduling model through real-time operational data, quantitative and objective model performance analysis can be provided. This helps to quickly identify the advantages and disadvantages of the selected pre-selected scheduling model under different operating conditions, ensuring that the selected pre-selected scheduling model can efficiently and accurately adapt to the actual power system operation requirements.

[0109] In some embodiments, based on real-time runtime data, at least one of the following is determined: computational complexity, model accuracy, model stability parameter, and model adaptability parameter for the pre-selected scheduling model, including at least one of the following:

[0110] Based on the data scale of the real-time running data and the model structure of the pre-selected scheduling model, the time complexity in terms of running time and the space complexity in terms of storage space of the pre-selected scheduling model are determined respectively, and the computational complexity is determined based on the time complexity and space complexity.

[0111] The data scale of real-time running data refers to the size or quantity of real-time running data. The data scale may include, but is not limited to, the number of data points, the dimensions of the data (such as multiple parameters such as voltage, current, and frequency), and the time span of the data. The model structure of the pre-selection scheduling model refers to the internal organization and algorithm logic of the pre-selection scheduling model. The model structure may include, but is not limited to, the hierarchical structure, the number of nodes, the connection method, and the type of algorithm used (such as neural networks, decision trees, linear regression, etc.).

[0112] Time complexity refers to the time resources required by a pre-selection scheduling model to process real-time running data. It reflects the trend of the time required for the pre-selection scheduling model to process data as the data size increases. Time complexity can be expressed as O(n), where n represents the data size. For example, the time complexity of a pre-selection scheduling model built based on Monte Carlo simulation can be expressed as O(N1×T1), where N1 is the number of times the real-time running data is sampled, and T1 is the running time of the pre-selection scheduling model. Similarly, the time complexity of a pre-selection scheduling model built based on probabilistic constraint optimization can be expressed as O(N2). 3 In the context of real-time scheduling, N² represents the dimension of the data being processed. For example, the time complexity of a pre-selection scheduling model built on a neural network can be expressed as O(L×N³×M), where L is the number of layers in the neural network, N³ is the number of nodes in each layer, and M is the number of training samples. Space complexity refers to the storage space resources required by the pre-selection scheduling model when processing real-time data. It reflects the amount of memory or disk space needed for data storage and processing. Space complexity can also be expressed as O(n), as detailed in the documentation on time complexity. By comprehensively considering both time and space complexity, computational complexity can provide an overall assessment of the computational resources and time required by the pre-selection scheduling model when processing real-time data.

[0113] For example, the server can determine the time complexity and space complexity of the pre-selected scheduling model based on the data scale of the real-time running data and the model structure of the pre-selected scheduling model, and then determine the computational complexity based on the time and space complexity. For instance, the server can determine the time complexity and space complexity of each pre-selected scheduling model based on the data dimension of the real-time running data, the number of model layers, the number of nodes per layer, etc., and the method used for the corresponding pre-selected scheduling model, and then combine the time and space complexity to obtain the computational complexity.

[0114] In some optional embodiments, when determining computational complexity, the server can assign appropriate weights to time complexity and space complexity based on the characteristics of the pre-selected scheduling model, and then weight and fuse the time and space complexity to obtain the computational complexity. For example, a pre-selected scheduling model with high storage space requirements can be assigned a larger weight.

[0115] In this embodiment, by combining the data scale of real-time running data and the model structure of the pre-selected scheduling model, the time and space complexity can be accurately determined, and then the computational complexity can be derived. This is beneficial for providing data support for model decision-making, thereby improving the overall system performance and resource utilization efficiency.

[0116] Determine the predicted running data of the pre-selected scheduling model at the running time of the real-time running data, and determine the model accuracy based on the difference between the real-time running data and the predicted running data.

[0117] In this context, operational prediction data refers to the power system operating state or parameter values ​​predicted by the pre-selected scheduling model based on its internal logic and algorithms at the time of real-time operation. The prediction error can be calculated by comparing the operational prediction data with the actual values ​​of the real-time operation data. Specifically, the operational prediction data can be data predicted by the pre-selected scheduling model based on data prior to the operating time. For example, the pre-selected scheduling model predicts based on the operating data at time T-1 to obtain the operational prediction data at time T. When the power system reaches time T, which is the time of real-time operation data, the real-time operating data at time T is compared with the operational prediction data at time T to determine the difference between them, thereby determining the model accuracy. Model accuracy can be measured by indicators such as prediction error and accuracy rate, for example, mean squared error and absolute error. Thus, the model accuracy can be evaluated based on the magnitude of the error.

[0118] For example, the server can input the operating data of the power system before the operating time of the real-time operating data into the pre-selected scheduling model, and predict the operating forecast data of the power system at the operating time through the internal logic and algorithm of the pre-selected scheduling model. Subsequently, the server can determine the accuracy of the model by calculating the mean square error between the real-time operating data and the operating forecast data.

[0119] In an optional embodiment, the model accuracy determined using the mean square error is expressed as follows:

[0120] (6)

[0121] in, Mean square error; For the first One set of predictive data, For the first Real-time running data, , The amount of data used for forecasting and real-time execution.

[0122] In this embodiment, by predicting in advance and comparing later, the predicted running data and the real-time running data being compared can be on the same time dimension to obtain accurate model accuracy.

[0123] Based on the selected pre-schedule model and real-time operating data, the pre-selection strategies for the power system under different fluctuation conditions are determined, and the model stability parameters are determined according to the fluctuation range of each pre-selection strategy.

[0124] The pre-selected strategy refers to the dispatching strategy or scheme generated by the pre-selected dispatching model based on real-time operational data, addressing different fluctuation conditions of the power system. It reflects the adaptability and flexibility of the pre-selected dispatching model under various operating conditions. Different fluctuation conditions may include, but are not limited to, load fluctuations, renewable energy generation fluctuations, and voltage fluctuations. The fluctuation range refers to the range of fluctuations or changes in the pre-selected strategy generated during the operation of the power system under different fluctuation conditions. A smaller fluctuation range indicates better stability of the pre-selected dispatching model, thus allowing the determination of the model's stability parameters. The model stability parameters can be comprehensively evaluated based on the performance of each pre-selected strategy under different fluctuation conditions (such as the adjustment frequency, adjustment magnitude, and contribution to system stability). By analyzing the performance of the pre-selected strategy within the fluctuation range, stability indices (such as volatility and stability coefficients) of the pre-selected strategy are calculated, and the stability parameters of the pre-selected dispatching model are determined based on these indices.

[0125] For example, the server can simulate the operation of a power system under different fluctuation conditions, and input real-time operating data into the corresponding pre-selected scheduling model for different fluctuation conditions to generate pre-selected strategies for the power system under different fluctuation conditions. By comparing the fluctuation range between different pre-selected strategies, the model stability parameters can be determined.

[0126] In this embodiment, by processing real-time running data multiple times under different fluctuation conditions through a pre-selected scheduling model, multiple pre-selected strategies can be obtained. Then, the stability parameters are determined based on the fluctuation range of the pre-selected strategies, which is beneficial for quantitatively evaluating the stability of the pre-selected scheduling model.

[0127] Based on real-time operational data, the adaptability index of the selected pre-selected scheduling model under operational conditions is determined. Based on the adaptability index and its corresponding weight, the model adaptability parameters are determined.

[0128] Adaptability metrics refer to various indicators based on real-time operational data to evaluate the adaptability and flexibility of a pre-selected scheduling model under operational conditions, reflecting its adaptability and adjustability under different operating conditions. Adaptability metrics may include, but are not limited to, accuracy, tracking performance, robustness, and response speed. Accuracy refers to the prediction error of the pre-selected scheduling model within a known time window. Tracking performance refers to the real-time adjustment capability of the pre-selected scheduling model when input fluctuations are large (e.g., the ability to automatically update model parameters). Robustness refers to the pre-selected scheduling model's ability to withstand disturbances from abnormal data or noisy inputs. Response speed refers to whether the running time of the pre-selected scheduling model meets real-time requirements. Metric weights refer to the proportion of each adaptability metric in the comprehensive evaluation of model adaptability, reflecting the importance of each metric in the overall evaluation; metric weights can be determined based on the needs of the actual application scenario and expert judgment. Model adaptability parameters can be obtained by weighted averaging or other comprehensive evaluation methods based on adaptability indicators and their corresponding weights. For example, the value of each adaptability indicator can be multiplied by its corresponding weight, and then all products can be added together to obtain the value of the model adaptability parameter. The higher the value of the model adaptability parameter, the stronger the adaptability of the pre-selected dispatch model to the state changes of the power system.

[0129] For example, the server can input real-time running data into the target pre-selected scheduling model to determine the parameters corresponding to the accuracy, tracking, robustness, and response speed of the target pre-selected scheduling model in the running state, and multiply the parameters of accuracy, tracking, robustness, and response speed by their respective corresponding index weights, and then add all the products to determine the model adaptability parameters.

[0130] In this embodiment, the adaptability index of the pre-selected scheduling model is determined based on real-time operation data, and the adaptability parameters are calculated by combining the index weights, which is beneficial for accurately evaluating the adaptability parameters of the model in the running state.

[0131] In some embodiments, step 208 includes:

[0132] Based on the matching degree between the operating status and the target pre-selected scheduling model and the real-time operating data, determine the adaptation parameters between the target pre-selected scheduling model and the operating status; if the adaptation parameters meet the model parameter update conditions, iteratively update the model parameters of the target pre-selected scheduling model until the termination conditions are met to obtain the global optimization parameters; based on the global optimization parameters corresponding to at least one pre-selected scheduling model, determine the target scheduling model from at least one pre-selected scheduling model.

[0133] The adaptation parameters refer to the parameters used to describe the degree of adaptation between the selected scheduling model and the operating state under specific operating conditions and real-time operating data. They reflect the adaptability and performance of the selected scheduling model under the current operating conditions. Model parameter update conditions refer to the conditions that must be met before iteratively updating the model parameters; for example, the adaptation parameters have not reached the preset threshold or target value; the model's performance under real-time operating data has not met expectations; the operating state has changed significantly, requiring adjustment of the model parameters to adapt to the new operating conditions, etc. Global optimization parameters refer to the parameters obtained by iteratively updating the model parameters of the selected scheduling model until the termination condition is met. Global optimization parameters enable the selected scheduling model to achieve optimal or near-optimal performance under a given operating state. When the adaptation parameters meet the model parameter update conditions, the model parameters of the selected scheduling model are iteratively updated. After each update, the performance of the selected scheduling model is re-evaluated, and a decision is made whether to continue updating based on the new adaptation parameters. The iterative process continues until the termination condition is met (such as reaching the preset number of iterations, or the adaptation parameters converging).

[0134] For example, taking particle swarm optimization as an example, the server can randomly generate multiple particles based on each pre-selected scheduling model and its corresponding initial model parameters. Each particle represents a combination of a pre-selected scheduling model and its corresponding initial model parameters, and the position and velocity of each particle are initialized. For each particle, the server can calculate the adaptation parameters of each particle based on the fitness function. For example, the server can comprehensively determine the adaptation parameters of the pre-selected scheduling model based on the performance indicators (such as accuracy, precision, F1 score, etc.) and business indicators (such as real-time performance, resource consumption, etc.) of each pre-selected scheduling model. For example, the adaptation parameters can be determined by weighted summation of the performance indicators and business indicators. Then, the server can compare the adaptation parameters with the individual optimal solutions and the global optimal solutions in the particle cluster to determine whether the adaptation parameters meet the model parameter update conditions. For example, if the adaptation parameters are better than the individual optimal solutions, the server can update the particle using the individual optimal solutions. Alternatively, if the adaptation parameters are greater than the global optimal solutions, the server can update the particle using the global optimal solutions and update the particle's position and velocity according to the update formula until the maximum number of iterations is reached or the adaptation parameters converge, thus obtaining the global optimal solution at this point. Finally, the server can determine the corresponding pre-selected scheduling model and its corresponding model parameter combination as the target scheduling model based on the global optimal solution after the adaptation parameters have converged.

[0135] In an optional embodiment, the update formula for particle swarm optimization is expressed as:

[0136] (7)

[0137] (8)

[0138] in, The first The th particle (i.e., the th) (the first pre-selected scheduling model) in the first The position and speed of each update round; The first The th particle (i.e., the th) (the first pre-selected scheduling model) in the first The position and speed of each update round; Inertial weights; These are the individual optimal solution and the global optimal solution, respectively. These are learning factors for individuals and for the global learning process, respectively. These are individual and global random factors, respectively.

[0139] In this embodiment, for each pre-selected scheduling model, its global optimization parameters are obtained through iterative updates. The global optimization parameters of each pre-selected scheduling model and its corresponding performance (such as model accuracy, stability, adaptability, etc.) are compared. Based on the comparison results, the pre-selected scheduling model with the best performance or that meets specific requirements is selected as the target scheduling model. This model can provide the optimal or near-optimal scheduling strategy under a given operating state, thereby improving the operating efficiency and stability of the power system.

[0140] In some embodiments, the uncertainty-based power system dispatching method further includes:

[0141] The system obtains the scheduling execution results of the power system based on the scheduling strategy, and determines the scheduling difference between the scheduling execution results and the expected scheduling results; if the scheduling difference meets the model update conditions, it determines the operating data of the power system after scheduling execution; it determines the real-time operating data of the power system based on the operating data after scheduling execution, and returns the steps of obtaining the real-time operating data of the power system and determining the operating status of the power system based on the real-time operating data, or returns the steps of determining the matching degree between the operating status and at least one pre-selected scheduling model based on the real-time operating data and at least one pre-selected scheduling model.

[0142] The dispatch execution result refers to the actual operating state and parameter changes of the power system after actual dispatching based on the dispatch strategy. The dispatch execution result may include, but is not limited to, the actual values ​​of parameters such as voltage, current, frequency, and power after dispatching, as well as the operating status of power equipment and load allocation. The dispatch execution result can be obtained in real time through the power system's monitoring system and data acquisition devices. The expected dispatch result refers to the expected operating state and parameter changes of the power system that can be achieved when generating the dispatch strategy, based on factors such as the power system's operating state, load forecast, and generation plan. The expected dispatch result is the target and desired outcome of the dispatch strategy. The dispatch difference refers to the difference between the dispatch execution result and the expected dispatch result, reflecting the deviation and uncertainty of the dispatch strategy during actual execution. The dispatch difference can be determined by calculating the deviation between the dispatch execution result and the expected dispatch result. The dispatch difference may include, but is not limited to, differences in parameter values ​​and differences in state changes. By comparing the dispatch execution result with the expected dispatch result, the execution deviation of the dispatch strategy can be quantified to analyze the execution effect of the dispatch strategy, which is beneficial for providing a basis for subsequent model updates and strategy adjustments.

[0143] Model update conditions refer to the conditions that trigger updates to the power system dispatch model, allowing the server to further update the model or reselect a suitable model to generate a dispatch strategy. Specifically, model update conditions can include dispatch discrepancies exceeding a set threshold, or the power system failing to reach the expected state after implementing the dispatch strategy. When dispatch discrepancies meet the model update conditions, it indicates that the current dispatch strategy or model cannot accurately reflect the actual operating state of the power system and requires updating and adjustment.

[0144] Post-dispatch execution operational data refers to the various operational data collected by the power system after dispatch execution. When the model update conditions are met, the operational data of the power system at this time can be collected as real-time operational data, and the steps of obtaining the real-time operational data of the power system and determining the operational status of the power system based on the real-time operational data can be returned to re-execute the entire scheme and redetermine the pre-selected dispatch model; or the steps of determining the matching degree between the operational status and at least one pre-selected dispatch model based on the real-time operational data and at least one pre-selected dispatch model can be returned to redetermine the matching degree between the pre-selected dispatch model and the operational status and redetermine the target dispatch model.

[0145] In this embodiment, by obtaining the scheduling execution result and determining the scheduling difference with the expected scheduling result, it is possible to determine whether the scheduling strategy is in line with the current operating state of the power system. When the scheduling difference meets the model update conditions, the real-time operating data is determined using the operating data after scheduling execution, so as to redetermine the pre-selected scheduling model or redetermine the matching degree between the pre-selected scheduling model and the operating state. This ensures that a more suitable scheduling model can be selected or the parameters and structure of the existing model can be adjusted to improve the accuracy and effectiveness of the scheduling strategy.

[0146] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0147] Based on the same inventive concept, this application also provides an uncertainty-based power system dispatching device for implementing the uncertainty-based power system dispatching method described above. The solution provided by this device is similar to the implementation described in the above method; therefore, the specific limitations in one or more uncertainty-based power system dispatching device embodiments provided below can be found in the limitations of the uncertainty-based power system dispatching method described above, and will not be repeated here.

[0148] In one exemplary embodiment, such as Figure 4 As shown, an uncertainty-based power system dispatching device is provided, comprising: a state determination module 402, a model screening module 404, a matching degree determination module 406, a model matching module 408, and an operation dispatching module 410, wherein:

[0149] The status determination module 402 is used to acquire real-time operating data of the power system and determine the operating status of the power system based on the real-time operating data.

[0150] Model 404 is used to select at least one pre-selected scheduling model from at least one pre-trained candidate scheduling model based on the state type of the operating state. The candidate scheduling model is trained based on the uncertainty parameters of the power system and historical operating data. The uncertainty parameters are used to characterize the fluctuation of the power system under the corresponding operating state of each state type.

[0151] The matching degree determination module 406 is used to determine the matching degree between the running state and at least one pre-selected scheduling model based on real-time running data and at least one pre-selected scheduling model.

[0152] The model matching module 408 is used to determine the target scheduling model from at least one pre-selected scheduling model based on the matching degree between the running status and at least one pre-selected scheduling model and the real-time running data.

[0153] The operation and scheduling module 410 is used to generate a scheduling strategy for the power system based on real-time operation data and a target scheduling model, and to schedule the power system according to the scheduling strategy.

[0154] In an optional embodiment, the matching degree determination module 406 is further configured to, for each pre-selected scheduling model, determine the model performance parameters of the pre-selected scheduling model based on real-time running data; the model performance parameters are used to characterize the degree of matching between the pre-selected scheduling model and the corresponding state type; determine the weight coefficient of the model performance parameters based on the relative importance of the model performance parameters when the pre-selected scheduling model performs scheduling strategy generation; and determine the matching degree between the pre-selected scheduling model and the running state based on the model performance parameters and the corresponding weight coefficients.

[0155] In an optional embodiment, the model performance parameters include at least one of the computational complexity, model accuracy, model stability parameters, and model fitness parameters for the pre-selected scheduling model. The matching degree determination module 406 is further configured to determine at least one of the computational complexity, model accuracy, model stability parameters, and model fitness parameters for the pre-selected scheduling model based on real-time running data.

[0156] In an optional embodiment, the matching degree determination module 406 is further configured to determine the time complexity and space complexity of the target pre-selected scheduling model in terms of running time and storage space, respectively, based on the data scale of the real-time running data and the model structure of the target pre-selected scheduling model, and determine the computational complexity based on the time complexity and space complexity; or determine the running prediction data of the target pre-selected scheduling model at the running time of the real-time running data, and determine the model accuracy based on the difference between the real-time running data and the running prediction data; or determine the pre-selected strategies of the power system under different fluctuation conditions based on the real-time running data of the target pre-selected scheduling model, and determine the model stability parameters based on the fluctuation range of each pre-selected strategy; or determine the adaptability index of the target pre-selected scheduling model in the running state based on the real-time running data, and determine the model adaptability parameters based on the adaptability index and the index weights corresponding to the adaptability index.

[0157] In an optional embodiment, the model matching module 408 is further configured to determine the adaptation parameters between the target pre-selected scheduling model and the running state based on the matching degree between the running state and the target pre-selected scheduling model and real-time running data; if the adaptation parameters meet the model parameter update conditions, the model parameters of the target pre-selected scheduling model are iteratively updated until the termination conditions are met to obtain global optimization parameters; and the target scheduling model is determined from at least one pre-selected scheduling model based on the global optimization parameters corresponding to each of the at least one pre-selected scheduling model.

[0158] In an optional embodiment, the uncertainty-based power system dispatching device further includes a model update module, used to obtain the dispatching execution result of dispatching the power system based on the dispatching strategy, determine the dispatching difference between the dispatching execution result and the expected dispatching result; if the dispatching difference meets the model update condition, determine the operating data of the power system after dispatching execution; determine the real-time operating data of the power system based on the operating data after dispatching execution, and return the steps of obtaining the real-time operating data of the power system and determining the operating status of the power system based on the real-time operating data, or return the steps of determining the matching degree between the operating status and at least one pre-selected dispatching model based on the real-time operating data and at least one pre-selected dispatching model.

[0159] The modules in the aforementioned uncertainty-based power system dispatching device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can invoke and execute the corresponding operations of each module.

[0160] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 5As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operating system and computer programs stored in the non-volatile storage media. The database stores power system operating data, operating status, scheduling models, weighting coefficients, and other data. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements an uncertainty-based power system scheduling method.

[0161] Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0162] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the uncertainty-based power system dispatching method in the various embodiments above.

[0163] In one exemplary embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the uncertainty-based power system dispatching methods described in the various embodiments above.

[0164] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the uncertainty-based power system dispatching methods in the various embodiments described above.

[0165] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0166] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0167] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0168] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A power system dispatching method based on uncertainty, characterized in that, The method includes: Acquire real-time operating data of the power system and determine the operating status of the power system based on the real-time operating data; Based on the state type of the operating state, at least one pre-selected scheduling model is determined from at least one pre-trained candidate scheduling model; the candidate scheduling model is trained based on the uncertainty parameters of the power system and historical operating data, and the uncertainty parameters are used to characterize the fluctuation of the power system under the corresponding operating state of each state type; Based on the real-time operation data and the at least one pre-selected scheduling model, determine the matching degree between the operation state and each of the at least one pre-selected scheduling model; Based on the matching degree between the operating status and each of the at least one pre-selected scheduling models and the real-time operating data, a target scheduling model is determined from the at least one pre-selected scheduling model; Based on the real-time operating data, a scheduling strategy for the power system is generated through the target scheduling model, and the power system is scheduled according to the scheduling strategy. The method further includes: Obtain the scheduling execution result of the power system based on the scheduling strategy, and determine the scheduling difference between the scheduling execution result and the expected scheduling result; If the scheduling differences satisfy the model update conditions, the operating data of the power system after the scheduling is executed are determined. The real-time operating data of the power system is determined based on the operating data after the scheduling execution, and the steps of obtaining the real-time operating data of the power system and determining the operating status of the power system based on the real-time operating data are returned, or the steps of determining the matching degree between the operating status and the at least one pre-selected scheduling model based on the real-time operating data and the at least one pre-selected scheduling model are returned.

2. The method according to claim 1, characterized in that, The step of determining the matching degree between the operating state and each of the at least one pre-selected scheduling model based on the real-time operating data and the at least one pre-selected scheduling model includes: For each of the pre-selected scheduling models, the model performance parameters of the pre-selected scheduling model are determined based on the real-time running data; the model performance parameters are used to characterize the degree of matching between the pre-selected scheduling model and the corresponding state type. The weighting coefficients of the model performance parameters are determined based on their relative importance when generating scheduling strategies for the pre-selected scheduling model. The matching degree between the pre-selected scheduling model and the running state is determined based on the model performance parameters and the corresponding weight coefficients.

3. The method according to claim 2, characterized in that, The model performance parameters include at least one of the following: computational complexity, model accuracy, model stability parameters, and model adaptability parameters of the pre-selected scheduling model. The step of determining the model performance parameters for the pre-selected scheduling model based on the real-time operational data includes: Based on the real-time operational data, determine at least one of the following parameters for the pre-selected scheduling model: computational complexity, model accuracy, model stability, and model adaptability.

4. The method according to claim 3, characterized in that, The step of determining at least one of the following parameters based on the real-time operational data: computational complexity, model accuracy, model stability parameter, and model adaptability parameter for the pre-selected scheduling model, includes at least one of the following: Based on the data scale of the real-time running data and the model structure of the pre-selected scheduling model, the time complexity in terms of running time and the space complexity in terms of storage space of the pre-selected scheduling model are determined respectively, and the computational complexity is determined based on the time complexity and the space complexity. Determine the predicted running data of the pre-selected scheduling model at the running time of the real-time running data, and determine the model accuracy based on the difference between the real-time running data and the predicted running data; Based on the real-time operating data, the pre-selected strategies for the power system under different fluctuation conditions are determined using the pre-selected scheduling model, and the model stability parameters are determined according to the fluctuation range of each pre-selected strategy. Based on the real-time operation data, the adaptability index of the pre-selected scheduling model under the operation state is determined, and the model adaptability parameters are determined according to the adaptability index and the index weight corresponding to each adaptability index.

5. The method according to claim 2, characterized in that, The step of determining the target scheduling model from the at least one pre-selected scheduling models based on the matching degree between the operating status and each of the at least one pre-selected scheduling models and the real-time operating data includes: Based on the matching degree between the operating status and the target pre-selected scheduling model and the real-time operating data, determine the adaptation parameters between the target pre-selected scheduling model and the operating status; If the adaptation parameters meet the model parameter update conditions, the model parameters of the pre-selected scheduling model are iteratively updated until the termination conditions are met, and the global optimization parameters are obtained. The target scheduling model is determined from the at least one pre-selected scheduling model based on the global optimization parameters corresponding to each of the at least one pre-selected scheduling model.

6. The method according to claim 4, characterized in that, The method further includes: Based on the characteristics of the pre-selection scheduling model, weights are assigned to the time complexity and the space complexity respectively, and the time complexity and the space complexity are weighted and fused to obtain the computational complexity.

7. A power system dispatching device based on uncertainty, characterized in that, The device includes: The status determination module is used to acquire real-time operating data of the power system and determine the operating status of the power system based on the real-time operating data. A model selection model is used to determine at least one pre-selected scheduling model from at least one pre-trained candidate scheduling model based on the state type of the operating state; the candidate scheduling model is trained based on the uncertainty parameters of the power system and historical operating data, and the uncertainty parameters are used to characterize the fluctuation of the power system under the corresponding operating state of each state type; The matching degree determination module is used to determine the matching degree between the running state and the at least one pre-selected scheduling model based on the real-time running data and the at least one pre-selected scheduling model. The model matching module is used to determine the target scheduling model from the at least one pre-selected scheduling model based on the matching degree between the running status and the at least one pre-selected scheduling model and the real-time running data. The operation scheduling module is used to generate a scheduling strategy for the power system based on the real-time operation data and the target scheduling model, and to schedule the power system according to the scheduling strategy. The model update module is used to obtain the scheduling execution result of the power system based on the scheduling strategy, determine the scheduling difference between the scheduling execution result and the expected scheduling result; if the scheduling difference meets the model update condition, determine the operating data of the power system after scheduling execution; determine the real-time operating data of the power system based on the operating data after scheduling execution, and return the step of obtaining the real-time operating data of the power system and determining the operating status of the power system based on the real-time operating data, or return the step of determining the matching degree between the operating status and the at least one pre-selected scheduling model based on the real-time operating data and the at least one pre-selected scheduling model.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to 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, it implements the steps of the method according to any one of claims 1 to 6.

Images