CN-115659554-B - Power supply unit division method considering uncertainty of photovoltaic and load

Abstract

The invention discloses a power supply unit division method considering photovoltaic and load uncertainty, which comprises a photovoltaic output and load uncertainty modeling method based on a K-means clustering multi-scene method, a photovoltaic output and load uncertainty power supply unit division model construction method considering the aim of balancing the maximum load rate of a power supply line in each power supply unit with the minimum number of power supply units, and a power supply unit division model solving method considering the photovoltaic output and the load uncertainty, so that a power supply unit division result is obtained. The invention makes up the defects of the traditional power supply unit dividing method under the background of large-scale access of the distributed photovoltaic and provides a theoretical basis for the construction of a novel power distribution system.

Inventors

• YANG BO
• LIU SHULI
• YANG SHENQUAN
• WANG KEXIN
• QI LUJIE
• LI ZHAO
• WANG CHEN
• ZHANG WEN
• DENG SHAOZHI
• LI KAI
• ZHANG XIAOLEI
• LIU ZHAO
• LU ZHIPENG
• LI HAO
• ZHENG ZHIJIE
• LIANG RONG
• WANG YANSHUO
• YANG YANG
• CUI CAN
• ZHAO REN
• WANG YAOLEI

Assignees

• 国网山东省电力公司经济技术研究院

Dates

Publication Date

20260508

Application Date

20220818

Claims (4)

1. 1. A method of partitioning power supply units taking photovoltaic and load uncertainty into account, the method comprising: Step 1, a method for modeling uncertainty of photovoltaic output and load based on a K-means clustering multi-scene method; Step 2, a photovoltaic output and load uncertainty power supply unit division model construction method is considered with the aim of balancing the minimum number of power supply units and the maximum load rate of the internal power supply line in each power supply unit; step 3, a power supply unit division model solving method considering photovoltaic output and load uncertainty comprises five parts, namely an inter-station power supply unit and intra-station power supply unit sequential division frame, a net load quantity calculating method considering DG output, an inter-station power supply unit division flow, an intra-station power supply unit division flow and a final division scheme; (1) Inter-station and intra-station power supply unit sequential division frame Considering the power supply unit division of the load characteristic matching between the connection feeder lines, firstly dividing the power supply units between the stations based on the load rate constraint of the feeder lines, fully excavating the power supply capability between the stations of the transformer substation, under the corresponding load rate constraint, generating N step groups of power supply unit division schemes between the stations by changing the positions of the feeder lines between the stations for multiple iterations, taking the average value of the load synchronous rate in each power supply unit between the stations as an index, evaluating the advantages and disadvantages of the power supply unit division between the stations, selecting a plurality of groups of power supply unit division schemes between the stations, carrying out the power supply unit division in the stations considering the load characteristic matching between the connection feeder lines on the basis, forming the power supply unit division scheme under the load rate constraint of the feeder lines between the stations, thereby realizing the sequential division of the power supply units between the stations, (2) Payload calculation method considering DG output The average value of the obtained net load characteristic curves considering the DG output force is obtained to obtain the net load quantity considering the DG output force, , wherein, Is the average value of the time sequence output of the net load, N is the number of scenes, mu i (t) is the time sequence output of the net load of the ith scene, (3) Inter-station power supply unit division flow Firstly, determining the number of outgoing lines among transformer stations according to the requirements of a load scale and a power supply model in a power supply grid, dividing by controlling the load rate of feeder lines among the stations, excavating the power supply capacity among the transformer stations, and introducing a specific dividing flow by taking a group of power supply units among the stations of the S 1 、S 2 transformer station as an example: step 1, determining the farthest distance of connection between every two substations according to the geographical bordering condition of each substation, and setting the load rate of the inter-substation feeder line by assuming that the feeder lines a i and b j are initial positions of the inter-substation feeder lines of the two substations, wherein k=1 , Step 2, initially making i=1 and j=1, calculating weighted distances from each load point in the transformer substations S 1 and S 2 to the feeder line a i and the feeder line b j , attributing the load with the smallest weighted distance into the power supply range of the feeder lines a i and b j , The specific expression of the weighted distance and the weighted factor is as follows: , , , Wherein, the inter-station feeder line a i and b j form a power supply unit, omega 1 represents a weighting factor, q is an amplification factor of the weighting factor related to the peak-valley difference rate of the power supply unit, eta max represents the maximum value of the time sequence output average value of the net load in the power supply unit after the load point and the photovoltaic are added into the feeder line a i or the feeder line b j , eta min represents the minimum value of the time sequence output average value of the net load in the power supply unit after the load or the photovoltaic is added, when the load point and the photovoltaic are added to be favorable for improving the peak-valley difference of the power supply unit, the value of the weighting factor omega 1 is reduced, the weighting distance is also reduced, the possibility that the load point belongs to the power supply unit is increased, and the endogenous load characteristic of the power supply unit is further ensured to be matched, Representing the variance of the magnitude of the net load in the power supply unit at the t-th moment after the load point and the photovoltaic are added to the feeder a i or the feeder b j , describing the fluctuation degree of the magnitude of the net load in the power supply unit, when the variance of the magnitude of the net load is reduced by the addition of the load point and the photovoltaic, the fluctuation degree and the uncertainty of the magnitude of the net load are reduced by the addition of the load, the weighting factor omega 2 is reduced, the weighting distance is reduced correspondingly, the load point or the photovoltaic is added to the power supply unit, d m,i ' and d m,i represent the Euclidean distance and the weighting distance of the mth load to the feeder a i respectively, d n,j ' and d n,j represent the Euclidean distance and the weighting distance of the nth load to the feeder b j respectively, Step 3, angle updating is carried out on the feeder lines a i and b j according to the loaded position, so that the feeder lines pass through the loaded geographic position center, Step 4, saving the power supply range division scheme of the feeder lines a i and b j , Step 5, calculating probability information of payload load rates of the feeder lines a i and b j , Step 6, judging whether the load rate of the feeder line a i under a certain confidence coefficient is smaller than a preset load rate constraint, if the maximum value of the load rate of the feeder line a i under a certain confidence coefficient is smaller than the preset load rate constraint, jumping to step 2, continuing to divide the power supply range of the feeder line a i , otherwise continuing to step 7, Step 7, outputting a division scheme of a power supply range of the feeder line a i-1 , Step 8, judging whether the maximum value of the load rate of the feeder b j under a certain confidence is smaller than a preset load rate constraint, if the maximum value of the load rate of the feeder b j under a certain confidence is smaller than the preset load rate constraint, jumping to step 2, continuing to divide the power supply range of the feeder b j , otherwise continuing to step 9, Step 9, outputting a division scheme of the power supply range of the feeder line b j-1 , Step 10, judging whether the iteration times N step are reached, if so, performing step 11, otherwise rotating the initial angles of the feed lines a i and b j by delta theta 1 , returning to step 2, Step 11, calculating the maximum load rate of the transfer line under a certain confidence level in each group of power supply unit division schemes, Step 12, outputting a power supply unit division scheme meeting the constraint of maximum load rate of the internal power supply line of the power supply unit, optimizing the power supply unit division scheme between stations by taking the minimum average value of the load synchronous rate of the power supply units between stations as an index, And 13, outputting a plurality of groups of power supply unit division schemes among stations.
2. 2. The power supply unit division method considering photovoltaic and load uncertainty according to claim 1 is characterized in that the uncertainty modeling method of photovoltaic output and load based on a K-means clustering multi-scene method specifically comprises the following steps: Processing the DG output uncertainty and the DG load uncertainty by adopting a multi-scenario method, collecting the output and the load of the DG at each moment in a full scenario, processing the DG output and the load at each moment by adopting a K-means clustering method, finally obtaining the typical output and the occurrence probability, the load time sequence characteristic and the occurrence probability of the DG at each moment, obtaining the net load scenario of the load of the user side by combining the DG output characteristic with the load characteristic of the user side of the DG access position, obtaining the distribution of all scenarios of the net load of the user side at each moment in a power supply unit according to the following scenario modeling, Wherein A is the distribution of all scenes at each moment, N is the number of scenes, and tau is the net load value of each scene at each moment.
3. 3. The power supply unit division method considering uncertainty of photovoltaic and load according to claim 1 is characterized in that the power supply unit division model construction method specifically comprises the following steps: The method comprises the steps of establishing a power supply unit division model objective function considering photovoltaic output and load uncertainty, wherein the model objective function specifically comprises the following steps: On the basis of minimum quantity, the maximum load rate balance of transfer supply lines in each power supply unit is required to be met, and the dividing advantages and disadvantages of the power supply units can be further judged, wherein the specific objective function is shown in the following formula: , , , Wherein, the Representing the average value of the expected maximum value of the maximum load rate of the transfer line in each power supply unit, gamma imax representing the expected maximum value of the maximum load rate of the transfer line in the ith power supply unit, beta representing the expected variance of the maximum load rate of the transfer line in each power supply unit, so as to describe the balance among the maximum load rates of the transfer line in each power supply unit, Y, Y zj and Y zn representing the numbers of all power supply units, inter-station power supply units and intra-station power supply units respectively, The method comprises the following steps of establishing a constraint condition of a power supply unit division model considering photovoltaic output and load uncertainty, wherein the constraint condition is specifically as follows: (1) Maximum load factor constraint of internal power supply line in power supply unit Assuming that the average value of the net load size in the first power supply unit at time t 1 is the largest, the constraint of the maximum load rate of the internal power supply line in the power supply unit is based on time t 1 , and Q l t1 is the size of the net load of the first power supply unit at time t 1 , and the constraint conditions are as follows: , Where cos phi is the power factor, L max is the maximum value of the line transmission capacity, Using opportunistic constraint planning can translate the formula into: Where ε is the confidence level that the constraint is satisfied, transforming the above equation may result in: the uncertainty is processed by adopting a multi-scene method, the distribution of all scenes of the net load of the power supply unit at each moment is obtained by the following formula, , wherein, The payload value per scene for each moment of time for the first power supply unit, N being the number of scenes, (2) Feeder load rate constraint Assuming that the average value of the payload size carried by feeder i is maximum at time t 2 , the feeder load factor constraint should be based on time t 2 , P i t2 is the payload size carried by feeder i at time t 2 , The i-th feeder load factor constraint is as follows: , the above formula is transformed by adopting the opportunity constraint, and the transformation is as follows: , The uncertainty is processed by adopting a multi-scene method, the distribution of the net load of the feeder line at each moment in all scenes is obtained by the following formula, , Wherein, the The net load value of each scene at each moment of the feeder line i is calculated, and N is the scene number.
4. 4. The method for solving the power supply unit division model by considering the uncertainty of the photovoltaic output and the load according to claim 1 is characterized by further comprising the following steps: (4) In-station power supply unit division flow The division of the power supply units in the station is based on the division scheme of the power supply units between stations, the power supply units which consider the matching of the load characteristics among the interconnecting feeder lines are divided in the power supply areas in the stations of each transformer substation by taking a group of the interconnecting feeder lines in the station as a research unit, Firstly, taking a transformer substation as a center, generating a plurality of central lines in an area in the substation by an angle equipartition principle, mutually connecting every two adjacent central lines to form an in-substation power supply unit, secondly, calculating the weighted distance from each load point to each central line, defining the weighted distance to be the same as the definition in the division of the in-substation power supply units, dividing each load point into the power supply range of each central line according to the principle of minimum weighted distance and meeting the opportunity constraint of the maximum load rate of the in-substation power supply unit, thereby ensuring the load characteristic matching of the in-substation power supply unit, updating the angle of each central line according to the load position of the central line after the division of the power supply range of each central line is finished, leading the central line to pass through the center of the loaded geographic position, finally, carrying out multiple iterations to finish the division of the in-substation power supply unit, (5) Acquisition of final partitioning scheme Traversing all the feasible schemes, comparing the number of power supply units in each scheme with the variance of the maximum load rate of the internal power supply line of each power supply unit, and selecting the power supply unit division scheme with the minimum variance of the maximum load rate of the internal power supply line of each power supply unit on the basis of the minimum number of the power supply units.

Description

Power supply unit division method considering uncertainty of photovoltaic and load Technical Field The patent provides a power supply unit division method considering photovoltaic and load probability information based on a traditional power supply unit division method of a power distribution network, and belongs to the field of grid division of the power distribution network. Background As an important link of grid planning of a power distribution network and construction of a novel power distribution system, the power supply unit division method of the power distribution network needs to be closely matched with the development trend of the current age. However, due to the intermittence and volatility of the distributed photovoltaic output, the large-scale access of the distributed photovoltaic output will have a great influence on the traditional distribution grid power supply unit division method. In addition, the random characteristics of the power load may also cause variation in the division result of the power supply unit, affected by weather, user behavior, and the like. In order to completely implement the new development concept, a power supply unit dividing method considering uncertainty of photovoltaic and load is provided under the background of large-scale access of distributed photovoltaic, so that the construction of a novel power distribution system is supported. Disclosure of Invention In order to solve the defects and shortcomings existing in the prior art, the patent provides a power supply unit dividing method considering uncertainty of photovoltaic and load, so as to make up for the defects of a traditional power supply unit dividing method under the background of large-scale access of distributed photovoltaic, and provide a theoretical basis for construction of a novel power distribution system. Specifically, the power supply unit dividing method considering the uncertainty of the photovoltaic and the load provided by the application comprises the following steps: (1) The uncertainty modeling method of the photovoltaic output and the load of the multi-scene method based on K-means clustering; (2) The photovoltaic output and load uncertainty power supply unit division model construction method is considered with the aim of balancing the maximum load rate of the internal power supply line in each power supply unit with the minimum number of power supply units; (3) The power supply unit division model solving method considering photovoltaic output and load uncertainty comprises five parts, namely an inter-station power supply unit and intra-station power supply unit sequential division thought, a net load quantity calculating method considering DG output, an inter-station power supply unit division flow, an intra-station power supply unit division flow and a final division scheme. The uncertainty modeling method of the photovoltaic output and the load of the multi-scene method based on K-means clustering in the step (1) specifically comprises the following steps: The scene analysis method is a common effective method for processing uncertainty problems, and the essence of the scene analysis method is that the randomness problems are discretized and discretized into a plurality of scenes, so that the uncertain mathematical problems in the whole scenes are converted into the definite mathematical problems in the single scenes to be solved, and the establishment of a complex random mathematical model which is difficult to solve is avoided. However, such a method has problems such as large data size and long calculation time. In the research, if the data size to be processed is large, a data clustering method is mostly adopted. Data clustering is used to group given data sets, and data sets of larger size are divided into small data sets of smaller size that are easier to process, often referred to as such, based on the inherent nature of the data and the correlation between the data. Object similarity in the same class is extremely large and object similarity in different "classes is extremely small. K-means clustering is one of the partitioning methods, which is a typical distance-based clustering method. The K-means clustering takes Euclidean distance as a similarity evaluation criterion, considers that the distance between objects in the same cluster is as small as possible, and the distance between objects in different clusters is as large as possible. Note that the sample dataset d= { X 1,X2,...,Xn }, n represents the number of samples, the i-th sample X i＝{xi1,xi2,...,xip }, p represents the number of attributes or index categories of the samples, and K represents the number of clusters. In the K-means cluster, the Euclidean distance between any two samples X i and X f is in the form: At the beginning of clustering, K samples are randomly selected as initial clustering centers, and then non-clustering center samples are classified into a class C h represented by a clustering center h