A method and system for power resource allocation

Abstract

The present invention discloses a power resource allocation method and system, which relates to the technical field of power resource allocation. It includes training a prediction model for both the supply and demand sides of the power grid. A prediction model that can consider the time lag of both the supply and demand sides of the power grid and analyze the complex coupling relationship between the supply and demand sides is trained, providing a reliable basis for subsequent power resource allocation strategies and the evaluation of allocation effects. Define the operating condition stability indicators for both the supply and demand sides of the power grid, thereby setting the monitoring period to ensure the timeliness and adaptability of power resource allocation and scheduling. Analyze the allocation effect of the initial power distribution strategy in multiple dimensions, and combine evaluation indicators to adjust the power distribution strategy in the next monitoring period. Evaluate the power distribution strategy comprehensively from multiple dimensions, thereby adjusting the power distribution strategy in the next monitoring period, optimizing the power resource allocation of the power grid system, and ensuring the reliability and accuracy of power resource allocation.

Description

Technical Field

[0001] The present invention relates to the technical field of power resource allocation, and in particular to a power resource allocation method and system. Background Art

[0002] With the continuous expansion of the scale of the power system, the power resource allocation scheme faces many challenges. Traditional power distribution mainly relies on the planned distribution method, which is difficult to adapt to the increasingly complex power demand and diverse power generation resources. At present, the problem of the peak-valley difference of the power grid is prominent, the access of renewable energy brings uncertainty, the distribution of power resources between regions is uneven, and the power market mechanism is not yet perfect. To solve these problems, measures such as adopting an intelligent power distribution system, optimizing the power dispatching mechanism, and strengthening cross-regional power allocation are required. By real-time monitoring of power supply and demand and flexibly adjusting the distribution strategy, it is ensured that the power resources are reasonably allocated, the operation efficiency and reliability of the power system are improved, and the power demand for economic and social development is met.

[0003] In the prior art, the strategy of power resource allocation is often relatively fixed, responsible for power distribution for a long period of time, and cannot adapt to the changes on both sides of the power grid supply and demand that are complex and changeable, resulting in poor timeliness, accuracy, and adaptability of power resource allocation, and unable to ensure the normal operation requirements on both sides of the power grid supply and demand.

[0004] Therefore, how to improve the timeliness, accuracy, and adaptability of power resource allocation is a technical problem to be solved at present. Summary of the Invention

[0005] The purpose of the present invention is to solve the problem of poor timeliness, accuracy, and adaptability of power resource allocation in the prior art, and a power resource allocation method is proposed, which includes,

[0006] Collect historical data on both sides of the power grid supply and demand, train a prediction model for both sides of the power grid supply and demand, and the prediction model for both sides of the power grid supply and demand is used to predict the supply and demand information on both sides of the power grid supply and demand in the future for a period of time;

[0007] Define the working condition stability index for both sides of the power grid supply and demand, and set the monitoring period through the working condition stability index for both sides of the power grid supply and demand;

[0008] Based on the monitoring period, collect relevant data on both sides of the power grid supply and demand in real time, and combine with the predicted supply and demand information on both sides of the power grid supply and demand to perform an optimization algorithm to generate an initial power distribution strategy, and set a threshold according to the predicted supply and demand information on both sides of the power grid supply and demand;

[0009] Allocate the power resources on the power grid according to the initial power distribution strategy, analyze the distribution effect of the initial power distribution strategy in multiple dimensions to obtain evaluation indicators, and adjust the power distribution strategy in the next monitoring period in combination with the evaluation indicators, so as to optimize the power resource allocation of the power grid system.

[0010] In some embodiments of the present application, training the power grid supply-demand prediction model includes:

[0011] The supply and demand sides of the power grid include the power supply side and the power demand side of the power grid;

[0012] Draw the time series diagrams of the power generation power on the power supply side of the power grid and the power consumption load on the power demand side of the power grid respectively, calculate the correlation coefficients between the power generation power and the power consumption load under different time lag values, traverse the correlation coefficients under different time lag values, and count the maximum value and the median value of the correlation coefficients. Denote the correlation coefficient interval of the maximum value and the median value as the strong correlation interval, and determine the time lag value of the power supply side - power demand side according to the strong correlation interval;

[0013] Establish the respective data time axes of the power supply side and the power demand side for the historical data of the power supply side and the power demand side of the power grid according to the timestamps, align the respective data time axes of the power supply side and the power demand side according to the time lag value of the power supply side - power demand side, determine the time window based on the periodicity of the historical data on both sides and the prediction time, intercept the respective data time axes of the power supply side and the power demand side according to the time window, determine the corresponding data axis of the power supply side - power demand side, and divide the training set and the test set on the basis of the corresponding data axis of the power supply side - power demand side;

[0014] Evaluate the contribution of the data on the power supply side and the power demand side of the power grid to the prediction results, so as to determine the respective proportions of the data on both sides in the training set and the test set;

[0015] Train and optimize the power grid supply-demand prediction model based on the divided training set and test set.

[0016] In some embodiments of the present application, define the operating stability indicators of the power grid supply and demand sides, including:

[0017] Collect the operating stability parameters of the power supply side and the power demand side of the power grid respectively, classify the operating stability parameters of the power supply side and the power demand side of the power grid respectively, determine the representative interval and the coefficient of variation of the operating stability parameters under each category, evaluate the stability of the representative interval of the operating stability parameters under each category, and obtain the evaluation indicators;

[0018] Calculate the respective operating stability indicators of the power supply side and the power demand side of the power grid according to the evaluation indicators and the coefficients of variation of the respective operating stability parameters of the power supply side and the power demand side of the power grid.

[0019] In some embodiments of the present application, the monitoring period is set based on the operating condition stability indexes on both the supply and demand sides of the power grid, including

[0020] Screen parameters in the power grid data that can describe the supply-demand relationship on both the supply and demand sides of the power grid, statistically calculate the representative intervals and coefficient of variation of the parameters of each supply-demand relationship, and generate a supply-demand stability index based on the representative intervals and coefficient of variation of the parameters of each supply-demand relationship;

[0021] Combine the operating condition stability indexes on both the supply and demand sides of the power grid and the supply-demand stability index of the power grid to determine the comprehensive stability level of a power grid, and map a monitoring period based on the comprehensive stability level of the power grid.

[0022] In some embodiments of the present application, an optimization algorithm is combined with the predicted supply-demand information on both the supply and demand sides of the power grid to generate an initial power distribution strategy, including

[0023] The optimization algorithm includes linear programming, quadratic programming, dynamic programming, genetic algorithm, and particle swarm algorithm, and the optimization objective and power grid operation constraint conditions are obtained;

[0024] Based on the predicted supply-demand information on both the supply and demand sides of the power grid, the optimization objective, and the power grid operation constraint conditions, apply the optimization algorithm to find the optimal initial power distribution strategy.

[0025] In some embodiments of the present application, thresholds are set according to the predicted supply-demand information on both the supply and demand sides of the power grid, including

[0026] Set deviation thresholds in multiple dimensions according to the predicted supply-demand information on both the supply and demand sides of the power grid. The deviation threshold is the degree of deviation between the actual value and the predicted value.

[0027] In some embodiments of the present application, analyze the distribution effect of the initial power distribution strategy in multiple dimensions to obtain evaluation indexes, including

[0028] The multiple dimensions include the model prediction dimension, power supply reliability dimension, energy utilization efficiency dimension, and electricity cost dimension. Analyze the distribution effect in the model prediction dimension, power supply reliability dimension, energy utilization efficiency dimension, and electricity cost dimension based on the deviation threshold to obtain multiple evaluation indexes, and each evaluation index corresponds to one dimension.

[0029] In some embodiments of the present application, combine the evaluation indexes to adjust the power distribution strategy in the next monitoring period, including

[0030] Integrate the evaluation indexes under all dimensions to obtain a distribution evaluation index;

[0031] When the distribution evaluation index is higher than the preset distribution evaluation index, do not adjust the power distribution strategy in the next monitoring period;

[0032] Otherwise, adjust the power distribution strategy in the next monitoring period according to the evaluation indicators in each dimension.

[0033] Correspondingly, the present application also provides a power resource allocation system, including

[0034] The first module is used to collect historical data on both sides of the power grid supply and demand, train a prediction model for both sides of the power grid supply and demand, and the prediction model for both sides of the power grid supply and demand is used to predict the supply and demand information on both sides of the power grid in a future period of time;

[0035] The second module is used to define the operating stability indicators for both sides of the power grid supply and demand, and set the monitoring period based on the operating stability indicators for both sides of the power grid supply and demand;

[0036] The third module is used to collect relevant data on both sides of the power grid supply and demand in real time on the basis of the monitoring period, and perform an optimization algorithm in combination with the predicted supply and demand information on both sides of the power grid supply and demand to generate an initial power distribution strategy, and set a threshold according to the predicted supply and demand information on both sides of the power grid supply and demand;

[0037] The fourth module is used to allocate power resources on the power grid according to the initial power distribution strategy, analyze the distribution effect of the initial power distribution strategy in multiple dimensions to obtain evaluation indicators, and adjust the power distribution strategy in the next monitoring period in combination with the evaluation indicators, so as to optimize the power resource allocation of the power grid system.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0039] 1. Train a prediction model for both sides of the power grid supply and demand, and train a prediction model that can consider the time lag of both sides of the power grid supply and demand and analyze the complex coupling relationship between the supply and demand of both sides, providing a reliable basis for subsequent power resource allocation strategies and distribution effect evaluations. Define the operating stability indicators for both sides of the power grid supply and demand, and determine the operating stability indicators by considering the respective operating stability of both sides of the power grid supply and demand and the overall supply and demand stability, so as to set the monitoring period and ensure the timeliness and adaptability of power resource allocation and scheduling.

[0040] 2. Analyze the distribution effect of the initial power distribution strategy in multiple dimensions, and adjust the power distribution strategy in the next monitoring period in combination with the evaluation indicators, comprehensively evaluate the power distribution strategy from multiple dimensions, so as to adjust the power distribution strategy in the next monitoring period, optimize the power resource allocation of the power grid system, and ensure the reliability and accuracy of power resource allocation. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic flow chart of a power resource allocation method proposed by the present invention;

[0042] Figure 2Schematic diagram of a power resource allocation system proposed by the present invention. Detailed implementation manners

[0043] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0044] Refer to Figure 1 , a power resource allocation method, including the following steps:

[0045] Step S101, collect historical data on both the supply and demand sides of the power grid, and train a prediction model for both the supply and demand sides of the power grid. The prediction model for both the supply and demand sides of the power grid is used to predict the supply and demand information on both the supply and demand sides of the power grid for a period of time in the future.

[0046] In this embodiment, in order to ensure the timeliness and advance of power resource allocation, it is necessary to predict the supply and demand information on both the supply and demand sides of the power grid in advance. Supply side: Collect historical power generation data of different types of power generation equipment (such as thermal power, hydropower, wind power, photovoltaic, etc.), including power generation power, equipment operation status, fuel consumption, meteorological conditions (such as wind speed, light intensity, water level), etc. At the same time, collect the transmission data of the power grid, such as the load and loss of transmission lines. Demand side: Collect historical electricity consumption data of different industries (industry, commerce, residents), covering factors such as electricity consumption load, user electricity consumption behavior (such as electricity consumption habits during peak and valley periods), economic development level, holidays, etc. Clean the collected data, and process missing values and outliers. For example, for missing power generation power data, interpolation method can be used for filling; for abnormal electricity consumption load data, it can be identified and corrected through statistical analysis methods. Perform data normalization or standardization processing to make data with different features comparable. For example, normalize the power generation power and electricity consumption load data to the interval [0, 1].

[0047] In some embodiments of the present application, training the prediction model for both the supply and demand sides of the power grid includes,

[0048] Both the supply and demand sides of the power grid include the supply side and the demand side of the power grid;

[0049] Draw the time series diagrams of the power generation power on the supply side of the power grid and the electricity consumption load on the demand side of the power grid respectively, calculate the correlation coefficients between the power generation power and the electricity consumption load under different time lag values, traverse the correlation coefficients under different time lag values, and count the maximum value and the median value of the correlation coefficients. Denote the correlation coefficient interval of the maximum value and the median value as the strong correlation interval, and determine the time lag value of the supply side - demand side according to the strong correlation interval;

[0050] Establish the respective data time axes of the power supply side and the demand side of the power grid according to the time stamps of the historical data on both sides of the power supply side and the demand side of the power grid. Align the respective data time axes of the power supply side and the demand side according to the time lag value of the power supply side - demand side. Determine the time window based on the periodicity of the historical data on both sides and the prediction time. Intercept the respective data time axes of the power supply side and the demand side according to the time window to determine the corresponding data axes of the power supply side - demand side. Divide the training set and the test set based on the corresponding data axes of the power supply side - demand side;

[0051] Evaluate the contribution of the data on both the power supply side and the demand side of the power grid to the prediction results, so as to determine the respective proportions of the data on both sides in the training set and the test set;

[0052] Train and optimize the prediction models for both the power supply and demand sides of the power grid based on the divided training set and test set.

[0053] In this embodiment, the power supply side of the power grid describes the situation of the power supply side with the power generation power, and the demand side of the power grid describes the situation of the power consumption side, that is, the demand side, with the power consumption load. There is usually a time lag between the power supply side data and the demand side data. From a causal relationship perspective, the change in the power consumption demand on the demand side will prompt the power supply side to adjust the power generation plan. For example, when the power consumption of industrial users suddenly increases during the daytime working hours, the power supply side may need a period of time (such as the dispatching response time, the generator set startup time, etc.) to increase the power generation to meet the demand. Conversely, the change in the power generation capacity on the power supply side will also affect the power consumption situation on the demand side, but this impact is relatively indirect. For example, when the power generation on the power supply side decreases due to equipment failures, the demand side may respond by power rationing, adjusting the power consumption plan, etc., and there is also a certain time lag in the middle. Considering the time lag of both the supply and demand sides to divide the training set and the test set can make the model prediction effect better.

[0054] In this embodiment, the time series diagrams of the power generation power on the power supply side and the power consumption load on the demand side are respectively drawn, and the correlation coefficients between the power generation power on the power supply side and the power consumption load on the demand side at different time lags are calculated. The correlation coefficient interval of the maximum value and the median value is denoted as the strong correlation interval, and this range is used as the strong correlation interval with relatively strong correlation. Within this strong correlation interval, the time lag situation between the supply and demand sides is determined through the time lag value corresponding to the average value. The approximate time lag duration can be determined by analyzing the historical data. For example, through analysis, it is found that after the change in the power consumption load on the demand side, the power generation power on the power supply side starts to have an obvious adjustment on average after 1 hour. Then, the demand side data can be moved forward by 1 hour to be preliminarily aligned with the power supply side data. Due to the time lag, it is necessary to align the data of the power supply side and the demand side in time.

[0055] In this embodiment, a time window is determined based on the periodicity of historical data on both sides and the prediction time. Considering the periodicity of each type of data and the prediction time (the duration to be predicted by this model), the size of the time window is jointly determined. If short-term supply-demand balance is predicted, such as the supply-demand situation in the next 1 hour, a smaller time window can be selected, for example, dividing the data at 1-hour intervals; if long-term supply-demand trends are predicted, such as the supply-demand situation in the next week, a larger time window can be selected, for example, dividing the data at 1-day intervals, and in this way, the training set and the test set are divided. Evaluate the contribution of data on both the power supply side and the demand side of the power grid to the prediction result, and analyze the importance of power supply side and demand side data to the model prediction. The contribution degree of each feature (power supply side and demand side data) to the prediction result can be determined through a feature selection algorithm (such as the feature importance evaluation of a random forest). If the demand side data has a greater impact on the prediction result, the proportion of demand side data in the training set can be appropriately increased; conversely, if the power supply side data is more important, the proportion of power supply side data is increased.

[0056] Step S102, define the working condition stability indicators on both the supply and demand sides of the power grid, and set the monitoring period through the working condition stability indicators on both the supply and demand sides of the power grid.

[0057] In this embodiment, the parameters that can describe the stable situation on both the supply and demand sides of the power grid are screened out, integrated separately, and the working condition stability indicators on both the supply and demand sides of the power grid are obtained. Then, in combination with the overall supply-demand relationship parameters of the power grid, the monitoring period is set.

[0058] It can be understood that the stable situation of the power grid is jointly reflected from three aspects: the working condition stability indicators on both the supply and demand sides of the power grid and the overall supply-demand relationship parameters of the power grid, so as to set the monitoring period, which is convenient for timely adjustment of subsequent power distribution strategies. This monitoring period can be dynamically adjusted according to the actual situation of the power grid.

[0059] In some embodiments of the present application, the working condition stability indicators on both the supply and demand sides of the power grid are defined, including

[0060] Separate the working condition stability parameters on the power supply side and the demand side of the power grid are collected respectively, the working condition stability parameters on the power supply side and the demand side of the power grid are classified respectively, the representative interval and coefficient of variation of the working condition stability parameters under each category are determined, and the stability of the representative interval of the working condition stability parameters under each category is evaluated to obtain the evaluation indicators;

[0061] According to the evaluation indicators and coefficients of variation of the working condition stability parameters on the power supply side and the demand side of the power grid respectively, the working condition stability indicators on the power supply side and the demand side of the power grid are calculated respectively.

[0062] In this embodiment, the operating condition stability parameters on the power supply side of the power grid include equipment operation, power generation capacity, and reserve capacity. For equipment operation, the failure rate of power generation equipment reflects the reliability of power generation equipment. The lower the failure rate, the higher the power supply stability. The mean time between failures (MTBF) of power generation equipment is the average running time between two failures. The longer the time, the more stable the equipment. For power generation capacity, the fluctuation range of power generation capacity reflects the change range of power generation power within a certain period. The smaller the fluctuation range, the more stable the power generation. The power generation power. For reserve capacity, the reserve capacity ratio measures the guarantee degree of reserve power generation capacity for the total power generation capacity. The higher the ratio, the stronger the ability to handle emergencies. The reserve capacity response time is the time from the need to activate reserve capacity to actual operation. The shorter the time, the more timely the power supply stability can be guaranteed. The operating condition stability parameters on the demand side include power consumption load and user behavior.

[0063] In this embodiment, the representative interval of the operating condition stability parameter is the interval value with a higher occurrence frequency. By analyzing the interval value, the evaluation index of the operating condition stability parameter is obtained. Combining the coefficient of variation, the operating condition stability indicators of the power supply side and demand side of the power grid are calculated respectively (The stability indicators of the operating conditions on the supplyside or demand side of the power grid). The specific formula is as follows:

[0064] ;

[0065] Among them, is the operating condition stability indicator of the power supply side or demand side of the power grid, is the number of operating condition stability parameters of the power supply side or demand side of the power grid, is the combined weight of the th operating condition stability parameter of the power supply side or demand side of the power grid, is the evaluation index of the th operating condition stability parameter of the power supply side or demand side of the power grid, is the coefficient of variation of the th operating condition stability parameter of the power supply side or demand side of the power grid. The coefficient of variation describes the stability of this type of operating condition stability parameter. is the first constant of the th operating condition stability parameter of the power supply side or demand side of the power grid, represents the correction of the evaluation index by the coefficient of variation. After summing and then taking the class average (slightly smaller than the average value, which can better reflect the operating condition stability), as the operating condition stability indicator.

[0066] In some embodiments of the present application, the monitoring period is set based on the operating condition stability indicators on both the supply and demand sides of the power grid, including:

[0067] Screen parameters in the power grid data that can describe the supply-demand relationship on both the supply and demand sides of the power grid, statistically calculate the representative intervals and coefficient of variation of the parameters of each supply-demand relationship, and generate supply-demand stability indicators based on the representative intervals and coefficient of variation of the parameters of each supply-demand relationship;

[0068] Determine the comprehensive stability level of a power grid by combining the operating condition stability indicators on both the supply and demand sides of the power grid and the supply-demand stability indicators of the power grid, and map a monitoring period based on the comprehensive stability level of the power grid.

[0069] In this embodiment, the supply-demand coupling relationship clearly reveals the internal connection between the power supply side and the demand side. Changes in the power generation capacity on the power supply side will directly affect the degree of electricity consumption satisfaction on the demand side, while fluctuations in the electricity load on the demand side will also affect the power generation plan and equipment operation on the power supply side. When combining the operating condition stability indicators on both sides, this kind of association needs to be considered, and no single indicator can be viewed in isolation. This mutual influence relationship provides a logical basis for the combination of the operating condition stability indicators on both sides. Based on the supply-demand coupling relationship, key indicators that truly reflect the coupling situation on both sides can be screened out, so as to be used for the combination of the operating condition stability indicators. When constructing the coupling index system, indicators of supply-demand relationship parameters such as the supply-demand balance index (the ratio of the difference between the actual power generation power and the electricity load to the electricity load), the adaptability of the reserve capacity response speed to load fluctuations, etc. are all direct manifestations of the supply-demand coupling relationship. Generate supply-demand stability indicators based on the representative intervals and coefficient of variation of the parameters of each supply-demand relationship. The calculation process is the same as that of the operating condition stability indicators on both sides above, and will not be elaborated here. Determine the comprehensive stability level of a power grid by combining the operating condition stability indicators on both the supply and demand sides of the power grid and the supply-demand stability indicators of the power grid. The specific calculation formula is as follows:

[0070] ;

[0071] Wherein, is the comprehensive stability level of the power grid, , , are the combined weights of the power supply side, the demand side, and the supply-demand relationship of the power grid respectively, , , are the operating condition stability indicators of the power supply side, the demand side, and the supply-demand relationship of the power grid respectively, represents the minimum value among the three, , are both preset constants, It represents the correction of the minimum operating condition stability index to the sum of the operating condition stability indices of the three. The two constants are used to balance the magnitude of the correction function and the magnitude of the stability level. [] is the rounding symbol.

[0072] Step S103: Based on the monitoring period, collect relevant data on both the supply and demand sides of the power grid in real time, and combine the predicted supply and demand information on both the supply and demand sides of the power grid to perform an optimization algorithm to generate an initial power distribution strategy, and set thresholds according to the predicted supply and demand information on both the supply and demand sides of the power grid.

[0073] In this embodiment, a prediction model is used to predict the supply and demand information on both the supply and demand sides of the power grid. Methods such as time series analysis, regression analysis, and neural network models are used to predict the information on both the supply and demand sides of the power grid. According to the specific objectives and constraints of power distribution, a suitable optimization algorithm is selected. Commonly used optimization algorithms include linear programming, quadratic programming, dynamic programming, genetic algorithms, particle swarm algorithms, etc.

[0074] In some embodiments of the present application, an optimization algorithm is combined with the predicted supply and demand information on both the supply and demand sides of the power grid to generate an initial power distribution strategy, including

[0075] The optimization algorithm includes linear programming, quadratic programming, dynamic programming, genetic algorithms, and particle swarm algorithms, and obtains the optimization objective and the power grid operation constraint conditions;

[0076] Based on the predicted supply and demand information on both the supply and demand sides of the power grid, the optimization objective, and the power grid operation constraint conditions, the optimization algorithm is applied to find the optimal initial power distribution strategy.

[0077] In this embodiment, the optimization objective can be to minimize the power generation cost, maximize the energy utilization rate, minimize the transmission loss, etc. under the premise of meeting the supply and demand balance. Other objectives such as environmental friendliness and power grid security can also be considered. The supply and demand prediction information, the power grid operation constraint conditions (such as the output limit of power generation equipment, the capacity limit of transmission lines, etc.), and the optimization objective are input into the optimization algorithm. The optimization algorithm performs iterative calculations to find the optimal power distribution strategy that satisfies all constraint conditions. According to the calculation results of the optimization algorithm, an initial power distribution strategy is generated. The power distribution strategy includes the output plan of each power generation equipment, the power flow distribution of the transmission line, the power consumption plan of users, etc.

[0078] In some embodiments of the present application, setting thresholds according to the predicted supply and demand information on both the supply and demand sides of the power grid includes

[0079] Deviation thresholds in multiple dimensions are set according to the predicted supply and demand information on both the supply and demand sides of the power grid. The deviation threshold is the degree of deviation between the actual value and the predicted value.

[0080] In this embodiment, the predicted supply and demand information on both sides of the power grid contains rich data, such as power generation capacity prediction, electricity load prediction, supply-demand balance prediction, etc. These data can reflect the operating status of the power grid at different times and under different conditions, providing a solid foundation for setting multi-dimensional evaluation thresholds. By analyzing and mining these data, the key factors affecting the power distribution effect can be identified, and then reasonable deviation thresholds can be set, reasonable deviations can be set for each evaluation dimension, and the rationality of the subsequent distribution strategy can be evaluated.

[0081] Step S104, perform power resource allocation on the power grid according to the initial power distribution strategy, analyze the distribution effect of the initial power distribution strategy in multiple dimensions to obtain evaluation indicators, and adjust the power distribution strategy in the next monitoring period in combination with the evaluation indicators, so as to optimize the power resource allocation of the power grid system.

[0082] In this embodiment, the process of monitoring the initial power distribution strategy is analyzed in multiple dimensions to determine the evaluation indicators for each dimension. A comprehensive indicator is determined in combination with the evaluation indicators to judge whether it is necessary to adjust the power distribution strategy.

[0083] In some embodiments of the present application, the distribution effect of the initial power distribution strategy is analyzed in multiple dimensions to obtain evaluation indicators, including

[0084] The multiple dimensions include the model prediction dimension, the power supply reliability dimension, the energy utilization rate dimension, and the electricity cost dimension. Based on the deviation threshold, analyze the distribution effect in the model prediction dimension, the power supply reliability dimension, the energy utilization rate dimension, and the electricity cost dimension to obtain multiple evaluation indicators, and each evaluation indicator corresponds to a dimension.

[0085] In this embodiment, the model prediction dimension includes the situations of model prediction such as electricity load and power generation capacity, the power supply reliability dimension includes the number of power outages, the power outage duration, the planned power outage duration, the unplanned power outage duration, etc., the energy utilization rate dimension includes the input and output energy of equipment, the transmission line loss, the actual electricity consumption of users, etc., and the electricity cost dimension includes the power generation cost, the transmission cost, the electricity cost, the peak-valley electricity price, etc.

[0086] In some embodiments of the present application, adjusting the power distribution strategy in the next monitoring period in combination with the evaluation indicators includes

[0087] Integrate the evaluation indicators under all dimensions to obtain the distribution evaluation indicator;

[0088] When the distribution evaluation indicator is higher than the preset distribution evaluation indicator, do not adjust the power distribution strategy in the next monitoring period;

[0089] Otherwise, adjust the power distribution strategy in the next monitoring period according to the evaluation indicators under each dimension.

[0090] In this embodiment, the evaluation indexes under all dimensions are integrated to obtain the distribution evaluation index, and the specific formula is as follows:

[0091] ;

[0092] Wherein, is the distribution evaluation index, , , , are the influence weights of the model prediction dimension, power supply reliability dimension, energy utilization rate dimension and power consumption cost dimension respectively, , , , are the evaluation indexes of the model prediction dimension, power supply reliability dimension, energy utilization rate dimension and power consumption cost dimension respectively, , are respectively the maximum value and the minimum value in, is a preset constant, represents the correction of the sum of the four dimensions by the average value determined by the maximum value and the minimum value.

[0093] Because there are multiple power distribution optimization objectives and not all dimensions can be taken into account, it is possible that not all dimensions are the optimal solutions. As long as the distribution evaluation index meets the requirements, a slightly worse evaluation index for a single dimension can be accepted without adjusting the distribution strategy. Otherwise, adjustments will be made. Based on the analysis of which dimension is worse, a specific adjustment plan for the power distribution strategy will be formulated. The adjusted power distribution strategy will be implemented in the next monitoring cycle.

[0094] Correspondingly, the present application also provides a power resource distribution system, as Figure 2 shown, including,

[0095] The first module is used to collect historical data on both sides of the power grid supply and demand, train the power grid supply and demand prediction model, and the power grid supply and demand prediction model is used to predict the supply and demand information on both sides of the power grid in the future for a period of time;

[0096] The second module is used to define the working condition stability index on both sides of the power grid supply and demand, and set the monitoring cycle through the working condition stability index on both sides of the power grid supply and demand;

[0097] The third module is used to collect relevant data on both sides of the power grid supply and demand in real time on the basis of the monitoring cycle, and combine the predicted supply and demand information on both sides of the power grid to perform an optimization algorithm to generate an initial power distribution strategy, and set a threshold according to the predicted supply and demand information on both sides of the power grid;

[0098] The fourth module is used to allocate power resources on the power grid according to the initial power distribution strategy, analyze the distribution effect of the initial power distribution strategy in multiple dimensions to obtain evaluation indicators, and adjust the power distribution strategy in the next monitoring period in combination with the evaluation indicators, so as to optimize the power resource allocation of the power grid system.

[0099] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0100] 1. Train the prediction models on both the supply and demand sides of the power grid, and train a prediction model that can consider the time lag on both the supply and demand sides of the power grid and analyze the complex coupling relationship between the supply and demand sides, providing a reliable basis for subsequent power resource allocation strategies and the evaluation of allocation effects. Define the working condition stability indicators on both the supply and demand sides of the power grid, and determine the working condition stability indicators by considering the respective working condition stability of the supply and demand sides of the power grid and the overall stability of the supply and demand, so as to set the monitoring period and ensure the timeliness and adaptability of power resource allocation and scheduling.

[0101] 2. Analyze the distribution effect of the initial power distribution strategy in multiple dimensions, and adjust the power distribution strategy in the next monitoring period in combination with the evaluation indicators. Evaluate the power distribution strategy comprehensively from multiple dimensions, and then adjust the power distribution strategy in the next monitoring period to optimize the power resource allocation of the power grid system, ensuring the reliability and accuracy of power resource allocation.

[0102] Through the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be implemented through hardware or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various implementation scenarios of the present invention.

[0103] Those skilled in the art can understand that the drawings are only schematic diagrams of a preferred implementation scenario, and the modules or processes in the drawings are not necessarily essential for implementing the present invention.

[0104] Those skilled in the art can understand that the modules in the system in the implementation scenario can be distributed in the system of the implementation scenario according to the description of the implementation scenario, or can be correspondingly changed and located in one or more systems different from the present implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple sub-modules.

[0105] The above are only the preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes, shall be covered by the protection scope of the present invention.

Claims

1. A method for allocating power resources, characterized in that, including collecting historical data on both the supply and demand sides of the power grid, training a prediction model for both the supply and demand sides of the power grid, which is used to predict the supply and demand information on both the supply and demand sides of the power grid for a period of time in the future; defining the operating stability indicators for both the supply and demand sides of the power grid, and setting the monitoring period based on the operating stability indicators for both the supply and demand sides of the power grid; collecting relevant data on both the supply and demand sides of the power grid in real time based on the monitoring period, and combining the predicted supply and demand information on both the supply and demand sides of the power grid to perform an optimization algorithm to generate an initial power distribution strategy, and setting a threshold based on the predicted supply and demand information on both the supply and demand sides of the power grid; allocating power resources on the power grid according to the initial power distribution strategy, analyzing the distribution effect of the initial power distribution strategy in multiple dimensions to obtain evaluation indicators, and adjusting the power distribution strategy in the next monitoring period in combination with the evaluation indicators, so as to optimize the power resource allocation of the power grid system; wherein training the prediction model for both the supply and demand sides of the power grid includes both the supply and demand sides of the power grid include the power supply side and the power demand side of the power grid; drawing the time series diagrams of the power generation power on the power supply side of the power grid and the power consumption load on the power demand side of the power grid respectively, calculating the correlation coefficients between the power generation power and the power consumption load under different time lag values, traversing the correlation coefficients under different time lag values, statistically obtaining the maximum value and the median value of the correlation coefficients, recording the correlation coefficient interval of the maximum value and the median value as the strong correlation interval, and determining the time lag value of the power supply side - power demand side according to the strong correlation interval; establishing the data time axes of the power supply side and the power demand side respectively for the historical data on both the power supply side and the power demand side of the power grid according to the time stamps, aligning the data time axes of the power supply side and the power demand side according to the time lag value of the power supply side - power demand side, determining the time window based on the periodicity of the historical data on both sides and the prediction time, intercepting the data time axes of the power supply side and the power demand side according to the time window, determining the corresponding data axes of the power supply side - power demand side, and dividing the training set and the test set based on the corresponding data axes of the power supply side - power demand side; evaluating the contribution of the data on both the power supply side and the power demand side of the power grid to the prediction result, so as to determine the respective proportions of the data on both sides in the training set and the test set; training and optimizing the prediction model for both the supply and demand sides of the power grid based on the divided training set and test set.

2. The power resource allocation method according to claim 1, wherein defining the operating stability indicators for both the supply and demand sides of the power grid includes collecting the operating stability parameters of the power supply side and the demand side of the power grid respectively, classifying the operating stability parameters of the power supply side and the demand side of the power grid respectively, determining the representative intervals and variation coefficients of the operating stability parameters under each category, evaluating the stability of the representative intervals of the operating stability parameters under each category, and obtaining the evaluation indicators; calculating the operating stability indicators of the power supply side and the demand side of the power grid respectively according to the evaluation indicators and variation coefficients of the respective operating stability parameters of the power supply side and the demand side of the power grid.

3. The power resource allocation method according to claim 2, wherein setting the monitoring period based on the operating stability indicators for both the supply and demand sides of the power grid includes Screen out the parameters in the power grid data that can describe the supply-demand relationship on both the supply and demand sides of the power grid, count the representative intervals and coefficient of variation of the parameters of each supply-demand relationship, and generate a supply-demand stability index based on the representative intervals and coefficient of variation of the parameters of each supply-demand relationship; Determine the comprehensive stability level of a power grid by combining the operating condition stability index on both the supply and demand sides of the power grid and the supply-demand stability index of the power grid, and map a monitoring period according to the comprehensive stability level of the power grid.

4. The power resource allocation method according to claim 1, wherein Combine the predicted supply-demand information on both the supply and demand sides of the power grid to perform an optimization algorithm to generate an initial power distribution strategy, including, The optimization algorithm includes linear programming, quadratic programming, dynamic programming, genetic algorithm, and particle swarm algorithm, and obtain the optimization objective and the power grid operation constraint conditions; Apply the optimization algorithm based on the predicted supply-demand information on both the supply and demand sides of the power grid, the optimization objective, and the power grid operation constraint conditions to find the optimal initial power distribution strategy.

5. The power resource allocation method according to claim 1, wherein Set thresholds according to the predicted supply-demand information on both the supply and demand sides of the power grid, including, Set deviation thresholds in multiple dimensions according to the predicted supply-demand information on both the supply and demand sides of the power grid, and the deviation threshold is the degree of deviation between the actual value and the predicted value.

6. The power resource allocation method according to claim 5, wherein Analyze the distribution effect of the initial power distribution strategy in multiple dimensions to obtain evaluation indicators, including, The multiple dimensions include the model prediction dimension, power supply reliability dimension, energy utilization rate dimension, and electricity consumption cost dimension. Analyze the distribution effects in the model prediction dimension, power supply reliability dimension, energy utilization rate dimension, and electricity consumption cost dimension based on the deviation threshold to obtain multiple evaluation indicators, and each evaluation indicator corresponds to a dimension.

7. The power resource allocation method according to claim 1, wherein Adjust the power distribution strategy in the next monitoring period in combination with the evaluation indicators, including, Integrate the evaluation indicators under all dimensions to obtain a distribution evaluation indicator; When the distribution evaluation indicator is higher than the preset distribution evaluation indicator, do not adjust the power distribution strategy in the next monitoring period; Otherwise, adjust the power distribution strategy in the next monitoring period according to the evaluation indicators under each dimension.

8. A power resource allocation system, characterized in that, For implementing the power resource allocation method according to any one of claims 1-7, the system includes, The first module is used to collect the historical data on both the supply and demand sides of the power grid, train the power grid supply-demand prediction model, and the power grid supply-demand prediction model is used to predict the supply-demand information on both the supply and demand sides of the power grid in the future for a period of time; The second module is used to define the operating condition stability index on both the supply and demand sides of the power grid and set the monitoring period through the operating condition stability index on both the supply and demand sides of the power grid; The third module is used to collect the relevant data on both the supply and demand sides of the power grid in real time based on the monitoring period, and combine the predicted supply-demand information on both the supply and demand sides of the power grid to perform an optimization algorithm to generate an initial power distribution strategy, and set thresholds according to the predicted supply-demand information on both the supply and demand sides of the power grid; The fourth module is used to allocate the power resources on the power grid according to the initial power distribution strategy, analyze the distribution effect of the initial power distribution strategy in multiple dimensions to obtain evaluation indicators, and adjust the power distribution strategy in the next monitoring period in combination with the evaluation indicators, so as to optimize the power resource allocation of the power grid system.

Images