Urban transformer area interconnection power control method, device and equipment and storage medium

Abstract

The invention discloses an urban transformer area interconnection power control method, device and equipment and a storage medium, and the method comprises the steps: obtaining a photovoltaic prediction error sequence based on a photovoltaic actual output curve and a prediction output curve, and carrying out the fitting of a Gaussian mixture model, and obtaining photovoltaic output error probability distribution; constructing a plurality of groups of random numbers obeying the error probability distribution, and performing load flow calculation according to network parameters of the urban power distribution network and a distributed power supply reference value to obtain node voltage and branch load flow probability distribution; randomly generating a plurality of sample data calculation system out-of-limit risk index values according to node voltage and branch power flow probability distribution, and determining a high-risk transformer area according to the index values; and combining each high-risk transformer area and the adjacent low-risk transformer area, calculating the complementation degree, and selecting the transformer area combination with the highest complementation degree to carry out flexible direct interconnection and power optimization control. According to the invention, imbalance of urban distribution network source-load space-time distribution can be stabilized, the energy utilization rate is improved, and the stability of urban power grid operation is improved.

Description

Technical Field

[0001] The present application relates to the technical field of power systems, and in particular to a method, device, equipment and storage medium for controlling power interconnection between urban substations. Background Art

[0002] With the rapid development of urbanization, the scale of cities and their load volume continue to increase. The continuous growth of urban electricity demand has put forward higher requirements for the power supply capacity of the power system, which is mainly manifested in the following aspects: (1) The contradiction between power supply capacity and the increase in urban electricity consumption: With the growth of urban population, electricity load is highly concentrated, and electricity consumption has increased significantly, while the capacity of existing power supply lines can no longer meet the surge in transmission demand. The expansion of distribution lines must take into account factors such as urban commercial real estate development planning and urban road construction. The expansion of overhead lines and cables is difficult. (2) The contradiction between power quality and changes in power load: The intelligent manufacturing industry and data information industry are developing rapidly. A large number of loads sensitive to power quality such as electronic computers, chip production, communication control, and high-precision medical equipment have been added to urban high-tech industrial zones. Any slight power fluctuation may cause damage to production products, data loss in data centers, or system paralysis, and even cause medical accidents. (3) The current is not distributed according to the natural current: Under the background of energy system reform and new power system construction, the penetration rate of distributed generation (DG) continues to increase, and the voltage of the distribution network no longer follows the step-by-step decrease from the feeder to the load end. The voltage at the DG access point will rise due to the DG access. Moreover, the farther the DG access point is from the system power supply, the more obvious the effect of the line end voltage being raised. The risk of the distribution system node voltage exceeding the limit increases. When the DG output is excessive, the distribution current backflow problem will also occur.

[0003] With the rapid development of power electronics and their control technology, flexible DC interconnection technology based on power electronics has been gradually applied to distribution systems to address the aforementioned new power quality issues in distribution networks. Compared to traditional distribution networks, flexible interconnection devices (FIDs) in flexible distribution networks have the ability to continuously regulate the power flows of multiple branches connected to a node in multiple directions. To address the urgent need for flexible interconnection in urban distribution networks, FIDs are used to implement flexible DC interconnection on existing urban AC distribution lines. Through interconnection between substations, power generation is mutually complementary, smoothing out peak and valley power consumption within or between substations. This also adjusts the spatial and temporal distribution imbalance of distributed renewable energy between the city center and surrounding suburban power supply areas. This allows for dynamic capacity expansion of urban distribution systems without expanding overhead lines or cables or increasing supply voltage levels. This alleviates system power flow congestion and voltage over-limit issues, promotes the absorption of distributed renewable energy within the region, and reduces power backflow to the upper transmission system. Secondly, the presence of a DC link in the flexible interconnection effectively isolates AC fault currents, reducing the scope of the fault. Combined with its rapid and independent control of active and reactive power flows across multiple connected branches, a fault quickly forms power supply islands in non-fault-related power outage zones, ensuring uninterrupted power supply to critical loads in the distribution system or shortening outage durations and minimizing outage losses. Furthermore, DC grids eliminate the power quality issues inherent in AC distribution networks, often caused by factors such as frequency, power angle, and reactance. This ensures that high-quality power, such as zero harmonics and excellent voltage stability, is met by high-precision urban loads, providing a DC interface for these loads.

[0004] When performing flexible direct current interconnection transformation on the substation, how to select the substation for flexible direct current interconnection is a technical problem that needs to be solved urgently by those skilled in the art. Summary of the Invention

[0005] Based on this, in response to the above technical problems, the present application provides a method, device, equipment and storage medium for urban substation interconnection power control.

[0006] In view of this, the first aspect of the present application provides a method for controlling power of interconnected urban substations, comprising:

[0007] A photovoltaic prediction error sequence is obtained based on the photovoltaic actual output curve and the photovoltaic predicted output curve, and the photovoltaic prediction error sequence is fitted using a Gaussian mixture model to obtain a photovoltaic output error probability distribution;

[0008] For nodes containing photovoltaic generators, several groups of random numbers that obey the photovoltaic output error probability distribution are constructed. Power flow calculations are performed based on the network parameters of the urban distribution network, the distributed power supply reference values, and the random numbers in each group. The node voltage and branch power flow probability distribution corresponding to each group of random numbers are obtained.

[0009] Randomly generate a plurality of sample data according to the node voltage and the branch power flow probability distribution, calculate a system over-limit risk index value according to the sample data, determine high-risk nodes and branches and low-risk nodes and branches according to the system over-limit risk index value, and obtain high-risk substations and low-risk substations;

[0010] Combining each of the high-risk substations with the adjacent low-risk substations to obtain multiple substation combinations, calculating the complementarity of each of the combined substations based on the typical daily load curve of each substation in each substation combination, and selecting the substation combination with the highest complementarity for flexible direct current interconnection;

[0011] A power optimization control model for a flexible DC interconnected area combination is established, and the power optimization control model is solved to obtain a power control strategy for the flexible DC interconnected area combination.

[0012] Optionally, obtaining a photovoltaic prediction error sequence based on the photovoltaic actual output curve and the photovoltaic predicted output curve includes:

[0013] The difference between the actual photovoltaic output value and the predicted photovoltaic output value at the same time in the actual photovoltaic output curve and the predicted photovoltaic output curve is calculated to obtain the photovoltaic prediction error sequence.

[0014] Optionally, obtaining a photovoltaic prediction error sequence based on the photovoltaic actual output curve and the photovoltaic predicted output curve includes:

[0015] Combine the photovoltaic output values ​​in the actual photovoltaic output curve and the photovoltaic predicted output curve to obtain a photovoltaic output joint vector;

[0016] Extracting statistical features from the photovoltaic output joint vector to obtain a photovoltaic feature vector;

[0017] Performing cluster analysis on the photovoltaic feature vectors as photovoltaic samples to generate actual photovoltaic output scenarios and predicted photovoltaic output scenarios on typical days;

[0018] The difference between the actual photovoltaic output value and the predicted photovoltaic output value at the same time in the typical photovoltaic day actual output scenario and the typical photovoltaic day predicted output scenario is calculated to obtain a photovoltaic prediction error sequence.

[0019] Optionally, performing cluster analysis on the photovoltaic feature vectors as photovoltaic samples to generate a typical photovoltaic day actual output scenario and a typical photovoltaic day predicted output scenario includes:

[0020] S1, taking the photovoltaic characteristic vector as a photovoltaic sample, and calculating a distance index value according to the difference between the characteristic value in each photovoltaic sample and the minimum value of the characteristic values ​​in all photovoltaic samples;

[0021] S2. Sort all the photovoltaic samples in descending order or ascending order based on the distance index value, and evenly divide the sorted photovoltaic samples based on a set number of clusters to obtain multiple photovoltaic sample sets;

[0022] S3, sorting the photovoltaic samples and taking the photovoltaic sample at the center as the initial cluster center;

[0023] S4, calculating the Euclidean distance between each photovoltaic sample and each initial cluster center, and assigning each photovoltaic sample to the nearest cluster center based on the Euclidean distance to obtain multiple clusters;

[0024] S5. Generate new cluster centers in each cluster and return to step S4 until convergence to obtain multiple clustering results;

[0025] S6. Calculate the total silhouette coefficient of each clustering result, select the photovoltaic sample in the clustering result with the largest total silhouette coefficient for splitting, and obtain the actual photovoltaic output scenario and the predicted photovoltaic output scenario on a typical day.

[0026] Optionally, the system over-limit risk index value includes a node voltage over-limit risk index and a branch power over-limit risk index; and calculating the system over-limit risk index value based on the sample data includes:

[0027] Calculating a node voltage loss value based on the node voltage, the voltage upper limit value, and the voltage lower limit value in the sample data, and calculating a node voltage over-limit risk index based on the node voltage loss value;

[0028] The branch power loss value is calculated based on the branch transmission power, power limit and load rate limit in the sample data, and the branch power over-limit risk index is calculated based on the branch power loss value.

[0029] Optionally, the calculating the complementarity of each combined area according to the typical daily load curve of each area in each area combination includes:

[0030] Superimposing the typical daily load curves of each area in each area combination to obtain a superimposed curve of each area combination;

[0031] Calculating the ratio of the average value to the maximum value of the sum of the load rates of the superimposed curves of each of the substation combinations in each time period to obtain the first complementarity of each of the substation combinations;

[0032] Calculating the ratio of the minimum value to the maximum value of the sum of the load rates of the superimposed curves of each of the substation combinations in each time period to obtain a second complementarity of the substation combination;

[0033] The product of the first complementarity and the second complementarity of each of the metro area combinations is calculated to obtain the complementarity of each of the metro area combinations.

[0034] Optionally, the objective function of the power optimization control model is:

[0035]

[0036] Where, is the DC system power loss at time t; is the AC system power loss at time t, T is the number of moments in the control cycle;

[0037] The constraints of the power optimization control model include:

[0038] The DC bus side should transmit power balance constraints:

[0039]

[0040] Transmission power balance constraints in the AC side area:

[0041]

[0042] Transmission capacity constraints of the DC converter stations on both sides of the flexible interconnection device:

[0043]

[0044] Where, 、 are the active power transmitted by the DC converter on one side of the flexible interconnection device and the active power transmitted by the DC converter on the other side at time t respectively; is the DC load power connected to the DC side at time t; 、 are the charging power and discharging power of the DC side energy storage at time t respectively; 、 are the active power input to the one-side and the second-side substations of the flexible interconnection device at time t respectively; 、 are the AC load power of the flexible interconnection device at time t, which is the power of the station area on one side and the station area on the second side respectively; 、 are the capacity of the DC converter on one side of the flexible interconnection device and the capacity of the DC converter on the other side respectively; 、 They are the reactive power transmitted by the DC converter on one side of the flexible interconnection device at time t and the reactive power transmitted by the DC converter on the two sides.

[0045] A second aspect of the present application provides a city substation interconnection power control device, comprising:

[0046] an error distribution acquisition unit, configured to acquire a photovoltaic prediction error sequence based on a photovoltaic actual output curve and a photovoltaic predicted output curve, and to fit the photovoltaic prediction error sequence using a Gaussian mixture model to obtain a photovoltaic output error probability distribution;

[0047] A power flow calculation unit is used to construct several groups of random numbers that obey the photovoltaic output error probability distribution for nodes containing photovoltaic generator groups, perform power flow calculations based on network parameters of the urban distribution network, distributed power supply reference values, and each group of random numbers, and obtain the node voltage and branch power flow probability distribution corresponding to each group of random numbers;

[0048] an over-limit risk assessment unit, configured to randomly generate a plurality of sample data according to the node voltage and the branch power flow probability distribution, calculate a system over-limit risk index value according to the sample data, determine high-risk nodes and branches and low-risk nodes and branches according to the system over-limit risk index value, and obtain high-risk substations and low-risk substations;

[0049] a complementarity calculation unit, configured to combine each of the high-risk substations with the adjacent low-risk substations to obtain a plurality of substation combinations, calculate the complementarity of each of the combined substations based on the typical daily load curve of each substation in each of the substation combinations, and select the substation combination with the highest complementarity for flexible direct current interconnection;

[0050] The power control unit establishes a power optimization control model for the flexible direct current interconnected area combination, solves the power optimization control model, and obtains a power control strategy for the flexible direct current interconnected area combination.

[0051] A third aspect of the present application provides an electronic device, the device comprising a processor and a memory;

[0052] The memory is used to store program code and transmit the program code to the processor;

[0053] The processor is used to execute the urban substation interconnection power control method described in any one of the first aspects according to the instructions in the program code.

[0054] In a fourth aspect, the present application provides a computer-readable storage medium for storing program code. When the program code is executed by a processor, the method for controlling the interconnected power of urban substations described in any one of the first aspects is implemented.

[0055] It can be seen from the above technical solutions that this application has the following advantages:

[0056] The present application provides a method for controlling power interconnection in urban substations, which consists of two parts: risk analysis of node voltage and feeder current exceeding the limit based on the uncertainty of distributed energy output prediction error, and interconnection power control strategy based on substation load complementarity. The former evaluates the risk of node and feeder exceeding the limit in the urban distribution network based on the idea of ​​probabilistic current, analyzes whether the distribution network needs to be transformed into flexible direct current interconnection, and provides substations with high risk of exceeding the limit as pre-selected sites for medium and low voltage flexible direct current interconnection transformation. The latter further analyzes the complementarity between substation load curves on the obtained pre-selected sites, and measures the feasibility of flexible direct current interconnection of quantitative substations by constructing a complementarity index. Flexible direct current interconnection is carried out between overloaded substations, ordinary substations, and substations with excess distributed energy output, thereby smoothing out the imbalance in the temporal and spatial distribution of sources and loads in the urban distribution network, improving energy utilization, and improving the stability of urban power grid operation. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 A flow chart of a method for controlling power in urban area interconnection provided in an embodiment of the present application;

[0059] Figure 2 A schematic diagram of a process for solving Gaussian mixture model parameters using the expectation-maximization algorithm provided in an embodiment of the present application;

[0060] Figure 3 A schematic diagram of a probabilistic power flow calculation process provided in an embodiment of the present application;

[0061] Figure 4 A schematic structural diagram of a city substation interconnection power control device provided in an embodiment of the present application;

[0062] Figure 5 A schematic structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0063] In order to help those skilled in the art better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0064] For easier understanding, please refer to Figure 1 , an embodiment of the present application provides a method for controlling power of interconnected urban substations, including:

[0065] Step 110: Obtain a photovoltaic prediction error sequence based on the photovoltaic actual output curve and the photovoltaic predicted output curve, and fit the photovoltaic prediction error sequence using a Gaussian mixture model to obtain a photovoltaic output error probability distribution;

[0066] In one embodiment, the photovoltaic actual output curve and the photovoltaic predicted output curve (historical data) can be obtained from the photovoltaic historical data, and then the difference between the photovoltaic actual output value and the photovoltaic predicted output value at the same time in the photovoltaic actual output curve and the photovoltaic predicted output curve is directly calculated to obtain the photovoltaic prediction error sequence.

[0067] In another embodiment, considering that if the photovoltaic prediction error sequence calculated from all photovoltaic historical data (a year of photovoltaic historical data has 356 samples) is used to construct the photovoltaic output error probability distribution, it will lead to a large amount of subsequent calculations and produce a dimensionality curse. In order to reduce the amount of data when constructing the photovoltaic output error probability distribution and improve the calculation efficiency, the embodiment of the present application combines the photovoltaic output values ​​in the photovoltaic actual output curve and the photovoltaic predicted output curve to obtain a photovoltaic output joint vector; extracts statistical features from the photovoltaic output joint vector to obtain a photovoltaic feature vector; uses the photovoltaic feature vector as a photovoltaic sample for cluster analysis to generate a photovoltaic typical day actual output scenario and a photovoltaic typical day predicted output scenario; calculates the difference between the photovoltaic actual output value and the photovoltaic predicted output value at the same time in the photovoltaic typical day actual output scenario and the photovoltaic typical day predicted output scenario to obtain a photovoltaic prediction error sequence.

[0068] This application reduces all photovoltaic historical data by clustering, and extracts representative photovoltaic typical day actual output data and photovoltaic typical day predicted output data. Specifically, the photovoltaic output values ​​in the photovoltaic actual output curve and the photovoltaic predicted output curve are combined to obtain a photovoltaic output joint vector; statistical features are extracted from the photovoltaic output joint vector to obtain a photovoltaic feature vector; the photovoltaic feature vector is used as the input of the K-means clustering algorithm to generate photovoltaic typical day actual output scenarios and photovoltaic typical day predicted output scenarios; the difference between the photovoltaic actual output value and the photovoltaic predicted output value at the same time in the photovoltaic typical day actual output scenario and the photovoltaic typical day predicted output scenario is calculated to obtain a photovoltaic prediction error sequence.

[0069] Actual PV output and predicted PV output are two different random variables. If actual and predicted PV output data are independently clustered to generate two typical scenarios, the clustering results cannot effectively reflect how the error between actual and predicted PV output varies over time. The actual and predicted PV output of a given PV power station are two different variables in the same time series. Therefore, clustering typical daily actual and predicted PV output scenarios falls under the category of multivariate time series clustering. It is necessary to fuse the actual and predicted PV output curves for the day into a joint vector to obtain the PV output joint vector. Assuming that every hour is used as a sampling point, the actual photovoltaic output curve of a day contains a total of 24 actual photovoltaic output values, and the predicted photovoltaic output curve of a day contains 24 predicted photovoltaic output values. After the actual photovoltaic output values ​​and the predicted photovoltaic output values ​​of a day are merged into a photovoltaic output joint vector, a photovoltaic sample (i.e., a photovoltaic output joint vector) contains 48-dimensional data, with the first 24 dimensions being the actual photovoltaic output values ​​and the last 24 dimensions being the predicted photovoltaic output values.

[0070] The PV output joint vector can be used as input for cluster analysis. Based on the aforementioned joint vector formation process, a PV output joint vector contains 48-dimensional data. Directly using the joint vector for clustering on a year's worth of PV output data can lead to a curse of dimensionality. Therefore, before performing cluster analysis, it is necessary to extract features from the PV output joint vector and reduce the dimensionality of the high-dimensional data. This allows for cluster analysis using low-dimensional eigenvalue vectors instead of the high-dimensional original data vectors.

[0071] This application extracts 6 statistical features from the photovoltaic joint vector, the photovoltaic feature vector. These 6 statistical features are:

[0072] (1)

[0073] Where, is the actual photovoltaic output value at time t; is the predicted photovoltaic output value at time t; is the number of time series contained in one day's data. If the sampling interval is 1 hour, then is 24; The average value of the actual photovoltaic output reflects the average daily level of the actual photovoltaic output; The average value of photovoltaic output prediction reflects the overall accuracy of the photovoltaic output prediction model; The actual photovoltaic output variance reflects the daily fluctuation of the actual photovoltaic output; is the photovoltaic output prediction variance, reflecting the stability of the photovoltaic output prediction model; is the mean PV output error, reflecting the degree to which the PV output prediction model overestimates or underestimates the actual output; is the photovoltaic error variance, reflecting the volatility of the forecast.

[0074] The photovoltaic feature vector is used as the input data of the K-means clustering algorithm to generate the actual output scene of the photovoltaic typical day and the predicted output scene of the photovoltaic typical day. In the traditional K-means clustering algorithm, the initial cluster center needs to be given in advance, but the algorithm itself does not provide a selection principle for the initial cluster center. The initial cluster center is usually randomly selected, which makes the clustering result less stable. Therefore, the embodiment of the present application introduces the distance index value as an evaluation index Assists in the selection of initial cluster centers. The evaluation index value reflects the distance between all samples and the minimum value sample. The calculation formula is:

[0075] (2)

[0076] Where, is the photovoltaic eigenvector No. eigenvalues; is the first photovoltaic sample among all photovoltaic samples (one photovoltaic feature vector is one photovoltaic sample). The minimum value of the eigenvalues; is the number of photovoltaic eigenvectors; is the number of eigenvalues ​​in the photovoltaic eigenvector. Sort the photovoltaic feature vectors in descending (or ascending) order. The similarity between photovoltaic feature vectors is positively correlated with the eigenvalue, so adjacent photovoltaic samples have high similarity. Divide the sorted photovoltaic samples into k equal parts (k is the number of clusters) to obtain k photovoltaic sample sets. Select the photovoltaic sample at the center of each photovoltaic sample set as the initial cluster center.

[0077] The steps for generating the actual photovoltaic output scenario and the predicted photovoltaic output cluster scenario on a typical day using the K-means clustering method are as follows:

[0078] 1) Initialization of the number of clusters: Set the number of m categories to be , when initialized, take i=1;

[0079] 2) Cluster center initialization: calculation of evaluation indicators , according to the evaluation index Select k i Initial cluster centers;

[0080] 3) Classification: Calculate the Euclidean distance between each photovoltaic sample and each cluster center, and assign the photovoltaic sample to the nearest cluster center according to the Euclidean distance to form k i clusters;

[0081] 4) Cluster center solution: Calculate the average value of each eigenvalue in all photovoltaic samples within the cluster in turn to form a new eigenvector consisting of the average value of the six eigenvalues, and use this as the new cluster center;

[0082] 5) Convergence judgment: Determine whether the clustering has converged based on the following formula:

[0083] (3)

[0084] Where, For category Photovoltaic samples; For category The cluster center of . Reaching a limit condition or value When the change range of meets the requirements, the clustering converges;

[0085] 6) Order , repeat steps 2) to 5) until the last category number Clustering is also completed.

[0086] 7) Determine the optimal number of clusters: The above number of clusters is different according to the number of categories. The total silhouette coefficient of each clustering result is calculated using the following formula. The number of clusters corresponding to the cluster group with the largest total silhouette coefficient is the optimal number of clusters.

[0087] (4)

[0088] (5)

[0089] Where, is the minimum value of the average distance between photovoltaic sample i and photovoltaic samples of other clusters; is the average distance between PV sample i and other PV samples in the same cluster; is the silhouette coefficient of photovoltaic sample i; The number of clusters is The total silhouette coefficient corresponding to the clustering result.

[0090] 8) Generate a typical day PV actual output curve and a typical day PV forecast output curve: The clustering result corresponding to the cluster group with the largest total silhouette coefficient is the final clustering result. The final clustering result includes the optimal number of clusters, cluster centers, and the eigenvector samples contained in each cluster. Based on the analysis in the previous section, each PV eigenvector corresponds to a joint vector of actual PV output and forecast PV output (the joint vector has 48 dimensions: the first 24 dimensions are actual output values, and the last 24 dimensions are forecast values). Therefore, the joint vector corresponding to the cluster center is split to obtain a typical day PV actual output curve and a typical day PV forecast output curve. By splitting the joint vector corresponding to all cluster centers in this way, multiple typical day PV actual output scenarios and multiple typical day PV forecast output scenarios are obtained.

[0091] Photovoltaic output exhibits a multimodal distribution due to weather conditions. A single normal distribution cannot capture this multimodal characteristic. This application introduces a Gaussian mixture model, which incorporates Gaussian weights to describe the multimodal nature of the distribution. The Gaussian mixture model achieves clustering by optimizing the mixture weights and Gaussian distribution parameters through expectation maximization. It can handle clusters of varying sizes, shapes, and orientations. The parameters of the Gaussian mixture model consist of the weight of the mth Gaussian component (representing the probability of the mth scenario), the mean of the mth Gaussian component (representing the accuracy of the photovoltaic forecast error under the mth scenario), and the variance (representing the stability of the photovoltaic forecast under the mth scenario).

[0092] The output curve P corresponding to a typical photovoltaic day actual output scenario obtained through the above process is PV (i.e., the actual output curve of photovoltaic typical day) and the predicted output scenario of photovoltaic typical day P pre (i.e. the typical day photovoltaic output curve) is:

[0093] (6)

[0094] The actual output scenario of photovoltaic typical day P PV and the typical daily predicted output scenario of photovoltaic pre The photovoltaic prediction error series is:

[0095] (7)

[0096] The probability density function of the photovoltaic prediction error sequence based on the Gaussian fuzzy model is:

[0097] (8)

[0098] Where K is the total number of Gaussian components; 、 、 are the weight, mean and variance of the kth Gaussian component respectively. The unknown parameters of the Gaussian mixture model are (include 、 、 ) can be solved using the Expectation-Maximization algorithm (EM), the specific steps are as follows Figure 2 As shown:

[0099] (1) Parameter initialization: set the initial value 、 、 ;

[0100] (2) Step E (Expectation): According to the initialization 、 、 or the result of the previous iteration 、 、 , calculate for each data point The posterior probability of belonging to the kth component :

[0101] (9)

[0102] Where, 、 、 Indicates the initial value 、 、 The result of the tth iteration;

[0103] (3) M-step (Maximization): Update parameters to maximize expectations

[0104] (10)

[0105] (4) Convergence judgment: When any of the following convergence criteria is met, the algorithm stops and outputs the latest parameters.

[0106] Convergence criterion 1: Repeat the E step and the M step until the minimum threshold of parameter change is met :

[0107] (11)

[0108] Convergence criterion 2: reaching the maximum number of iterations.

[0109] Step 120: For nodes containing photovoltaic generators, construct several groups of random numbers that obey the photovoltaic output error probability distribution, perform power flow calculations based on the network parameters of the urban distribution network, the distributed power reference values, and each group of random numbers, and obtain the node voltage and branch power flow probability distribution corresponding to each group of random numbers.

[0110] Among the many methods for probabilistic power flow calculation, simulation is a common approach. This approach randomly extracts a large amount of input data based on the probability distribution of the input variables, generates a large number of scenarios, performs deterministic power flow calculations on each set of scenarios, and then calculates the power flow results of all scenarios to form a probability distribution of the system state variables. The present embodiment uses the Monte Carlo simulation method for probabilistic power flow calculations.

[0111] The power system flow equation can be simplified as follows:

[0112] (12)

[0113] Where, Inject power vectors into system nodes; is the system branch power flow vector; is the system network parameter; is the node voltage.

[0114] The PV DG output error can be added to the node injection power as a summand of the node injection power. The above formula can be rewritten as:

[0115] (13)

[0116] Where, Inject power error vectors into system nodes.

[0117] The steps of probabilistic power flow calculation based on Monte Carlo simulation are as follows: Figure 3 As shown. Obtain the network parameters of the urban distribution network and the benchmark values ​​of distributed power sources in the system. The network parameters include network topology, branch impedance parameters, and node load. The benchmark value of distributed power sources uses the photovoltaic typical day forecast output curve generated in step 110 as a benchmark. For nodes containing photovoltaic generators, construct a probability distribution that obeys the photovoltaic output error. Several sets of random numbers are generated. Based on the network parameters of the urban distribution network and the benchmark values ​​of distributed generation, deterministic power flow calculations are performed using each set of random numbers, such as the Newton-Raphson method and the PQ decomposition method. The probability distribution of node voltages and branch power flows in the distribution system corresponding to each set of random numbers is obtained. Statistical analysis is performed on all results to calculate the mathematical expectation and probability distribution of the output variables.

[0118] Step 130: randomly generate a number of sample data based on the probability distribution of node voltages and branch power flows, calculate a system over-limit risk index value based on the sample data, determine high-risk nodes and branches and low-risk nodes and branches based on the system over-limit risk index value, and obtain high-risk substations and low-risk substations;

[0119] The embodiment of the present application uses a utility function to establish the relationship between the loss value and the risk index. The utility function can be:

[0120] (14)

[0121] Where, Represents the loss value of the object under investigation. The first-order derivative of the utility function and the second-order derivative are all greater than 0, that is , A first-order derivative greater than 0 indicates that the dissatisfaction expressed by the utility function increases with the increase in the object's loss value. A second-order derivative greater than 0 indicates that the rate of change in dissatisfaction increases with the increase in the object's loss value. If the object under investigation is node voltage, the greater the voltage fluctuation, the greater the voltage limit risk.

[0122] For nodes with DG access, the node voltage will increase due to the power backflow into the distribution network. Excessive voltage will reduce the insulation level of the access point line. The utility function is used to construct the node voltage limit risk indicator:

[0123] (15)

[0124] Where N is the number of random samples; is the voltage loss value of node i in sample s, and n is the number of nodes in the system. The node voltage loss value is defined as follows:

[0125] (16)

[0126] Where, is the node voltage of node i in sample s; 、 They are the upper and lower voltage limits allowed for the node voltage respectively.

[0127] The embodiment of the present application uses the branch load rate indicator to represent the severity of the branch overload risk. The more the system branch load rate exceeds the limit value L0, the higher the risk severity. When the system branch load rate is lower than the limit value L0, the system risk is 0. The utility function is used to construct the branch power limit risk indicator:

[0128] (17)

[0129] Where, is the power loss value of branch j in sample s, and m is the number of branches in the system. The branch power loss value is defined as follows:

[0130] (18)

[0131] Where, L j is the ratio of the actual transmission power of branch j to its limit.

[0132] N sample data are randomly generated based on the node voltage and branch power flow probability distribution generated in step 120. The sample data are substituted into formula (15) and formula (17) to calculate the system over-limit risk index value (including the node voltage over-limit risk index and the branch power over-limit risk index). The over-limit risk of each node in the distribution network is evaluated based on the node voltage over-limit risk index, and the over-limit risk of each branch is evaluated based on the branch power over-limit risk index. High-risk nodes (such as nodes with node voltage over-limit risk index values ​​exceeding the first node over-limit risk threshold are evaluated as high-risk nodes), high-risk branches (such as branches with branch power over-limit risk index values ​​exceeding the first branch over-limit risk threshold are evaluated as high-risk branches), as well as low-risk nodes (such as nodes with node voltage over-limit risk index values ​​lower than the second node over-limit risk threshold are evaluated as low-risk nodes), and low-risk branches (such as branches with branch power over-limit risk index values ​​lower than the second branch over-limit risk threshold are evaluated as low-risk branches), thereby obtaining high-risk substations and low-risk substations. This application selects nodes and branches with high risk indicators as pre-selected sites for flexible direct current interconnection.

[0133] Step 140: Combine each high-risk substation with an adjacent low-risk substation to obtain multiple substation combinations. Calculate the complementarity of each combined substation based on the typical daily load curve of each substation in each substation combination, and select the substation combination with the highest complementarity for flexible direct current interconnection.

[0134] The complementarity of the substations is expressed as the complementarity between the typical daily load curves of the substations. The greater the complementarity of the typical daily load curves, the greater the complementarity of the substations, and the greater the feasibility of flexible direct current interconnection between the corresponding substations. If there is a significant time difference between the valley and peak periods of the typical daily load curves of the two substations, for example, the peak power consumption of substation A is during the day and the valley power consumption is at night, and the valley power consumption of substation B is during the day and the peak power consumption is at night, then the trend of the comprehensive typical daily load curve obtained by superimposing the typical daily load curves of substations A and B will tend to be flat. Based on the shape characteristics of the superimposed comprehensive typical daily load curve, this application quantifies the degree of complementarity of the typical daily load curves of the substations into a first complementarity index and a second complementarity index.

[0135] The first complementarity is the ratio of the average value to the maximum value of the sum of the load rates of different areas in each period:

[0136] (19)

[0137] Where, T is the number of statistical cycle times; n is the number of substations participating in the complementary analysis; is the load size of station x at time t.

[0138] The second complementarity is the ratio of the minimum to maximum sum of the load rates of different areas in each period:

[0139] (20)

[0140] The comprehensive complementarity is the product of the first complementarity and the second complementarity:

[0141] (twenty one)

[0142] The steps for feasibility analysis of flexible direct current interconnection in substations using the above complementarity indicators are as follows:

[0143] (1) Collect, organize, verify and normalize the load data of each substation to form a typical daily load curve for each substation;

[0144] (2) Superimposing and combining the typical daily load curves of the high-risk substation obtained by the analysis in step 130 with the low-risk substations adjacent to the high-risk substation to form a variety of substation combinations; calculating the complementarity of the superimposed curves based on the comprehensive complementarity evaluation index to obtain the complementarity of each substation combination;

[0145] (3) Sort the substation combinations in descending order according to their complementarity, and select the substation combination with the highest complementarity for flexible direct current interconnection transformation.

[0146] Step 150: Establish a power optimization control model for the flexible DC interconnection area, solve the power optimization control model, and obtain a power control strategy for the flexible DC interconnection area.

[0147] The flexible interconnection device in the flexible distribution network connects the DC side to two voltage source converters (VSC) in a back-to-back manner. The flexible interconnection device in the flexible distribution network has the ability to continuously control the power flow of multiple branches connected to a node in multiple directions. The FID transmission active power and reactive power model is:

[0148] (twenty two)

[0149] From formula (22), we can see that controlling the voltage phase angle and amplitude of the converters on both sides can control the active power and reactive power flowing through the FID. Therefore, there are four controlled quantities, namely, the active power output on one side, and reactive power , the active power output on both sides and reactive power . In different operating scenarios of the distribution network, the control modes of the four variables of FID are also different. In normal operating mode, usually one side of the VSC realizes stable control of the DC voltage, and the other side realizes stable control of the DC transmission power. In addition, since the FID isolates the AC systems on both sides through the DC link, the output reactive power of the VSCs on both sides is controlled separately; in the power grid fault mode, the VSC on the non-fault side provides DC voltage stability control, and the VSC on the fault side provides frequency stability control. Based on the above analysis, the embodiment of the present application proposes an FID operation optimization control strategy with minimal network loss in the urban flexible interconnected distribution network, and establishes a power optimization control model for the flexible direct current interconnected substation based on this control strategy. The objective function of the model is:

[0150] (twenty three)

[0151] Where, is the DC system power loss at time t; is the AC system power loss at time t, T is the number of moments in the control cycle;

[0152] The constraints of the power optimization control model include:

[0153] (1) The DC bus side should transmit power balance constraints:

[0154] (twenty four)

[0155] Where, 、 are the active power transmitted by the DC converter on one side of FID at time t and the active power transmitted by the DC converter on the other side of FID at time t respectively; is the DC load power connected to the DC side at time t; 、 are the charging power and discharging power of the DC side energy storage at time t respectively.

[0156] (2) Transmission power balance constraints in the AC side area:

[0157] (25)

[0158] Where, 、 are the active power input from the FID one side area and the active power input from the FID two side areas at time t respectively; 、 are the AC load powers of the FID one-side and second-side substations at time t.

[0159] (3) Transmission capacity constraints of DC converter stations on both sides of FID:

[0160] (26)

[0161] Where, 、 are the VSC capacities on one side and two sides respectively; 、 They are the reactive power transmitted by the DC converter on one side of FID at time t and the reactive power transmitted by the DC converter on the other side of FID at time t.

[0162] In addition to meeting the above FID-related constraints, the basic power flow constraints of the distribution network must also be met, which will not be repeated here. The power optimization control model can be solved using commercial solvers such as CPLEX and GUROBI or conventional intelligent algorithms to provide the FID controllable variables that optimize the network loss under the system operation constraints. 、 、 、 control scheme.

[0163] This application comprehensively evaluates the feasibility of flexible direct current (FDC) interconnection between substations by constructing the complementarity of typical daily load curves between them. For the preselected FDC installation sites (i.e., nodes and branches with high over-limit risk) generated in the previous steps, this section analyzes the load complementarity of the substations to identify substations with a matching complementarity with the preselected sites, forming a pair of FDC nodes (FDC interconnection requires the connection of two or more substations). Finally, a power control strategy based on the FID is developed to minimize distribution network losses.

[0164] This application consists of two parts: an analysis of the risk of node voltage and feeder current exceeding the limit based on the uncertainty of the output forecast error of distributed energy resources, and an interconnection power control strategy based on the complementarity of the load of the substation. The former evaluates the risk of nodes and feeders exceeding the limit in the urban distribution network based on the idea of ​​probabilistic current, analyzes whether the distribution network needs to be transformed into a flexible direct current interconnection, and gives the substation with high risk of exceeding the limit as the pre-selected site for the transformation of medium and low voltage flexible direct current interconnection. The latter further analyzes the complementarity between the load curves of the substations on the obtained pre-selected sites, and measures the feasibility of flexible direct current interconnection of the substations by constructing a complementarity index. Flexible direct current interconnection is carried out between overloaded substations, ordinary substations, and substations with excess distributed energy output, so as to smooth out the imbalance in the temporal and spatial distribution of sources and loads in the urban distribution network, improve energy utilization, and improve the stability of urban power grid operation.

[0165] Please refer to Figure 4 The embodiment of the present application further provides a city substation interconnection power control device, comprising:

[0166] The error distribution acquisition unit 410 is used to acquire a photovoltaic prediction error sequence based on the photovoltaic actual output curve and the photovoltaic predicted output curve, and fit the photovoltaic prediction error sequence using a Gaussian mixture model to obtain a photovoltaic output error probability distribution;

[0167] The power flow calculation unit 420 is used to construct several groups of random numbers that obey the probability distribution of photovoltaic output errors for nodes containing photovoltaic generators, perform power flow calculations based on network parameters of the urban distribution network, distributed power generation reference values, and each group of random numbers, and obtain the node voltage and branch power flow probability distribution corresponding to each group of random numbers;

[0168] The over-limit risk assessment unit 430 is configured to randomly generate a plurality of sample data based on the probability distribution of node voltages and branch power flows, calculate a system over-limit risk index value based on the sample data, and determine high-risk nodes and branches and low-risk nodes and branches based on the system over-limit risk index value, thereby obtaining high-risk and low-risk substation areas.

[0169] The complementarity calculation unit 440 is configured to combine each high-risk substation with an adjacent low-risk substation to obtain multiple substation combinations, calculate the complementarity of each combined substation based on the typical daily load curve of each substation in each substation combination, and select the substation combination with the highest complementarity for flexible direct current interconnection;

[0170] The power control unit 450 establishes a power optimization control model for the flexible DC interconnection substation combination, solves the power optimization control model, and obtains a power control strategy for the flexible DC interconnection substation combination.

[0171] As a further improvement, the error distribution acquisition unit 410 is specifically configured to calculate the difference between the actual photovoltaic output value and the predicted photovoltaic output value at the same time in the actual photovoltaic output curve and the predicted photovoltaic output curve to obtain a photovoltaic prediction error sequence; and to fit the photovoltaic prediction error sequence using a Gaussian mixture model to obtain a photovoltaic output error probability distribution.

[0172] Alternatively, the photovoltaic output values ​​in the actual photovoltaic output curve and the photovoltaic predicted output curve are combined to obtain a photovoltaic output joint vector;

[0173] Extract statistical features from the photovoltaic output joint vector to obtain a photovoltaic feature vector;

[0174] The photovoltaic feature vectors are used as photovoltaic samples for cluster analysis to generate the actual photovoltaic output scenario and the predicted photovoltaic output scenario on a typical day.

[0175] Calculate the difference between the actual photovoltaic output value and the predicted photovoltaic output value at the same time in the typical photovoltaic day actual output scenario and the typical photovoltaic day predicted output scenario to obtain the photovoltaic prediction error sequence;

[0176] The Gaussian mixture model is used to fit the photovoltaic prediction error series to obtain the photovoltaic output error probability distribution.

[0177] As a further improvement, the over-limit risk assessment unit 430 is specifically configured to randomly generate a plurality of sample data according to the probability distribution of the node voltage and the branch power flow;

[0178] Calculate the node voltage loss value based on the node voltage, voltage upper limit value and voltage lower limit value in the sample data, and calculate the node voltage over-limit risk index based on the node voltage loss value;

[0179] Calculate the branch power loss value based on the branch transmission power, power limit and load rate limit in the sample data, and calculate the branch power limit risk index based on the branch power loss value;

[0180] According to the node voltage over-limit risk index and the branch power over-limit risk index, high-risk nodes and branches and low-risk nodes and branches are determined, and high-risk substations and low-risk substations are obtained.

[0181] As a further improvement, the complementarity calculation unit 440 is specifically configured to combine each high-risk area with an adjacent low-risk area to obtain a plurality of area combinations;

[0182] Superimpose the typical daily load curves of each substation in each substation combination to obtain the superimposed curve of each substation combination;

[0183] Calculate the ratio of the average value to the maximum value of the sum of the load rates of the superimposed curves of each substation combination in each time period to obtain the first complementarity of each substation combination;

[0184] Calculate the ratio of the minimum value to the maximum value of the sum of the load rates of the superimposed curves of each substation combination in each time period to obtain the second complementarity of the substation combination;

[0185] Calculate the product of the first complementarity and the second complementarity of each area combination to obtain the complementarity of each area combination;

[0186] Select the substation combination with the highest degree of complementarity for flexible direct current interconnection.

[0187] This application consists of two parts: an analysis of the risk of node voltage and feeder current exceeding the limit based on the uncertainty of the output forecast error of distributed energy resources, and an interconnection power control strategy based on the complementarity of the load of the substation. The former evaluates the risk of nodes and feeders exceeding the limit in the urban distribution network based on the idea of ​​probabilistic current, analyzes whether the distribution network needs to be transformed into a flexible direct current interconnection, and gives the substation with high risk of exceeding the limit as the pre-selected site for the transformation of medium and low voltage flexible direct current interconnection. The latter further analyzes the complementarity between the load curves of the substations on the obtained pre-selected sites, and measures the feasibility of flexible direct current interconnection of the substations by constructing a complementarity index. Flexible direct current interconnection is carried out between overloaded substations, ordinary substations, and substations with excess distributed energy output, so as to smooth out the imbalance in the temporal and spatial distribution of sources and loads in the urban distribution network, improve energy utilization, and improve the stability of urban power grid operation.

[0188] Please refer to Figure 5 , an embodiment of the present application further provides an electronic device, the device including a processor 510 and a memory 520;

[0189] The memory 520 is used to store program codes and transmit the program codes to the processor 510;

[0190] The processor 510 is configured to execute the urban substation interconnection power control method in the aforementioned method embodiment according to the instructions in the program code.

[0191] An embodiment of the present application also provides a computer-readable storage medium, which is used to store program code. When the program code is executed by a processor, the urban substation interconnection power control method in the aforementioned method embodiment is implemented.

[0192] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0193] In the specification of this application and the above-mentioned drawings, the terms "first," "second," "third," "fourth," etc. (if any) are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. In addition, the terms "including" and "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus that includes a series of steps or elements is not necessarily limited to those steps or elements explicitly listed, but may include other steps or elements not explicitly listed or inherent to such process, method, product, or apparatus.

[0194] It should be understood that in this application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships can exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following items" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or plural.

[0195] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0196] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0197] In addition, the functional units in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0198] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for executing all or part of the steps of the method described in each embodiment of the present application through a computer device (which can be a personal computer, server, or network device, etc.). The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (full name: Read-Only Memory, English abbreviation: ROM), random access memory (full name: Random Access Memory, English abbreviation: RAM), disk or optical disk, and other media that can store program code.

[0199] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for controlling power of interconnected urban substations, characterized in that: include: A photovoltaic prediction error sequence is obtained based on the photovoltaic actual output curve and the photovoltaic predicted output curve, and the photovoltaic prediction error sequence is fitted using a Gaussian mixture model to obtain a photovoltaic output error probability distribution; For nodes containing photovoltaic generators, several groups of random numbers that obey the photovoltaic output error probability distribution are constructed. Power flow calculations are performed based on the network parameters of the urban distribution network, the distributed power supply reference values, and the random numbers in each group. The node voltage and branch power flow probability distribution corresponding to each group of random numbers are obtained. Randomly generate a plurality of sample data according to the node voltage and the branch power flow probability distribution, calculate a system over-limit risk index value according to the sample data, determine high-risk nodes and branches and low-risk nodes and branches according to the system over-limit risk index value, and obtain high-risk substations and low-risk substations; Combining each of the high-risk substations with the adjacent low-risk substations to obtain multiple substation combinations, calculating the complementarity of each of the combined substations based on the typical daily load curve of each substation in each substation combination, and selecting the substation combination with the highest complementarity for flexible direct current interconnection; A power optimization control model for a flexible DC interconnected area combination is established, and the power optimization control model is solved to obtain a power control strategy for the flexible DC interconnected area combination.

2. The urban area interconnection power control method according to claim 1, characterized in that: The step of obtaining a photovoltaic prediction error sequence based on the photovoltaic actual output curve and the photovoltaic predicted output curve includes: The difference between the actual photovoltaic output value and the predicted photovoltaic output value at the same time in the actual photovoltaic output curve and the predicted photovoltaic output curve is calculated to obtain the photovoltaic prediction error sequence.

3. The urban area interconnection power control method according to claim 1, characterized in that: The step of obtaining a photovoltaic prediction error sequence based on the photovoltaic actual output curve and the photovoltaic predicted output curve includes: Combine the photovoltaic output values ​​in the actual photovoltaic output curve and the photovoltaic predicted output curve to obtain a photovoltaic output joint vector; Extracting statistical features from the photovoltaic output joint vector to obtain a photovoltaic feature vector; Performing cluster analysis on the photovoltaic feature vectors as photovoltaic samples to generate actual photovoltaic output scenarios and predicted photovoltaic output scenarios on typical days; The difference between the actual photovoltaic output value and the predicted photovoltaic output value at the same time in the typical photovoltaic day actual output scenario and the typical photovoltaic day predicted output scenario is calculated to obtain a photovoltaic prediction error sequence.

4. The urban area interconnection power control method according to claim 3, characterized in that: The cluster analysis of the photovoltaic feature vectors as photovoltaic samples to generate a typical photovoltaic day actual output scenario and a typical photovoltaic day predicted output scenario includes: S1, taking the photovoltaic characteristic vector as a photovoltaic sample, and calculating a distance index value according to the difference between the characteristic value in each photovoltaic sample and the minimum value of the characteristic values ​​in all photovoltaic samples; S2. Sort all the photovoltaic samples in descending order or ascending order based on the distance index value, and evenly divide the sorted photovoltaic samples based on a set number of clusters to obtain multiple photovoltaic sample sets; S3, sorting the photovoltaic samples and taking the photovoltaic sample at the center as the initial cluster center; S4, calculating the Euclidean distance between each photovoltaic sample and each initial cluster center, and assigning each photovoltaic sample to the nearest cluster center based on the Euclidean distance to obtain multiple clusters; S5. Generate new cluster centers in each cluster and return to step S4 until convergence to obtain multiple clustering results; S6. Calculate the total silhouette coefficient of each clustering result, select the photovoltaic sample in the clustering result with the largest total silhouette coefficient for splitting, and obtain the actual photovoltaic output scenario and the predicted photovoltaic output scenario on a typical day.

5. The urban area interconnection power control method according to claim 1, characterized in that: The system over-limit risk index value includes a node voltage over-limit risk index and a branch power over-limit risk index; The calculating of the system over-limit risk index value according to the sample data includes: Calculating a node voltage loss value based on the node voltage, the voltage upper limit value, and the voltage lower limit value in the sample data, and calculating a node voltage over-limit risk index based on the node voltage loss value; The branch power loss value is calculated based on the branch transmission power, power limit and load rate limit in the sample data, and the branch power over-limit risk index is calculated based on the branch power loss value.

6. The urban area interconnection power control method according to claim 1, characterized in that: The calculating the complementarity of each combined area according to the typical daily load curve of each area in each area combination includes: Superimposing the typical daily load curves of each area in each area combination to obtain a superimposed curve of each area combination; Calculating the ratio of the average value to the maximum value of the sum of the load rates of the superimposed curves of each of the substation combinations in each time period to obtain the first complementarity of each of the substation combinations; Calculating the ratio of the minimum value to the maximum value of the sum of the load rates of the superimposed curves of each of the substation combinations in each time period to obtain a second complementarity of the substation combination; The product of the first complementarity and the second complementarity of each of the metro area combinations is calculated to obtain the complementarity of each of the metro area combinations.

7. The urban area interconnection power control method according to claim 1, characterized in that: The objective function of the power optimization control model is: Where, is the DC system power loss at time t; is the AC system power loss at time t, T is the number of moments in the control cycle; The constraints of the power optimization control model include: The DC bus side should transmit power balance constraints: Transmission power balance constraints in the AC side area: Transmission capacity constraints of the DC converter stations on both sides of the flexible interconnection device: Where, 、 are the active power transmitted by the DC converter on one side of the flexible interconnection device and the active power transmitted by the DC converter on the other side at time t respectively; is the DC load power connected to the DC side at time t; 、 are the charging power and discharging power of the DC side energy storage at time t respectively; 、 are the active power input to the one-side and the second-side substations of the flexible interconnection device at time t respectively; 、 are the AC load power of the flexible interconnection device at time t, which is the power of the station area on one side and the station area on the second side respectively; 、 are the capacity of the DC converter on one side of the flexible interconnection device and the capacity of the DC converter on the other side respectively; 、 They are the reactive power transmitted by the DC converter on one side of the flexible interconnection device at time t and the reactive power transmitted by the DC converter on the two sides.

8. A power control device for interconnected urban substations, characterized in that: include: an error distribution acquisition unit, configured to acquire a photovoltaic prediction error sequence based on a photovoltaic actual output curve and a photovoltaic predicted output curve, and to fit the photovoltaic prediction error sequence using a Gaussian mixture model to obtain a photovoltaic output error probability distribution; A power flow calculation unit is used to construct several groups of random numbers that obey the photovoltaic output error probability distribution for nodes containing photovoltaic generator groups, perform power flow calculations based on network parameters of the urban distribution network, distributed power supply reference values, and each group of random numbers, and obtain the node voltage and branch power flow probability distribution corresponding to each group of random numbers; an over-limit risk assessment unit, configured to randomly generate a plurality of sample data according to the node voltage and the branch power flow probability distribution, calculate a system over-limit risk index value according to the sample data, determine high-risk nodes and branches and low-risk nodes and branches according to the system over-limit risk index value, and obtain high-risk substations and low-risk substations; a complementarity calculation unit, configured to combine each of the high-risk substations with the adjacent low-risk substations to obtain a plurality of substation combinations, calculate the complementarity of each of the combined substations based on the typical daily load curve of each substation in each of the substation combinations, and select the substation combination with the highest complementarity for flexible direct current interconnection; The power control unit establishes a power optimization control model for the flexible direct current interconnected area combination, solves the power optimization control model, and obtains a power control strategy for the flexible direct current interconnected area combination.

9. An electronic device, characterized in that: The device includes a processor and a memory; The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the urban substation interconnection power control method described in any one of claims 1 to 7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium is used to store program code, and when the program code is executed by a processor, the urban substation interconnection power control method described in any one of claims 1 to 7 is implemented.

Images