Output prediction method, system, equipment and medium for distributed new energy power station

Abstract

The present invention discloses a method and system, equipment and medium for predicting the output of a distributed new energy power station. After setting a fluctuation characteristic value, the method calculates the fluctuation characteristic value based on the output data of each node in multiple historical time sections, and judges whether there is generalized volatility based on the calculation results of the fluctuation characteristic values ​​of the multiple historical time sections. If so, the fluctuation center position and the fluctuation characteristic value of each node in the predicted time section are predicted based on the fluctuation characteristic value of each node in the multiple historical time sections. Finally, based on the predicted fluctuation center position and the fluctuation characteristic value of each node, the actual output value of all nodes in the fluctuation area is predicted. The method combines the spatiotemporal coupling characteristics between distributed new energy power stations and the generalized volatility analysis, compensates for the output value of each node that is increased or decreased due to the influence of volatility at the prediction time, and improves the output prediction accuracy of the distributed new energy power station under the influence of group random factors.

Description

Technical Field

[0001] The present invention relates to the technical field of output prediction of distributed renewable energy power stations, and in particular to an output prediction method and system for distributed renewable energy power stations, electronic equipment, and a computer-readable storage medium. Background Art

[0002] Distributed renewable energy power stations (including distributed photovoltaic and distributed / distributed wind power) are characterized by their vast number and fixed geographical locations (adjacent locations). These distributed renewable energy power stations form a massive network spanning a vast geographical space. Each renewable energy power station has a metering device, and each can be considered a sensor node. Its sensor data represents the output data of the renewable energy power station. Therefore, a distributed renewable energy system spanning a vast geographical space can be considered a giant sensor network.

[0003] Influenced by external factors, the output of new energy power stations exhibits significant randomness. The randomness of the output of distributed new energy power stations can be divided into individual randomness and group randomness. For example, in the photovoltaic sector, a large bird landing on a photovoltaic module represents individual randomness, affecting only the output of that particular photovoltaic station and having no impact on neighboring stations. Conversely, an unpredictable cloud introduces group randomness. The cloud's aggregation, deformation, movement, and dissipation affect a certain number of photovoltaic stations. The impact of this group randomness on the output of photovoltaic nodes can be viewed as a fluctuation, operating across related distributed photovoltaic nodes. This individual randomness is caused by random variations caused by specific events at individual power stations, similar to the independent behavior of particles, and can be defined as generalized particle properties. Group randomness, on the other hand, is caused by random variations caused by large-scale or collective events, similar to the diffusion and propagation of waves, and can be defined as generalized volatility. Individual randomness has a limited impact, affecting only individual renewable energy power plants. Changes in each individual event are relatively independent and not directly correlated with changes in other power plants. These changes occur at the microscopic level, minimally impacting the overall distributed renewable energy system output. Group randomness, on the other hand, affects multiple renewable energy power plants, or even the entire system, exhibiting greater volatility. Changes are correlated across different renewable energy power plants, resulting in a wider impact. These changes occur at the macroscopic level, significantly impacting the overall system output. The characteristic of volatility is that fluctuations propagate far over time. When these fluctuations operate in a sensor network composed of distributed renewable energy power plants, they form strong spatiotemporal coupling. Changes at one node at one moment can propagate to other nodes at the next, resulting in corresponding changes. This spatiotemporal coupling can be a key factor in predicting the output of distributed renewable energy power plants. However, current research has not considered the impact of spatiotemporal coupling on the output prediction of distributed renewable energy power plants, resulting in poor output prediction accuracy. Summary of the Invention

[0004] The present invention provides a method and system for predicting the output of a distributed new energy power station, electronic equipment, and a computer-readable storage medium, which can improve the output prediction accuracy of the distributed new energy power station under the influence of group random factors.

[0005] According to one aspect of the present invention, a method for predicting the output of a distributed new energy power station is provided, comprising the following contents:

[0006] Constructing the geographical topology structure among distributed new energy power stations;

[0007] Set the fluctuation characteristic value;

[0008] Calculate the fluctuation characteristic value based on the output data of each node in multiple time sections, and determine whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections. If generalized volatility exists, obtain the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections;

[0009] Based on the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections, the fluctuation center position and the fluctuation characteristic value of each node in the next time section are predicted;

[0010] Based on the fluctuation center position in the next time section and the fluctuation characteristic value of each node, the actual output value of all nodes in the fluctuation area is predicted.

[0011] Furthermore, the fluctuation characteristic value is the power fluctuation rate, wherein the power fluctuation rate = (baseline output difference - actual output difference) / baseline mean, the baseline output difference is the difference between the baseline value at the next moment and the baseline value at the previous moment, the actual output difference is the difference between the output value at the next moment and the output value at the previous moment, and the baseline mean is the average value of the baseline value at the next moment and the baseline value at the previous moment.

[0012] Furthermore, the process of determining whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections includes the following:

[0013] In a certain time section, if an extreme value region appears in the power fluctuation rate of each node in a certain area of ​​the topological structure, and the power fluctuation rate gradually decreases outside the extreme value region, the extreme value region is regarded as the fluctuation center. In the subsequent time sections, if the distance between the new fluctuation center and the old fluctuation center does not exceed the preset threshold, the power fluctuation rate of the nodes near the fluctuation center migration line is higher than that of other areas, and the change in the power fluctuation rate sequence of a single node is negatively correlated with the change in the distance between the node and the fluctuation center, then it is determined that generalized volatility exists.

[0014] Furthermore, the process of predicting the actual output values ​​of all nodes in the fluctuation area based on the fluctuation center position and the fluctuation characteristic value of each node in the next time section includes the following:

[0015] For each node, a multiple regression model is constructed with the actual output of the node as the variable and the output baseline and power fluctuation rate of the node as the influencing factors. The actual output data, output baseline and power fluctuation rate of the node in the known time section are then substituted into the multiple regression model to calculate the coefficients of the multiple regression model. The power fluctuation rate and output baseline of the node in the next time section that are predicted are then substituted into the multiple regression model to calculate the actual output value of the node in the next time section. The above process is repeated to obtain the actual output values ​​of all nodes in the fluctuation area.

[0016] Furthermore, the fluctuation characteristic values ​​are the node influence ratio and the node correlation ratio, wherein the node influence ratio is the ratio of the total node influence of a certain node to the average of the total node influence of all nodes, the total node influence of each node is the sum of the absolute values ​​of the standardized covariance sequence of the output value of the node and the output values ​​of all other nodes in multiple time sections, and the node correlation ratio is the ratio of the number of nodes in the standardized covariance sequence whose absolute values ​​are greater than the preset correlation threshold to (N-1), where N is the number of nodes.

[0017] Furthermore, the process of determining whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections includes the following:

[0018] The node with the highest node correlation ratio in multiple time sections is taken as the fluctuation center. If the movement of the new and old fluctuation centers meets the continuity requirements, it is determined that generalized volatility exists.

[0019] Furthermore, the process of predicting the fluctuation center position and the fluctuation characteristic value of each node in the next time section based on the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections includes the following:

[0020] Based on the standardized covariance sequence of each node in the most recent multiple time sections, the standardized covariance sequence of each node in the next period of time is predicted, and the fluctuation characteristic value prediction value of each node is calculated according to the predicted standardized covariance sequence, and the new fluctuation center is determined.

[0021] In addition, the present invention also provides an output prediction system for a distributed new energy power station, comprising:

[0022] Topology construction module, used to build the geographical topology between distributed new energy power stations;

[0023] A fluctuation characteristic value setting module, used for setting the fluctuation characteristic value;

[0024] A generalized volatility judgment module is used to calculate the fluctuation characteristic value based on the output data of each node in multiple time sections, and judge whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections. If generalized volatility exists, the fluctuation center position and the fluctuation characteristic value of each node in the multiple time sections are obtained;

[0025] The fluctuation center prediction module is used to predict the fluctuation center position and the fluctuation characteristic value of each node in the next time section based on the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections;

[0026] The node output prediction module is used to predict the actual output values ​​of all nodes in the fluctuation area based on the fluctuation center position and the fluctuation characteristic value of each node in the next time section.

[0027] In addition, the present invention also provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor is configured to execute the steps of the above method by calling the computer program stored in the memory.

[0028] In addition, the present invention also provides a computer-readable storage medium for storing a computer program for output prediction of a distributed new energy power station, wherein the computer program executes the steps of the above-described method when running on a computer.

[0029] The present invention has the following beneficial effects:

[0030] The output prediction method of the distributed new energy power station of the present invention calculates the fluctuation characteristic value based on the output data of each node in multiple historical time sections after setting the fluctuation characteristic value, and judges whether there is generalized volatility based on the calculation results of the fluctuation characteristic values ​​of multiple historical time sections. If so, the fluctuation center position and the fluctuation characteristic value of each node in the predicted time section are predicted based on the fluctuation characteristic value of each node in the multiple historical time sections. Finally, the actual output value of all nodes in the fluctuation area is predicted based on the predicted fluctuation center position and the fluctuation characteristic value of each node. It combines the spatiotemporal coupling characteristics between distributed new energy power stations and the generalized volatility analysis, compensates for the output value of each node that is increased or decreased due to the influence of volatility at the prediction time, thereby improving the output prediction accuracy of the distributed new energy power station under the influence of group random factors.

[0031] In addition, the output prediction system of the distributed new energy power station of the present invention also has the above advantages.

[0032] In addition to the above-described objects, features and advantages, the present invention has other objects, features and advantages. The present invention will be further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are intended to explain the present invention and do not constitute an undue limitation of the present invention. In the accompanying drawings:

[0034] Figure 1 It is a flow chart of the output prediction method of the distributed new energy power station in the first embodiment of the present application.

[0035] Figure 2 It is a schematic diagram of the module structure of an output prediction system of a distributed new energy power station according to another embodiment of the present application. DETAILED DESCRIPTION

[0036] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0037] Example 1

[0038] Reference Figure 1 , Embodiment 1 of the present application provides a method for predicting the output of a distributed new energy power station, including the following contents:

[0039] Step S1: constructing a geographical topology structure among distributed new energy power stations;

[0040] Step S2: setting the fluctuation characteristic value;

[0041] Step S3: Calculating the fluctuation characteristic value based on the output data of each node in multiple time sections, and judging whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections. If generalized volatility exists, obtaining the fluctuation center position and the fluctuation characteristic value of each node in the multiple time sections;

[0042] Step S4: Based on the fluctuation center position and the fluctuation characteristic value of each node in the multiple time sections, predict the fluctuation center position and the fluctuation characteristic value of each node in the next time section;

[0043] Step S5: Based on the fluctuation center position in the next time section and the fluctuation characteristic value of each node, the actual output value of all nodes in the fluctuation area is predicted.

[0044] It can be understood that the output prediction method of the distributed new energy power station in this embodiment, after setting the fluctuation characteristic value, calculates the fluctuation characteristic value based on the output data of each node in multiple historical time sections, and judges whether there is generalized volatility based on the calculation results of the fluctuation characteristic values ​​of multiple historical time sections. If so, the fluctuation center position and the fluctuation characteristic value of each node in the predicted time section are predicted based on the fluctuation characteristic value of each node in multiple historical time sections. Finally, based on the predicted fluctuation center position and the fluctuation characteristic value of each node, the actual output value of all nodes in the fluctuation area is predicted. It combines the spatiotemporal coupling characteristics between distributed new energy power stations and the generalized volatility analysis, and compensates for the output value of each node at the prediction moment that is increased or decreased due to the influence of volatility, thereby improving the output prediction accuracy of the distributed new energy power station under the influence of group random factors.

[0045] It is understood that in step S1, a topological diagram of the new energy power station system is constructed based on the geographical location relationships between the distributed new energy power stations, thereby providing an analytical basis for subsequent generalized volatility analysis and volatility center prediction. In addition, the nodes referred to below are new energy power stations.

[0046] It can be understood that in step S2, optionally, the fluctuation characteristic value is the power fluctuation rate, wherein the power fluctuation rate = (baseline output difference - actual output difference) / baseline mean, the baseline output difference is the difference between the baseline value at the next moment and the baseline value at the previous moment, the actual output difference is the difference between the output value at the next moment and the output value at the previous moment, and the baseline mean is the average value of the baseline value at the next moment and the baseline value at the previous moment.

[0047] Among them, the baseline output of each node is a theoretical value predicted based on the historical output data of the node without considering the influence of volatility and randomness. For example, the historical output data of a certain node within a period of time is collected at a sampling frequency of every 15 minutes, and then combined with the weather forecast of the future time span, the node output within the next time span (for example, 24 hours) is predicted, and the prediction result is the output baseline prediction of the node. Of course, in other embodiments of the present invention, the influence of weather conditions can also be ignored, and the output baseline can be predicted simply based on historical output data. The improvement idea of ​​the present application is to compensate for the influence of local volatility on the basis of the baseline output prediction, and to optimize the output prediction result by utilizing the spatiotemporal coupling characteristics between distributed new energy power stations.

[0048] It can be understood that the power fluctuation rate is the rate of change of the fluctuation influence of the node output power. The higher the value, the greater the node is affected by the volatility. When making output forecasts, the power fluctuation rate at the prediction moment is added to the baseline output forecast value, that is, the output power value of the node at the prediction moment that is increased or decreased due to the fluctuation can be compensated, thereby improving the accuracy of the output forecast of distributed new energy power stations.

[0049] It is understood that in step S3, after performing a baseline output forecast based on the output data of each node in multiple historical time sections, the power fluctuation rate of each node in the multiple historical time sections is calculated, where a time section can be understood as a data collection moment. The process of determining whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections includes the following:

[0050] In a certain time section, if an extreme value region appears in the power fluctuation rate of each node in a certain area of ​​the topological structure, and the power fluctuation rate gradually decreases outside the extreme value region, the extreme value region is regarded as the fluctuation center. In the subsequent time sections, if the distance between the new fluctuation center and the old fluctuation center does not exceed the preset threshold, the power fluctuation rate of the nodes near the fluctuation center migration line is higher than that of other areas, and the change in the power fluctuation rate sequence of a single node is negatively correlated with the change in the distance between the node and the fluctuation center, then it is determined that generalized volatility exists.

[0051] Specifically, within a certain time section, if the power fluctuation rate of each node in a certain area of ​​the topological structure has an extreme point or an extreme area, and the power fluctuation rate gradually decreases outside the extreme area, the extreme area is regarded as the fluctuation center. Among them, a certain area of ​​the topological structure represents the fluctuation influence area, that is, the fluctuation area, which can be divided according to actual needs. The entire distributed new energy power station system can be selected, or a local area of ​​the distributed new energy power station system can be selected. In addition, the extreme point refers to the area where the power fluctuation rate of a node in the fluctuation area is the largest, and the extreme area refers to the area formed by these node clusters when multiple nodes with similar distances are at the maximum level. The maximum level refers to the power fluctuation rate of the node being within the top 5% of the highest. The specific percentage can be set according to actual conditions, for example, set to the highest 3%. In addition, for the extreme area, the center point of the extreme area in the topological structure diagram can be selected as the fluctuation center, the node where the mean power fluctuation rate in the extreme area is located can be selected as the fluctuation center, or any node in the extreme area can be selected as the fluctuation center. The specific selection can be made according to actual needs. In subsequent time sections, due to changes in the power fluctuation rate of the node, new fluctuation centers will gradually appear. If the new fluctuation center is adjacent to the old fluctuation center, that is, the distance between the new fluctuation center and the old fluctuation center does not exceed a preset threshold, which can be set to three times the average distance between adjacent nodes in the entire distributed new energy power station system, and the power fluctuation rate of nodes located near the fluctuation center migration line is higher and higher than that of other areas, and the power fluctuation rate sequence change of a single node is negatively correlated with the change in the distance between the node and the fluctuation center, that is, the farther the node is from the fluctuation center, the smaller its power fluctuation rate, and the closer the node is to the fluctuation center, the larger its power fluctuation rate. If a fluctuation center exists and the above conditions are met, it is determined that the output of the distributed new energy power station system has generalized volatility.

[0052] In addition, in step S3, after determining the presence of generalized volatility, the fluctuation center position and the power fluctuation rate of each node within multiple historical time sections are obtained to facilitate the subsequent prediction of the fluctuation center position and the power fluctuation rate of each node within the predicted time section. In addition, based on the power fluctuation rate of each node within multiple time sections, the change pattern of the fluctuation characteristic value outside the fluctuation center within the fluctuation area can be analyzed, including spatial and temporal patterns. The spatial pattern refers to whether the fluctuation characteristic value and the distance from the node to the fluctuation center are positively or negatively correlated from the fluctuation center to the periphery, and the magnitude of the correlation can be calculated. The temporal pattern refers to whether the change rate of the fluctuation characteristic value of each node gradually increases or decreases over time, thereby indicating whether the influence of the fluctuation area is increasing or decreasing.

[0053] It will be appreciated that in step S4, based on the power fluctuation rates of each node over multiple time slices, a power fluctuation rate sequence for each node can be obtained. Data methods such as linear regression and curve fitting are then used to perform predictions based on the power fluctuation rate sequence for each node, thereby obtaining the power fluctuation rate for each node at the prediction time. The fluctuation center is then determined based on the power fluctuation rate of each node at the prediction time. If the fluctuation center remains unchanged, the prediction time is extended and the prediction is repeated until a new fluctuation center emerges. Alternatively, if extending the prediction time fails to yield a new fluctuation center, this indicates that the volatility has significantly weakened or even disappeared, which is not considered in the present invention.

[0054] It can be understood that in step S5, the process of predicting the actual output values ​​of all nodes in the fluctuation area based on the fluctuation center position and the fluctuation characteristic value of each node in the next time section includes the following:

[0055] For each node, a multiple regression model is constructed with the actual output of the node as the variable and the output baseline and power fluctuation rate of the node as the influencing factors. The actual output data, output baseline and power fluctuation rate of the node in the known time section are then substituted into the multiple regression model to calculate the coefficients of the multiple regression model. The power fluctuation rate and output baseline of the node in the next time section that are predicted are then substituted into the multiple regression model to calculate the actual output value of the node in the next time section. The above process is repeated to obtain the actual output values ​​of all nodes in the fluctuation area.

[0056] Specifically, a prediction model is constructed for each node within the fluctuation region to facilitate output prediction for each node. For each node, a multivariate regression model is constructed using the node's actual output as a variable and the node's output baseline and power fluctuation rate as influencing factors. The actual output value, output baseline, and power fluctuation rate of the node within the historical time section are then substituted into the constructed multivariate regression model for calculation to obtain the coefficients of the multivariate regression model. The power fluctuation rate and output baseline of the node within the predicted time section are then substituted into the multivariate regression model to calculate the actual output value of the node within the predicted time section. Repeating the above process, the actual output values ​​of all nodes within the fluctuation region can be calculated.

[0057] Example 2

[0058] The difference between the second embodiment and the first embodiment lies in the specific contents of steps S2, S3, S4 and S5.

[0059] It can be understood that in step S2, the fluctuation characteristic value can also be a node influence ratio and a node correlation ratio, wherein the node influence ratio is the ratio of the total node influence of a certain node to the average of the total node influence of all nodes, the total node influence of each node is the sum of the absolute values ​​of the standardized covariance sequence of the output value of the node and the output values ​​of all other nodes in multiple time sections, and the node correlation ratio is the ratio of the number of nodes in the standardized covariance sequence whose absolute values ​​are greater than the preset correlation threshold to (N-1), where N is the number of nodes.

[0060] In addition, the standardized covariance sequence refers to the sequence composed of the standardized covariance coefficients of the output value of a node and the output values ​​of other nodes at a series of time sections. The covariance between nodes can be defined as: COV(X, Y) = E(XY)-E(X)E(Y), where X and Y represent the output value sequences of the two nodes, E(X) and E(Y) represent the means or mathematical expectations of sequences X and Y, respectively, and E(XY) represents the mean or mathematical expectation of the product of sequences X and Y. The standardized covariance = covariance / (first standard deviation × second standard deviation), where the first standard deviation and the second standard deviation represent the standard deviations of sequences X and Y, respectively. The standardized covariance between two nodes represents the consistency of the output changes of the two nodes under the influence of fluctuations. Its value ranges from [-1, 1]. The closer the absolute value is to 1, the stronger the correlation between the output changes of the two nodes, and the closer it is to 0, the weaker the correlation. The correlation threshold is used to determine whether there is a correlation between nodes. The absolute value of the standardized covariance between nodes ranges from [0 to 1]. The closer it is to 0, the lower the correlation. Therefore, by setting a correlation threshold, the output data of two nodes is considered correlated only when the absolute value of the standardized covariance is greater than the correlation threshold. Therefore, the standardized covariance between two nodes reflects the correlation of their output series. When affected by the same fluctuation region, the output time series of two nodes will show a certain degree of similarity. The closer the two nodes are to the fluctuation center, the higher the output correlation, and vice versa. The total influence of a node represents the sum of the correlations between the node and the rest of the nodes. The node influence ratio is used to normalize the total influence of a node. The two functions are to use the ratio of total influence to influence to determine the node's position (i.e., whether it is at the center). Secondly, the output correlation between nodes quantifies the impact of the fluctuation region. When predicting the node output at the next moment, the influence of the fluctuation region can be taken into account in the predicted value. The purpose of the node correlation ratio is to exclude nodes with weak correlation (i.e., the outermost nodes that may not be affected by the fluctuation area) and avoid randomness when searching for the fluctuation center.

[0061] It can be understood that the present invention, by adopting the node influence ratio and the node correlation ratio as the fluctuation characteristic values, can reflect the influence of the fluctuation area in the time series changes of the node characteristic values ​​and the correlation changes between nodes. By extracting the laws of spatiotemporal changes and adding them to the basic prediction, the output value of each node at the prediction moment that is increased or decreased due to the influence of fluctuations is compensated, thereby improving the accuracy of the output prediction.

[0062] It is understood that in step S3, the output data of each node within multiple historical time sections is collected, and the fluctuation characteristic value is calculated according to the definition formula of the node influence ratio and node correlation ratio described above. The output value of the corresponding node within the prediction time section is predicted based on the historical output data of each node. The purpose of this embodiment is to consider the more detailed impact of fluctuations on the output prediction value, thereby correcting the prediction value and improving the accuracy of the output prediction.

[0063] The process of determining whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections includes the following:

[0064] The node with the highest node correlation ratio in multiple time sections is taken as the fluctuation center. If the movement of the new and old fluctuation centers meets the continuity requirements, it is determined that generalized volatility exists.

[0065] It can be understood that the node with the highest node correlation ratio is usually the central node or key node of the distributed new energy power station system. This node has a strong correlation with the other nodes and reflects a large fluctuation influence. Therefore, if there are always one or more nodes with the highest node correlation ratio in a series of time sections, they can be regarded as the fluctuation center. Moreover, if the movement of the new and old fluctuation centers meets the continuity requirement, that is, the new fluctuation center is adjacent to the old fluctuation center and there are no other nodes between them, then it is determined that the output of the distributed new energy power station system has generalized fluctuation. For example, assuming that the output data of each node in 50 time sections are collected to calculate the fluctuation characteristic value, according to the fluctuation characteristic value calculation results, it is found that there is a fluctuation center A in time sections 1-25 and a fluctuation center B in time sections 26-50, and A and B are adjacent, then it is determined that there is generalized fluctuation. In addition, after determining the existence of generalized fluctuation, the fluctuation center position, the fluctuation characteristic value of each node (including the node influence ratio and the node correlation ratio) and the standardized covariance series in the most recent multiple time sections are obtained to facilitate the prediction of the fluctuation center and the characteristic value of each node in the next period of time.

[0066] It can be understood that in step S4, the process of predicting the fluctuation center position and the fluctuation characteristic value of each node in the next time section based on the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections includes the following:

[0067] Based on the standardized covariance sequence of each node in the most recent multiple time sections, data processing methods such as linear regression and curve fitting are used to predict the standardized covariance sequence of each node in the next period of time. The fluctuation characteristic value prediction value of each node is calculated based on the predicted standardized covariance sequence, and the new fluctuation center is determined.

[0068] For example, based on the standardized covariance sequence of each node in time sections 26-50, the standardized covariance sequence of each node in time section 51 and a series of subsequent time sections is predicted, and then based on the predicted standardized covariance sequence of each node, the node influence ratio prediction value and the node correlation ratio prediction value of each node are calculated, and the new fluctuation center position is determined according to the above definition of the fluctuation center.

[0069] It can be understood that this embodiment makes predictions based on a standardized covariance sequence with time series characteristics, and then indirectly calculates the predicted value of the fluctuation characteristic value. Compared with directly using the fluctuation characteristic value sequence for prediction, this can reduce calculation errors and facilitate more accurate prediction of the new fluctuation center position.

[0070] It can be understood that in step S4, the standardized covariance sequence, the predicted value of the fluctuation characteristic value and the new fluctuation center of each node in the future period have been predicted. In step S5, it is necessary to predict the actual output of each node based on the standardized covariance sequence, the predicted value of the fluctuation characteristic value and the new fluctuation center of each node in the predicted time period. Among them, first, according to the rule that the predicted value of the fluctuation characteristic value is from high to low and the position is from the fluctuation center (i.e., the new fluctuation center) to the periphery, the fluctuation area is first layered using a clustering method (such as the existing SVM clustering algorithm, the K-means clustering algorithm, etc.), and the highest cardinality of each layer is 1, that is, it represents the fluctuation center node, and decreases as the distance from the fluctuation center increases, and the lowest is 0; then, according to the volatility law, the fluctuation center node's own output is calculated. The fluctuation influence factor of each node in the layer is calculated, wherein the fluctuation center's fluctuation influence factor on each node in the layer = layer cardinality * standardized covariance between the fluctuation center node and the calculation node. In addition, the amount of fluctuation influence on the output of the fluctuation center node itself = the power change rate of the fluctuation center × the change in the fluctuation characteristic value of the fluctuation center (for example, the change in the node influence ratio). Based on the known series of output power values ​​and a series of fluctuation characteristic values ​​of the previous fluctuation center, the linear regression model is used to describe the relationship between the output power and the fluctuation characteristic value, which can be expressed as: P = β0 + β1x, x represents the fluctuation characteristic value, P represents the output power, β0 and β1 are coefficients, β0 represents the intercept, and β1 represents the power change rate of the fluctuation center. The values ​​of the coefficients β0 and β1 can be estimated by the least squares method, and the change in the fluctuation characteristic value of the fluctuation center = the fluctuation characteristic value of the old fluctuation center - the fluctuation characteristic value of the new fluctuation center, so that the amount of fluctuation influence on the output of the fluctuation center node itself can be calculated.

[0071] First, the output impact of the fluctuation is added to the output forecast value of the fluctuation center node at the prediction time as a correction value for the output forecast of the fluctuation center. The output impact of the fluctuation center node itself is also the standard value of other levels. The output forecast value of the fluctuation center node is obtained in step S3, that is, it is predicted based on the historical output data of the fluctuation center node.

[0072] Then, starting with the level with the greatest fluctuation impact, the output forecast correction value is calculated using the level's fluctuation impact factor and the node output forecast value. The output forecast correction value for the remaining nodes = the node's output forecast value × (1 + the output fluctuation impact rate of the fluctuation center node × the fluctuation impact factor). The output fluctuation impact rate of the fluctuation center node = the output fluctuation impact amount of the fluctuation center node / the output forecast correction value of the fluctuation center. This process is repeated for all levels until the corrected output forecast values ​​for all nodes are obtained.

[0073] In addition, after the correction of the predicted value at that moment is completed, the predicted corrected value at that moment is added to the power data sequence of the actual output to form a new power data sequence. Then, based on the new power data sequence, the standardized covariance sequence and the fluctuation characteristic value of each node are calculated. It is required that the error between the standardized covariance sequence and the predicted value of the fluctuation characteristic value obtained in step S4 does not exceed 5% (the specific value can be set according to actual needs). If the condition is not met, the fluctuation influence of the fluctuation center node at the predicted moment and the level cardinality of each level (selective depending on the specific situation) are adjusted and re-iterated. For example, if the error exceeds 5%, or there is a fluctuation characteristic value of more than half of the nodes that has an error of more than 5% between the corrected fluctuation characteristic value and the predicted fluctuation characteristic value, the fluctuation influence of the fluctuation center node is adjusted, and the new influence = the old influence × exp(-α*k), where α represents the center adjustment rate and k represents the number of iterations.

[0074] Repeat the above steps until the corrected forecast value sequence within the required forecast time period is obtained, and the output forecast is completed.

[0075] In addition, if Figure 2 As shown, another embodiment of the present invention further provides an output prediction system for a distributed new energy power station, preferably using the output prediction method as described above, including:

[0076] Topology construction module, used to build the geographical topology between distributed new energy power stations;

[0077] A fluctuation characteristic value setting module, used for setting the fluctuation characteristic value;

[0078] A generalized volatility judgment module is used to calculate the fluctuation characteristic value based on the output data of each node in multiple time sections, and judge whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections. If generalized volatility exists, the fluctuation center position and the fluctuation characteristic value of each node in the multiple time sections are obtained;

[0079] The fluctuation center prediction module is used to predict the fluctuation center position and the fluctuation characteristic value of each node in the next time section based on the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections;

[0080] The node output prediction module is used to predict the actual output values ​​of all nodes in the fluctuation area based on the fluctuation center position and the fluctuation characteristic value of each node in the next time section.

[0081] It can be understood that the output prediction system of the distributed new energy power station in this embodiment, after setting the fluctuation characteristic value, calculates the fluctuation characteristic value based on the output data of each node in multiple historical time sections, and judges whether there is generalized volatility based on the calculation results of the fluctuation characteristic values ​​of multiple historical time sections. If so, the fluctuation center position and the fluctuation characteristic value of each node in the predicted time section are predicted based on the fluctuation characteristic value of each node in multiple historical time sections. Finally, based on the predicted fluctuation center position and the fluctuation characteristic value of each node, the actual output value of all nodes in the fluctuation area is predicted. It combines the spatiotemporal coupling characteristics between distributed new energy power stations and the generalized volatility analysis, and compensates for the output value of each node at the prediction time that is increased or decreased due to the influence of volatility, thereby improving the output prediction accuracy of distributed new energy power stations under the influence of group random factors.

[0082] In addition, another embodiment of the present invention further provides an electronic device, including a processor and a memory, wherein the memory stores a computer program, and the processor is configured to execute the steps of the above method by calling the computer program stored in the memory.

[0083] In addition, another embodiment of the present invention further provides a computer-readable storage medium for storing a computer program for output prediction of a distributed new energy power station, wherein the computer program executes the steps of the above-described method when running on a computer.

[0084] Common computer-readable storage media include: floppy disks, flexible disks, hard disks, magnetic tape, any other magnetic media, CD-ROMs, any other optical media, punch cards, paper tape, any other physical medium with a pattern of holes, random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), flash-erasable programmable read-only memory (FLASH-EPROM), any other memory chip or cartridge, or any other medium that can be read by a computer. Instructions can further be transmitted or received via a transmission medium. The term transmission medium may include any tangible or intangible medium that can be used to store, encode, or carry instructions for execution by a machine, and includes digital or analog communication signals or other intangible media that facilitate communication of such instructions. Transmission media include coaxial cables, copper wire, and fiber optics, including the wires of a bus used to transmit a computer data signal.

[0085] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiment of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal translation scripting language JavaScript, etc.

[0086] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0087] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0088] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0089] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0090] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

[0091] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for predicting the output of a distributed new energy power station, characterized in that: Includes the following: Constructing the geographical topology structure among distributed new energy power stations; Setting a fluctuation characteristic value, wherein the fluctuation characteristic value is set to a power fluctuation rate, or the fluctuation characteristic value is set to a node influence ratio and a node correlation ratio; Calculate the fluctuation characteristic value based on the output data of each node in multiple time sections, and determine whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections. If generalized volatility exists, obtain the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections; Based on the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections, the fluctuation center position and the fluctuation characteristic value of each node in the next time section are predicted; Based on the fluctuation center position and the fluctuation characteristic value of each node in the next time section, the actual output value of all nodes in the fluctuation area is predicted; The process of determining whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections includes the following: In a certain time section, if the power fluctuation rate of each node in a certain area of ​​the topological structure has an extreme value area, and the power fluctuation rate gradually decreases outside the extreme value area, then the extreme value area is regarded as the fluctuation center. In the subsequent time sections, if the distance between the new fluctuation center and the old fluctuation center does not exceed the preset threshold, the power fluctuation rate of the nodes near the fluctuation center migration line is higher than that of other areas, and the power fluctuation rate sequence change of a single node is negatively correlated with the change in the distance between the node and the fluctuation center, then it is determined that generalized volatility exists; Alternatively, the process of determining whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections includes the following: The node with the highest node correlation ratio in multiple time sections is taken as the fluctuation center. If the movement of the new and old fluctuation centers meets the continuity requirements, it is determined that generalized volatility exists.

2. The method for predicting the output of a distributed new energy power station according to claim 1, wherein: The fluctuation characteristic value is the power fluctuation rate, where power fluctuation rate = (baseline output difference - actual output difference) / baseline mean, the baseline output difference is the difference between the baseline value at the next moment and the baseline value at the previous moment, the actual output difference is the difference between the output value at the next moment and the output value at the previous moment, and the baseline mean is the average value of the baseline value at the next moment and the baseline value at the previous moment.

3. The method for predicting the output of a distributed new energy power station according to claim 2, wherein: The process of predicting the actual output values ​​of all nodes in the fluctuation area based on the fluctuation center position and the fluctuation characteristic value of each node in the next time section includes the following: For each node, a multiple regression model is constructed with the actual output of the node as the variable and the output baseline and power fluctuation rate of the node as the influencing factors. The actual output data, output baseline and power fluctuation rate of the node in the known time section are then substituted into the multiple regression model to calculate the coefficients of the multiple regression model. The power fluctuation rate and output baseline of the node in the next time section that are predicted are then substituted into the multiple regression model to calculate the actual output value of the node in the next time section. Repeat the above process to obtain the actual output values ​​of all nodes in the fluctuation area.

4. The method for predicting the output of a distributed new energy power station according to claim 1, wherein: The fluctuation characteristic values ​​are the node influence ratio and the node correlation ratio, wherein the node influence ratio is the ratio of the total node influence of a certain node to the average of the total node influence of all nodes, the total node influence of each node is the sum of the absolute values ​​of the standardized covariance sequence of the output value of the node and the output values ​​of all other nodes at multiple time sections, and the node correlation ratio is the ratio of the number of nodes in the standardized covariance sequence whose absolute values ​​are greater than the preset correlation threshold to (N-1), where N is the number of nodes.

5. The method for predicting the output of a distributed new energy power station according to claim 1, wherein: The process of predicting the fluctuation center position and the fluctuation characteristic value of each node in the next time section based on the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections includes the following: Based on the standardized covariance sequence of each node in the most recent multiple time sections, the standardized covariance sequence of each node in the next period of time is predicted, and the fluctuation characteristic value prediction value of each node is calculated according to the predicted standardized covariance sequence, and the new fluctuation center is determined.

6. An output prediction system for a distributed new energy power station, using the output prediction method according to any one of claims 1 to 5, characterized in that: include: Topology construction module, used to build the geographical topology between distributed new energy power stations; A fluctuation characteristic value setting module, used for setting the fluctuation characteristic value; A generalized volatility judgment module is used to calculate the fluctuation characteristic value based on the output data of each node in multiple time sections, and judge whether generalized volatility exists based on the calculation results of the fluctuation characteristic values ​​of multiple time sections. If generalized volatility exists, the fluctuation center position and the fluctuation characteristic value of each node in the multiple time sections are obtained; The fluctuation center prediction module is used to predict the fluctuation center position and the fluctuation characteristic value of each node in the next time section based on the fluctuation center position and the fluctuation characteristic value of each node in multiple time sections; The node output prediction module is used to predict the actual output values ​​of all nodes in the fluctuation area based on the fluctuation center position and the fluctuation characteristic value of each node in the next time section.

7. An electronic device, characterized in that: The method comprises a processor and a memory, wherein a computer program is stored in the memory, and the processor is configured to execute the steps of the method according to any one of claims 1 to 5 by calling the computer program stored in the memory.

8. A computer-readable storage medium for storing a computer program for predicting the output of a distributed renewable energy power station, characterized in that: When the computer program is run on a computer, the steps of the method according to any one of claims 1 to 5 are executed.

Images