A line loss early warning method for low voltage active network considering theoretical limit and fluctuation amplitude

Abstract

The present invention relates to a low-voltage active network line loss early warning method considering theoretical limits and fluctuation amplitudes, and belongs to the field of power system line loss management. It includes: establishing a comparison table of theoretical limit values ​​of load rate and line loss rate and a fluctuation limit model of line loss; obtaining historical load, distributed power output, and line loss values ​​and clustering them to obtain the probability distribution of each clustered line loss rate; obtaining the predicted distributed power output and load, and then obtaining the line loss rate and the probability distribution of line loss at the next moment, the theoretical limit value of the line loss rate, and the fluctuation limit value of the line loss; based on the line loss rate and the probability distribution of line loss at the next moment, the theoretical limit value of the line loss rate, and the fluctuation limit value of the line loss, the probability of line loss exceeding the limit at the next moment is obtained and a line loss early warning is performed. When obtaining the probability of line loss exceeding the limit, the present invention takes into account the theoretical limit of line loss, the fluctuation limit value of line loss, and the probability of exceeding the line loss rate threshold, thereby improving the accuracy and effectiveness of the line loss early warning.

Description

Technical Field

[0001] The present invention relates to the technical field of power system line loss management, and in particular to a low-voltage active network line loss early warning method taking into account theoretical limits and fluctuation amplitudes. Background Art

[0002] Line loss is an important economic and technical indicator for evaluating the level of refined management of power grids. With the continuous development of power systems, low-voltage distribution networks, as an important part of power systems, have become a key link in optimizing energy utilization, improving power supply quality and economic benefits. However, in the low-voltage distribution network of the new power system, due to the large-scale access of distributed power sources and the increased diversity and uncertainty of load-side electrical equipment, the line loss of the low-voltage active network presents complex and changeable characteristics.

[0003] The line loss warning of the low-voltage active network can determine whether the line loss at the next moment will exceed the normal range or the preset threshold by predicting the line loss at the next moment, and issue a warning based on the judgment result. The existing line loss warning method usually only determines whether the line loss or line loss rate exceeds the preset fixed threshold, but this method has the following two defects: First, this method does not consider the theoretical limit value of the line loss rate under a given power grid structure. Since the theoretical limit value of the line loss rate is different under different load levels, it may appear that the theoretical limit value of the line loss rate under the current load level is higher than the preset threshold, resulting in the inability to accurately warn the line loss at the current moment; second, this method does not consider whether the amplitude of the line loss change at the current moment relative to the previous moment exceeds the limit, ignoring the monitoring of abnormal fluctuations in the line loss amplitude, which may lead to the omission of the line loss warning at the current moment.

[0004] It can be seen that providing a method that comprehensively considers the theoretical limit value and fluctuation range of the line loss rate under a given power grid structure to improve the accuracy and effectiveness of low-voltage active network line loss warning is an urgent problem to be solved. Summary of the invention

[0005] In view of the above analysis, the present invention aims to provide a low-voltage active network line loss early warning method taking into account the theoretical limit and fluctuation amplitude, so as to solve the problem that the existing low-voltage active network line loss early warning method does not take into account the theoretical limit value and fluctuation amplitude of the line loss rate under a given power grid structure, resulting in low accuracy and effectiveness of the low-voltage active network line loss early warning.

[0006] The present invention provides a low-voltage active network line loss early warning method considering theoretical limits and fluctuation amplitudes, the method comprising the following steps:

[0007] Obtain the rated capacity of the low-voltage active area distribution transformer and distributed power source, and the load of the low-voltage user, use equivalent network power flow analysis to establish a theoretical limit value model of the low-voltage active network line loss rate, and obtain a comparison table of the theoretical limit values ​​of the distribution transformer low-voltage side load rate and the low-voltage active network line loss rate based on the equivalent network power flow analysis and the theoretical limit value model of the low-voltage active network line loss rate;

[0008] Based on the amplitude variation of the load on the low-voltage side of the distribution transformer at adjacent moments in the low-voltage active area, a fluctuation limit model of the line loss of the low-voltage active network is established;

[0009] Obtain the historical low-voltage side load of the distribution transformer in the low-voltage active area, the historical output of distributed power sources, and the historical line loss values, and cluster them to obtain multiple cluster categories. The line loss rate of the samples in each cluster category is statistically analyzed to obtain the probability distribution of the line loss rate of the cluster category.

[0010] Construct a correlation equation and an autoregressive moving average model, obtain the predicted distributed power output based on the correlation equation, and obtain the predicted distribution transformer low-voltage side load based on the autoregressive moving average model; obtain the corresponding clustering category based on the predicted distribution transformer low-voltage side load and the distributed power output, and then obtain the low-voltage active network line loss rate and the probability distribution of line loss at the next moment, and obtain the theoretical limit value of the low-voltage active network line loss rate and the fluctuation limit value of the line loss based on the predicted distribution transformer low-voltage side load through the comparison table of the distribution transformer low-voltage side load rate and the theoretical limit value of the low-voltage active network line loss rate, and the fluctuation limit value model of the low-voltage active network line loss; obtain the probability of line loss exceeding the limit at the next moment and issue a line loss warning based on the low-voltage active network line loss rate and the probability distribution of line loss, the theoretical limit value of the line loss rate, and the fluctuation limit value of the line loss at the next moment.

[0011] Furthermore, a fluctuation limit model of low voltage active network line loss is established by the following method:

[0012] ,

[0013] Where z is the fluctuation limit of the low voltage active network line loss, is the preset minimum fluctuation limit, for The proportionality coefficient between is the load on the low-voltage side of the distribution transformer at the current moment, The load on the low-voltage side of the distribution transformer in the previous moment;

[0014] The value of is selected in a graded adaptive way, that is, the amplitude change of the low-voltage side load of the distribution transformer at adjacent moments is divided into several grades from small to large, each grade represents a different amplitude change range of the low-voltage side load of the distribution transformer at adjacent moments, and each grade corresponds to different The greater the change in the amplitude of the load on the low-voltage side of the distribution transformer at adjacent moments, the greater the corresponding The smaller the value.

[0015] Further, obtaining the line loss limit-exceeding probability and performing line loss warning based on the line loss rate of the low-voltage active network at the next moment, the probability distribution of the line loss, the theoretical limit value of the line loss rate, and the fluctuation limit value of the line loss includes:

[0016] The first over-limit probability is obtained based on the probability distribution of the line loss of the low-voltage active network at the next moment and the theoretical limit value of the line loss rate;

[0017] The second over-limit probability is obtained based on the probability distribution of the low-voltage active network line loss at the next moment, the fluctuation limit of the line loss and the line loss value at the current moment;

[0018] A line loss rate threshold is set, and a third over-limit probability is obtained based on the probability distribution of the line loss rate of the low-voltage active network at the next moment and the line loss rate threshold;

[0019] The line loss exceeding limit probability at the next moment is obtained based on the first exceeding limit probability, the second exceeding limit probability and the third exceeding limit probability. If the line loss exceeding limit probability at the next moment exceeds the warning threshold, a line loss warning is issued.

[0020] Furthermore, the probability of the line loss exceeding the limit at the next moment is:

[0021] ,

[0022] in, is the probability of line loss exceeding the limit, is the first crossing probability, is the second crossing probability, is the third limit-crossing probability.

[0023] Further, a historical distribution transformer low-voltage side load curve within a period of time in the low-voltage active area is obtained, and the historical distribution transformer low-voltage side load curve is divided by hour to obtain a plurality of third samples; the third sample corresponding to the distribution transformer low-voltage side load at time t includes to The load on the low-voltage side of the distribution transformer at four moments, the load on the low-voltage side of the distribution transformer at moment t in the previous year, the load on the low-voltage side of the distribution transformer at moment t in the previous month, and the load on the low-voltage side of the distribution transformer at moment t in the previous week; constructing an autoregressive moving average model, and training the autoregressive moving average model based on the third sample to obtain a trained autoregressive moving average model;

[0024] The autoregressive moving average model is:

[0025] ,

[0026] in, is the predicted value of the load on the low-voltage side of the distribution transformer at time t, is the third sample corresponding to the load on the low-voltage side of the distribution transformer at time t, is a constant, , is a coefficient whose value is not 0, is the error term.

[0027] Further, the historical distributed power output curve, historical temperature data curve, and historical sunshine data curve of the low-voltage active area within a period of time are obtained; the historical distributed power output curve, historical temperature data curve, and historical sunshine data curve are divided by hours to obtain a number of second samples, and the i-th second sample includes the distributed power output, temperature data, and sunshine data of the i-th hour of the area, which is expressed as: ,in, is the distributed power output of the i-th second sample, is the temperature data of the i-th second sample, is the sunshine data of the i-th second sample; constructing a correlation equation, and optimizing the weight coefficient of the correlation equation based on the second sample to obtain a final correlation equation;

[0028] Obtain the low-voltage side load of the distribution transformer, the output of the distributed power source, the temperature data, and the sunshine data at the current moment and the previous n moments; input the low-voltage side load of the distribution transformer into the trained autoregressive moving average model to obtain the predicted low-voltage side load of the distribution transformer at the next moment, and input the distributed power source output, the temperature data, and the sunshine data into the final correlation equation to obtain the predicted distributed power source output at the next moment;

[0029] Based on the predicted low-voltage side load of the distribution transformer and the output of the distributed generation, the corresponding clustering category is obtained. Based on the probability distribution of the line loss rate of the clustering category, the probability distribution of the low-voltage active network line loss rate at the next moment is obtained, and then the probability distribution of the low-voltage active network line loss at the next moment is obtained.

[0030] Further, the historical distribution transformer low-voltage side load curve, the historical distributed power output curve, and the historical low-voltage active network line loss value curve of the low-voltage active area within a period of time are obtained; the historical distribution transformer load curve, the historical distributed power output curve, and the historical line loss value curve are divided by hours to obtain a number of first samples, and the i-th first sample includes the distribution transformer low-voltage side load, distributed power output, and low-voltage active network line loss value of the i-th hour of the area;

[0031] The Euclidean distance between the low-voltage side load of the distribution transformer and the output of the distributed power source of each two first samples is calculated as the Euclidean distance between the corresponding two first samples; when the Euclidean distance between the two first samples is less than a threshold, the two first samples are assigned to a cluster category through clustering; if two cluster categories contain the same first samples, the two cluster categories are merged to obtain a merged cluster category, otherwise, no merging is required; all cluster categories are obtained based on the merged cluster categories and the cluster categories that do not need to be merged, and the average value of all first samples in each cluster category is calculated to obtain the cluster center of each cluster category.

[0032] Furthermore, for each cluster: calculate the line loss rate corresponding to each first sample in the cluster and obtain the maximum line loss rate and the minimum line loss rate; set the interval threshold, and set the line loss rate interval based on the maximum line loss rate, the minimum line loss rate, and the interval threshold; count the number of first samples of the line loss rate corresponding to each line loss rate interval, so as to obtain the probability distribution of the line loss rate of the cluster category.

[0033] Furthermore, an approximate equivalent network of the low-voltage network in the substation is constructed based on the rated capacity of the low-voltage active substation distribution transformer and distributed power sources, the number of low-voltage users and the load of each low-voltage user; the power flow distribution in the approximate equivalent network is solved based on the power flow equation, and the load on the low-voltage side of the distribution transformer is obtained according to the power flow distribution result; and a theoretical limit value model of the line loss rate of the low-voltage active network is established based on the total load on the low-voltage side of the distribution transformer and the load of the low-voltage users.

[0034] Furthermore, when the load rate on the low-voltage side of the distribution transformer is When the theoretical limit value of the corresponding low-voltage active network line loss rate is calculated according to the following steps, a comparison table of the theoretical limit value of the low-voltage side load rate of the distribution transformer and the low-voltage active network line loss rate is obtained:

[0035] Set the total load of low-voltage users to the rated capacity of the distribution transformer. ;

[0036] Starting from the low-voltage user farthest from the distribution transformer, set the load of the low-voltage user to its rated upper limit from far to near until the total load of the low-voltage user reaches the rated capacity of the distribution transformer. ;

[0037] The approximate equivalent network distribution transformer low-voltage side load is obtained through the following steps S11 to S14;

[0038] Obtaining a corresponding theoretical limit value of the low-voltage active network line loss rate based on the theoretical limit value model of the low-voltage active network line loss rate;

[0039] Said include , , , , and sequentially obtain the theoretical limit values ​​of the low-voltage active network line loss rate corresponding to the above four distribution transformer low-voltage side load rates, and establish a comparison table of the theoretical limit values ​​of the distribution transformer low-voltage side load rate and the low-voltage active network line loss rate;

[0040] The approximate equivalent network distribution transformer low-voltage side load is obtained through the following steps:

[0041] S11, setting the initial active and reactive power of the balancing node, i.e. the starting node;

[0042] S12, based on the voltage amplitude, phase angle, active power and reactive power of the starting node, the voltage amplitude and phase angle of the adjacent next node are obtained through the power flow equation, until the voltage amplitude and phase angle of the end node are obtained;

[0043] S13, based on the active and reactive power of the end node, the voltage amplitude and phase angle of the end node obtained in step S12, the active and reactive power of the adjacent previous node is obtained, until the active and reactive power of the starting node is obtained;

[0044] S14, determine whether the termination condition is met, if so, obtain the low-voltage side load of the distribution transformer according to the active and reactive power of the current starting node; otherwise, repeat steps S12 to S14.

[0045] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:

[0046] 1. The present invention clusters the historical low-voltage side load of the distribution transformer, the historical output of distributed power sources, and the historical line loss values ​​of the low-voltage active network in the substation area to obtain the probability distribution of the line loss rate, and obtains the predicted low-voltage side load of the distribution transformer and the output of distributed power sources through an autoregressive moving average model and a correlation equation, and then obtains the low-voltage active network line loss rate and the probability distribution of line loss at the next moment, thereby improving the accuracy of line loss prediction, providing a basis for improving the accuracy of line loss warning, and thus providing strong technical support for the refined management and efficient operation of the low-voltage distribution network.

[0047] 2. The present invention obtains the theoretical limit value of the low-voltage active network line loss rate through equivalent network power flow analysis, and takes the theoretical limit of the low-voltage active network line loss into consideration when obtaining the line loss over-limit probability, so that the line loss warning is closer to the actual power grid, thereby improving the accuracy and effectiveness of the low-voltage active network line loss warning.

[0048] 3. The present invention obtains the fluctuation limit of the low-voltage active network line loss through the amplitude change of the low-voltage side load of the distribution transformer at adjacent moments in the substation area. The fluctuation limit of the low-voltage active network line loss is taken into account when obtaining the line loss over-limit probability, thereby realizing the monitoring of the fluctuation amplitude of the low-voltage active network line loss, thereby improving the accuracy and effectiveness of the low-voltage active network line loss early warning.

[0049] 4. The present invention obtains the probability of exceeding the line loss rate threshold through the probability distribution of the line loss rate of the low-voltage active network at the next moment, and considers the probability of exceeding the line loss rate threshold when obtaining the line loss over-limit probability, thereby improving the accuracy and effectiveness of the low-voltage active network line loss warning.

[0050] In the present invention, the above-mentioned technical solutions can also be combined with each other to achieve more preferred combination solutions. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can become obvious from the description, or can be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the contents particularly pointed out in the description and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings are only used for the purpose of illustrating specific embodiments and are not to be considered as limiting the present invention. In the entire drawings, the same reference symbols represent the same components;

[0052] Figure 1 The present invention is a flowchart of a low-voltage active network line loss early warning method considering theoretical limits and fluctuation amplitudes according to an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0054] A specific embodiment of the present invention discloses a low voltage active network line loss early warning method taking into account theoretical limits and fluctuation amplitudes. Figure 1 As shown, the method comprises the following steps:

[0055] Step S1, obtaining the rated capacity of the low-voltage active area distribution transformer and the distributed power source, and the load of the low-voltage user, using the equivalent network power flow analysis to establish the theoretical limit value model of the low-voltage active network line loss rate, and based on the equivalent network power flow analysis and the theoretical limit value model of the low-voltage active network line loss rate, obtaining a comparison table of the theoretical limit value of the distribution transformer low-voltage side load rate and the low-voltage active network line loss rate;

[0056] Step S2: establishing a fluctuation limit model of low-voltage active network line loss based on the amplitude change of the low-voltage side load of the distribution transformer at adjacent moments in the low-voltage active station area;

[0057] Step S3, obtaining the historical low-voltage side load of the distribution transformer, the historical distributed power output, and the historical line loss values ​​of the low-voltage active area, and clustering them to obtain multiple cluster categories, and statistically analyzing the line loss rate of the samples in each cluster category to obtain the probability distribution of the line loss rate of the cluster category;

[0058] Step S4, constructing a correlation equation and an autoregressive moving average model, obtaining a predicted distributed power output based on the correlation equation, and obtaining a predicted distribution transformer low-voltage side load based on the autoregressive moving average model; obtaining corresponding clustering categories based on the predicted distribution transformer low-voltage side load and distributed power output, and then obtaining the low-voltage active network line loss rate and the probability distribution of line loss at the next moment, and obtaining the theoretical limit value of the low-voltage active network line loss rate and the fluctuation limit value of the line loss at the next moment based on the predicted distribution transformer low-voltage side load through the comparison table of the distribution transformer low-voltage side load rate and the theoretical limit value of the low-voltage active network line loss rate, and the fluctuation limit value model of the low-voltage active network line loss; obtaining the probability of line loss exceeding the limit at the next moment and performing a line loss warning based on the low-voltage active network line loss rate and the probability distribution of line loss, the theoretical limit value of the line loss rate, and the fluctuation limit value of the line loss at the next moment.

[0059] Specifically, in step S1, the distribution transformer is a distribution transformer; the inventory data of the low-voltage active substation equipment is obtained through the production management system, and the rated capacity of the distribution transformer and distributed power source, the number of low-voltage users and the load of each low-voltage user are obtained based on the inventory data of the equipment.

[0060] Furthermore, an approximate equivalent network of the low-voltage network in the substation is constructed based on the rated capacity of the low-voltage active substation distribution transformer and distributed power sources, the number of low-voltage users and the load of each low-voltage user; the power flow distribution in the approximate equivalent network is solved based on the power flow equation, and the load on the low-voltage side of the distribution transformer is obtained according to the power flow distribution result; and a theoretical limit value model of the line loss rate of the low-voltage active network is established based on the total load on the low-voltage side of the distribution transformer and the load of the low-voltage users.

[0061] Specifically, the approximate equivalent network of the low-voltage network in the area is constructed based on the rated capacity of the low-voltage active area distribution transformer and the distributed power source, the number of low-voltage users, and the load of each low-voltage user, including:

[0062] Construct a low-voltage single-phase line as an approximate equivalent network of the low-voltage network in the substation, and the length of the approximate equivalent network is 1 / 3 of the total length of the low-voltage line in the low-voltage active network; assume that low-voltage users are evenly distributed on the approximate equivalent network, the number of low-voltage users in the approximate equivalent network is 1 / 3 of the number of low-voltage users in the low-voltage active substation, and the rated capacity of the approximate equivalent network line is 1 / 3 of the rated capacity of the distribution transformer in the low-voltage active substation; assume that the distributed power source in the approximate equivalent network is located at 1 / 2 of the total length of its line, and the capacity of its distributed power source is 1 / 3 of the rated capacity of the distributed power source in the low-voltage active substation.

[0063] It should be noted that the line loss of the low-voltage active network in this application refers to the line loss of the low-voltage network in the substation, and does not include the loss of the substation distribution transformer itself. For the low-voltage network, the low-voltage side of the substation distribution transformer is the power supply point, and each low-voltage user is a load point. The line loss of the low-voltage active network is obtained by subtracting the sum of the loads of the load points from the power supply point output. In this application, the low-voltage network line is divided into three completely identical single-phase lines according to the principle of uniform distribution. Therefore, the length of each single-phase line is 1 / 3 of the total length of the low-voltage line, and the rated capacity of each single-phase line is 1 / 3 of the rated capacity of the low-voltage active substation distribution transformer. Since the approximate equivalent network is constructed for the substation low-voltage network, and the distribution transformer is a device that transforms from medium voltage to low voltage, the equivalent network does not need to include a distribution transformer, but the low-voltage side of the distribution transformer needs to be used as a balancing node. The load of the low-voltage users in the approximately equivalent network is set according to the load of the low-voltage users in the low-voltage active area, so that the total load of the low-voltage users in the approximately equivalent network is 1 / 3 of the total load of the low-voltage users in the low-voltage active area.

[0064] Furthermore, the power flow equation is:

[0065] ,

[0066] ,

[0067] ,

[0068] in, is the active power of node i, is the reactive power of node i, is the active output of the distributed generation at node i, is the reactive power output of the distributed generation at node i, is the voltage amplitude at node i, is the voltage amplitude at node j, is the real part of the admittance matrix element between node i and node j, is the imaginary part of the admittance matrix element between node i and node j, is the phase angle difference between node i and node j, node i and node j are adjacent nodes, is the lower limit of the node voltage, is the upper limit of the node voltage.

[0069] It should be noted that the above voltage upper and lower limits need to be set with reference to the rated voltage.

[0070] Specifically, the nodes in the equivalent network include load nodes and balance nodes. Each low-voltage user in the approximate equivalent network is taken as a load node, and the low-voltage side of the distribution transformer in the substation is taken as a balance node; the balance node is taken as the starting node, and its voltage amplitude is limited to 1.0 and the phase angle is limited to 0; the load node farthest from the balance node is taken as the end node. The load on the low-voltage side of the distribution transformer in the approximate equivalent network is obtained through the following steps:

[0071] S11, setting the initial active and reactive power of the balancing node, i.e. the starting node;

[0072] S12, based on the voltage amplitude, phase angle, active power and reactive power of the starting node, the voltage amplitude and phase angle of the adjacent next node are obtained through the power flow equation, until the voltage amplitude and phase angle of the end node are obtained;

[0073] S13, based on the active and reactive power of the end node, the voltage amplitude and phase angle of the end node obtained in step S12, the active and reactive power of the adjacent previous node is obtained, until the active and reactive power of the starting node is obtained;

[0074] S14, determine whether the termination condition is met, if so, obtain the low-voltage side load of the distribution transformer according to the active and reactive power of the current starting node; otherwise, repeat steps S12 to S14.

[0075] Specifically, in step S11, the initial active power of the starting node is set to the sum of the active power of each load node, and the initial reactive power is set to the sum of the reactive power of each load node. In step S12, the voltage amplitude of the starting node is limited to 1.0, and the phase angle is limited to 0, and the voltage amplitude and phase angle of each node are calculated in sequence from the starting node to the end node through the power flow equation. In step S13, the active and reactive power of the end node are obtained based on the load of the end node, and the active and reactive power of each node are calculated in sequence from the end node to the starting node through the power flow equation. In step S14, the termination condition is that the active and reactive power deviation of each load node is less than the threshold, and the sum of the active and reactive power of the current starting node is the load on the low-voltage side of the distribution transformer.

[0076] It can be understood that each low-voltage user is regarded as a load node, so the load of the low-voltage user is the load of the load node. Based on the load of the load node, the active and reactive power of the load node and the active and reactive output of the distributed power source at the load node can be obtained; similarly, based on the load of the balancing node, the active and reactive power of the balancing node and the active and reactive output of the distributed power source at the balancing node can be obtained. The load of the low-voltage user and the load on the low-voltage side of the distribution transformer in the substation can be obtained through the ledger data of the production management system, but the obtained load on the low-voltage side of the distribution transformer is only the load level actually reached on the low-voltage side of the distribution transformer, rather than the theoretical limit value that it can reach. The present invention obtains the theoretical limit value of the load on the low-voltage side of the distribution transformer by calculating the power flow of the approximate equivalent network, and then obtains the theoretical limit value of the line loss rate of the low-voltage active network, which provides a basis for subsequently improving the accuracy of line loss warning.

[0077] Furthermore, the theoretical limit value model of the low voltage active network line loss rate is:

[0078] ,

[0079] in, is the theoretical limit value of the line loss rate of the low voltage active network, It is the load on the low voltage side of the approximately equivalent network distribution transformer. It is the total load of low-voltage users in the approximately equivalent network.

[0080] Furthermore, when the load rate on the low-voltage side of the distribution transformer is , calculate the corresponding theoretical limit value of low-voltage active network line loss rate according to the following steps:

[0081] Set the total load of low-voltage users to the rated capacity of the distribution transformer. ;

[0082] Starting from the low-voltage user farthest from the distribution transformer, set the load of the low-voltage user to its rated upper limit from far to near until the total load of the low-voltage user reaches the rated capacity of the distribution transformer. ;

[0083] Through the above steps S11 to S14, an approximate equivalent network distribution transformer low-voltage side load is obtained;

[0084] Based on the theoretical limit value model of the low-voltage active network line loss rate, the corresponding theoretical limit value of the low-voltage active network line loss rate is obtained.

[0085] Furthermore, the include , , , , and the theoretical limit values ​​of the low-voltage active network line loss rate corresponding to the above four distribution transformer low-voltage side load rates are obtained in turn, and a comparison table of the theoretical limit values ​​of the distribution transformer low-voltage side load rate and the low-voltage active network line loss rate is established.

[0086] It should be noted that the theoretical limit value of the low-voltage active network line loss rate refers to the theoretical maximum value of the low-voltage network line loss rate in the substation area; the theoretical limit value of the low-voltage active network line loss rate is related to the load level on the low-voltage side of the distribution transformer. , the corresponding theoretical limit value of the line loss rate is also different. Therefore, the present invention solves the corresponding theoretical limit value of the low-voltage active network line loss rate based on the equivalent network power flow analysis and the theoretical limit value model of the low-voltage active network line loss rate at different distribution transformer low-voltage side load rate levels, thereby obtaining a comparison table of the theoretical limit values ​​of the distribution transformer low-voltage side load rate and the low-voltage active network line loss rate. The line loss mentioned in the present invention refers to the low-voltage active network line loss, and the line loss rate mentioned refers to the low-voltage active network line loss rate.

[0087] For example, assuming that the rated capacity of a low-voltage active distribution transformer in a certain area is 100kVA, the rated capacity of the distributed power source is 20kW, the total length of the low-voltage line is 3km, and the number of low-voltage users is 114, the approximate equivalent network constructed by the low-voltage distribution network topology structure is: a 1km long low-voltage line with a rated capacity of 33kVA, 38 low-voltage users and evenly distributed, with a low-voltage load point every 26m, and a distributed power source with a capacity of 6.7kW connected at 0.5km. The power flow distribution in the above approximate equivalent network is solved based on the power flow equation, and the load on the low-voltage side of the distribution transformer is obtained according to the power flow distribution result. When the load rate on the low-voltage side of the distribution transformer in the area is respectively , , , When the load rate on the low-voltage side of the distribution transformer is , the comparison table of the theoretical limit values ​​of the line loss rate of the low-voltage active network is shown in Table 1:

[0088] surface Comparison table of theoretical limit values ​​of load rate on the low-voltage side of distribution transformer and line loss rate of low-voltage active network at different load rate levels in a low-voltage active station area

[0089]

[0090] Specifically, in step S2, a fluctuation limit model of low voltage active network line loss is established by the following method:

[0091] ,

[0092] Where z is the fluctuation limit of the low voltage active network line loss, is the preset minimum fluctuation limit, for The proportionality coefficient between is the load on the low-voltage side of the distribution transformer at the current moment, It is the load on the low voltage side of the distribution transformer in the previous moment.

[0093] It should be noted that, except for the extreme conditions, the fluctuation limit of the line loss needs to satisfy the requirement that the amplitude change of the line loss of the low-voltage active network at adjacent moments does not exceed the amplitude change of the load on the low-voltage side of the distribution transformer corresponding to the adjacent moments. However, for the extreme conditions, that is, when the amplitude change of the load on the low-voltage side of the distribution transformer at adjacent moments is 0, the amplitude change of the line loss corresponding to the adjacent moments is often not 0 due to the actual situation. At this time, if the fluctuation limit of the line loss corresponding to the adjacent moments is set to 0, the line loss warning will inevitably be triggered; therefore, this application presets , thus avoiding the above situation. The value of is selected in a graded adaptive way, that is, the amplitude change of the low-voltage side load of the distribution transformer at adjacent moments is divided into several grades from small to large, each grade represents a different amplitude change range of the low-voltage side load of the distribution transformer at adjacent moments, and each grade corresponds to different The greater the change in the amplitude of the load on the low-voltage side of the distribution transformer at adjacent moments, the greater the corresponding The smaller the value.

[0094] For example, the amplitude of the low-voltage side load of a certain low-voltage active station area for five consecutive moments is shown in Table 2:

[0095] surface The amplitude of the load on the low-voltage side of a distribution transformer in a low-voltage active area for five consecutive moments

[0096]

[0097] set up ; When the amplitude change of the load on the low-voltage side of the distribution transformer at adjacent moments is less than 0.1kW, set , when the amplitude change of the load on the low-voltage side of the distribution transformer at adjacent moments is greater than or equal to 0.1kW and less than 0.5kW, set , otherwise, set ; The fluctuation limit of the line loss of the low-voltage active network in the low-voltage active area at the corresponding time can be obtained as shown in Table 3:

[0098] surface Fluctuation limit of line loss of a low voltage active network for 5 consecutive moments

[0099]

[0100] Specifically, in step S3, the historical distribution transformer low-voltage side load curve, historical distributed power output curve, and historical low-voltage active network line loss value curve of the low-voltage active substation over a period of time are obtained through the production management system; the historical distribution transformer load curve, historical distributed power output curve, and historical line loss value curve are divided by hours to obtain a number of first samples, and the i-th first sample includes the distribution transformer low-voltage side load, distributed power output, and low-voltage active network line loss value of the i-th hour in the substation.

[0101] Exemplarily, the historical distribution transformer low-voltage side load curve, the historical distributed power output curve, and the historical low-voltage active network line loss value curve within 90 days of the low-voltage active area are obtained; the historical distribution transformer low-voltage side load curve, the historical distributed power output curve, and the historical low-voltage active network line loss value curve are divided by hour to obtain 2160 first samples, and the i-th first sample is expressed as: ,in, is the load on the low-voltage side of the distribution transformer at the ith first sample (i.e. the ith hour in the distribution area), is the distributed generation output of the ith first sample (i.e. the ith hour in the area), It is the line loss value of the low voltage active network of the ith first sample (i.e. the ith hour of the substation).

[0102] It should be noted that the production management system samples the load on the low-voltage side of the distribution transformer, the output of distributed power sources, and the line loss values ​​of the low-voltage active network on an hourly basis. Therefore, the load on the low-voltage side of the distribution transformer at a certain moment is the load on the low-voltage side of the distribution transformer for that hour, the output of the distributed power source at a certain moment is the output of the distributed power source for that hour, and the line loss value of the low-voltage active network at a certain moment is the line loss value of the low-voltage active network for that hour.

[0103] Furthermore, the Euclidean distance between the low-voltage side load of the distribution transformer and the output of the distributed power source of each two first samples is calculated as the Euclidean distance between the corresponding two first samples; when the Euclidean distance between the two first samples is less than a threshold, the two first samples are assigned to a cluster category through clustering; if two cluster categories contain the same first samples, the two cluster categories are merged to obtain a merged cluster category, otherwise, no merging is required; all cluster categories are obtained based on the merged cluster categories and the cluster categories that do not need to be merged, and the average value of all first samples in each cluster category is calculated to obtain the cluster center of each cluster category.

[0104] Specifically, the Euclidean distance between the two first samples is:

[0105] ,

[0106] in, is the Euclidean distance between the i-th first sample and the j-th first sample.

[0107] Exemplarily, assuming that 8 first samples are obtained, the Euclidean distances between the 8 first samples are calculated to obtain 28 Euclidean distances. Assuming that the Euclidean distances between sample 1 and sample 2, sample 2 and sample 5, sample 3 and sample 4, sample 4 and sample 6, and sample 7 and sample 8 in the 28 Euclidean distances are less than a threshold, sample 1 and sample 2 are assigned to cluster 1, sample 2 and sample 5 are assigned to cluster 2, sample 3 and sample 4 are assigned to cluster 3, sample 4 and sample 6 are assigned to cluster 4, and sample 7 and sample 8 are assigned to cluster 5. Since cluster 1 and cluster 2 contain the same sample 2, and cluster 3 and cluster 4 contain the same sample 4, cluster 1 and cluster 2 are merged to obtain merged cluster 1, which contains sample 1, sample 2 and sample 5. Cluster 3 and cluster 4 are merged to obtain merged cluster 3, which contains sample 3, sample 4 and sample 6. All cluster categories are obtained based on merged cluster 1, merged cluster 3 and cluster 5 that does not need to be merged. The mean value of all first samples in each cluster category is calculated to obtain the cluster center of each cluster category. For example, the mean value of sample 1, sample 2 and sample 5 is calculated to obtain the cluster center of merged cluster 1.

[0108] It should be noted that calculating the mean of all first samples in a certain clustering category refers to calculating the mean of the low-voltage side load of the distribution transformer, the mean of the output of the distributed power source, and the mean of the low-voltage active network line loss values ​​of all the first samples in the clustering category respectively. Therefore, the cluster center of the clustering category contains the mean of the low-voltage side load of the distribution transformer, the mean of the output of the distributed power source, and the mean of the low-voltage active network line loss values ​​of all its first samples.

[0109] Furthermore, for each cluster: calculate the line loss rate corresponding to each first sample in the cluster and obtain the maximum line loss rate and the minimum line loss rate; set the interval threshold, and set the line loss rate interval based on the maximum line loss rate, the minimum line loss rate, and the interval threshold; count the number of first samples of the line loss rate corresponding to each line loss rate interval, so as to obtain the probability distribution of the line loss rate of the cluster category.

[0110] Specifically, the line loss rate corresponding to each first sample is the ratio of the line loss value of the low-voltage active network of the first sample to the load on the low-voltage side of the distribution transformer of the first sample. For each cluster category: set a fixed interval threshold, obtain the upper limit of the line loss rate based on the maximum line loss rate, obtain the lower limit of the line loss rate based on the minimum line loss rate, and set several line loss rate intervals based on the fixed interval threshold, the upper limit of the line loss rate, and the lower limit of the line loss rate; count the number of line loss rates contained in each line loss rate interval, that is, the number of first samples, calculate the ratio of the number of line loss rates contained in each line loss rate interval to the number of samples in the cluster to obtain the probability of the line loss rate corresponding to each line loss rate interval, thereby obtaining the probability distribution of the line loss rate of the cluster category.

[0111] Exemplarily, the line loss rate corresponding to each first sample in a certain cluster category of a certain low-voltage active station area is shown in Table 4:

[0112] surface Line loss rate corresponding to each first sample in a cluster category of a low-voltage active station area

[0113]

[0114] As shown in Table 4, the maximum line loss rate of this cluster is 2.96%, the minimum line loss rate is 2.52%, and the number of samples is 20. The fixed interval threshold is set to 0.1%, and the upper limit of the line loss rate is 3% based on the maximum line loss rate of 2.96%, and the lower limit of the line loss rate is 2.5% based on the minimum line loss rate of 2.52%. Based on 0.1%, 3%, and 2.5%, the line loss rate intervals are set to [2.5, 2.6), [2.6, 2.7), [2.7, 2.8), [2.8, 2.9), and [2.9, 3); the number of line loss rates included in each line loss rate interval is counted, for example, the number of line loss rates included in the line loss rate interval [2.5, 2.6) is 2; the probability of the line loss rate corresponding to each line loss rate interval is calculated by calculating the ratio of the number of line loss rates included in the line loss rate interval [2.5, 2.6) to the number of first samples in the cluster category 20, for example, the probability of the line loss rate corresponding to [2.5, 2.6) is 0.1, so that the probability distribution of the line loss rate of the cluster category is shown in Table 5:

[0115] surface Probability distribution of line loss rate of a certain cluster category in a low voltage active area

[0116]

[0117] Specifically, in step S4, the historical distributed power output curve, historical temperature data curve, and historical sunshine data curve of the low-voltage active area within a period of time are obtained; the historical distributed power output curve, historical temperature data curve, and historical sunshine data curve are divided by hours to obtain a number of second samples, and the i-th second sample includes the distributed power output, temperature data, and sunshine data of the i-th hour of the area, which is expressed as: ,in, is the distributed power output of the i-th second sample, is the temperature data of the i-th second sample, is the sunshine data of the i-th second sample; constructing a correlation equation, and optimizing the weight coefficient of the correlation equation based on the second sample to obtain a final correlation equation.

[0118] Furthermore, the correlation equation is:

[0119] ,

[0120] in, To provide distributed power generation, is the temperature data, is the sunshine data, , , is the weight coefficient.

[0121] Specifically, optimizing the weight coefficient of the association equation based on the second sample to obtain the final association equation includes:

[0122] Constructing a weight coefficient vector and setting an initial weight coefficient, constructing an input matrix based on the temperature data and sunshine data in the second sample, and constructing an actual value output vector based on the output of the distributed power source in the second sample;

[0123] The predicted value of the distributed power output is obtained by multiplying the input matrix with the weight coefficient vector to construct a predicted value output vector, and the square sum of the difference between the actual value and the predicted value of the distributed power output is obtained based on the actual value output vector and the predicted value output vector to establish an objective function;

[0124] Based on the objective function, a partial derivative of each weight coefficient is calculated to establish a system of equations, and the system of equations is solved to obtain the optimized weight coefficients to obtain the final correlation equation.

[0125] Specifically, the weight coefficient vector is:

[0126] ;

[0127] The input matrix is:

[0128] ,

[0129] Where m is the number of the second sample, is the temperature data of the i-th second sample, is the sunshine data of the i-th second sample;

[0130] The actual value output vector is:

[0131] ,

[0132] in, is the distributed power output of the i-th second sample.

[0133] Specifically, the predicted value output vector is:

[0134] ,

[0135] ,

[0136] in, is the predicted value of the distributed power output corresponding to the i-th second sample;

[0137] The objective function is:

[0138] .

[0139] Specifically, based on the objective function, the partial derivative of each weight coefficient is taken to establish the equation system:

[0140] ,

[0141] in, Represents the element in the i-th row and j-th column of the input matrix;

[0142] Convert the above equations into matrix form:

[0143] ,

[0144] in, represent The transposed matrix of

[0145] The above equation is solved by Gaussian elimination method or LU decomposition to obtain the optimized weight coefficient, and then the final correlation equation is obtained based on the optimized weight coefficient.

[0146] Further, a historical distribution transformer low-voltage side load curve within a period of time in the low-voltage active area is obtained, and the historical distribution transformer low-voltage side load curve is divided by hour to obtain a plurality of third samples; the third sample corresponding to the distribution transformer low-voltage side load at time t includes to The load on the low-voltage side of the distribution transformer at four moments, the load on the low-voltage side of the distribution transformer at moment t in the previous year, the load on the low-voltage side of the distribution transformer at moment t in the previous month, and the load on the low-voltage side of the distribution transformer at moment t in the previous week; construct an autoregressive moving average model, and train the autoregressive moving average model based on the third sample to obtain a trained autoregressive moving average model.

[0147] Furthermore, the autoregressive moving average model is:

[0148] ,

[0149] in, is the predicted value of the load on the low-voltage side of the distribution transformer at time t, is the third sample corresponding to the load on the low-voltage side of the distribution transformer at time t, is a constant, , is a coefficient whose value is not 0, is the error term.

[0150] It should be noted that, from the description of the third sample above, it can be seen that in the autoregressive moving average model to for to The load on the low-voltage side of the transformer at four moments: is the load on the low-voltage side of the distribution transformer at time t in the previous year, is the load on the low-voltage side of the distribution transformer at time t in the previous month, It is the load on the low-voltage side of the distribution transformer at time t of the previous week.

[0151] Furthermore, the third sample is input into the autoregressive moving average model for training to obtain the predicted distribution transformer low-voltage side load and distributed power output, and the root mean square error RMSE between the distribution transformer low-voltage side load and the distributed power output sample value and the predicted value is calculated. When the RMSE of the distribution transformer low-voltage side load and the distributed power output are both less than the error threshold, the trained autoregressive moving average model is obtained.

[0152] Furthermore, the low-voltage side load of the distribution transformer, the output of distributed power sources, temperature data, and sunshine data of the current moment and the previous n moments are obtained; the low-voltage side load of the distribution transformer is input into the trained autoregressive moving average model to obtain the predicted low-voltage side load of the distribution transformer at the next moment, and the distributed power source output, temperature data, and sunshine data are input into the final correlation equation to obtain the predicted distributed power source output at the next moment.

[0153] Furthermore, the corresponding clustering category is obtained based on the predicted low-voltage side load of the distribution transformer and the output of the distributed power source, and the probability distribution of the low-voltage active network line loss rate at the next moment is obtained based on the probability distribution of the line loss rate of the clustering category, and then the probability distribution of the low-voltage active network line loss at the next moment is obtained.

[0154] Specifically, the Euclidean distance between the predicted low-voltage side load of the distribution transformer and the output of the distributed power source and the cluster centers of all cluster categories is calculated, and the corresponding cluster category is obtained based on the minimum Euclidean distance, thereby obtaining the probability distribution of the corresponding cluster category line loss rate, and the probability distribution of the corresponding cluster category line loss rate is used as the probability distribution of the low-voltage active network line loss rate at the next moment, and the probability distribution of the low-voltage active network line loss at the next moment is obtained based on the predicted low-voltage side load of the distribution transformer and the probability distribution of the low-voltage active network line loss rate at the next moment.

[0155] For example, assuming that the load on the low-voltage side of the distribution transformer of a low-voltage active substation at the current moment and the previous three moments, as well as the load on the low-voltage side of the distribution transformer at the current moment in the previous year, month and week are [48.24kW, 52.43kW, 58.68kW, 60.72kW, 62.54kW, 46.32kW, 49.87kW, 47.64kW], the above-mentioned load on the low-voltage side of the distribution transformer is input into the trained autoregressive moving average model, and the predicted load on the low-voltage side of the distribution transformer at the next moment is 48.83kW. Assuming that the distributed power output of a low-voltage active area at the current moment and the previous four moments is [12.75kW, 12.35kW, 13.69kW, 16.44kW, 20.38kW], the temperature data is [16 degrees, 16 degrees, 17 degrees, 18 degrees, 19 degrees], and the sunshine data is [1, 1, 1, 1, 1] (in the sunshine data, 1 represents sunshine and 0 represents no sunshine), the above distributed power output, temperature data, and sunshine data are input into the final correlation equation to obtain the predicted distributed power output of 17.94kW at the next moment. Assuming that the predicted distribution transformer low-voltage side load and the probability distribution of the clustering category line loss rate corresponding to the distributed power output are shown in Table 5, since the line loss is the product of the distribution transformer low-voltage side load and the line loss rate, the upper and lower limits of the line loss interval can be obtained based on the product of 48.83kW and the upper and lower limits of the line loss rate interval, and then the low-voltage active network line loss rate and the probability distribution of line loss at the next moment are obtained as shown in Table 6:

[0156] surface The line loss rate and probability distribution of the low-voltage active network in a low-voltage active area at the next moment

[0157]

[0158] It can be understood that the present invention obtains the probability distribution of line loss rate by clustering the historical low-voltage side load of the distribution transformer, the historical distributed power output, and the historical low-voltage active network line loss values ​​of the substation, and obtains the predicted low-voltage side load of the distribution transformer and the distributed power output through the autoregressive moving average model and the correlation equation, and then obtains the low-voltage active network line loss rate and the probability distribution of line loss at the next moment, which improves the accuracy of line loss prediction, provides a basis for improving the accuracy of line loss warning, and thus provides strong technical support for the refined management and efficient operation of the low-voltage distribution network.

[0159] Specifically, according to the predicted load on the low-voltage side of the distribution transformer and the comparison table of the theoretical limit values ​​of the load rate on the low-voltage side of the distribution transformer and the line loss rate of the low-voltage active network, the theoretical limit value of the low-voltage active network line loss rate at the next moment is obtained by the linear difference method; the predicted load on the low-voltage side of the distribution transformer and the load on the low-voltage side of the distribution transformer at the current moment are substituted into the fluctuation limit model of the low-voltage active network line loss to obtain the fluctuation limit of the low-voltage active network line loss at the next moment.

[0160] Specifically, the corresponding load rate is obtained according to the predicted load on the low-voltage side of the distribution transformer, and the theoretical limit value of the low-voltage active network line loss rate at the next moment can be obtained based on the comparison table of the theoretical limit values ​​of the load rate on the low-voltage side of the distribution transformer and the line loss rate of the low-voltage active network by using existing linear interpolation methods, such as piecewise interpolation, Lagrange interpolation, polynomial interpolation, and tri-spline interpolation (not limited in the present invention). It can be understood that the ratio of the predicted load on the low-voltage side of the distribution transformer to the rated capacity of the distribution transformer is the corresponding load rate.

[0161] For example, according to the predicted low-voltage side load of 48.83kW and the rated capacity of 100kVA of the distribution transformer in the substation, the corresponding load rate is 48.83%. Through the linear difference method, based on Table 1, the theoretical limit value of the low-voltage active network line loss rate at the next moment is obtained to be 2.946kW; the predicted low-voltage side load of 48.83kW and the current low-voltage side load of 48.24kW are substituted into the fluctuation limit model of the low-voltage active network line loss, and the fluctuation limit of the low-voltage active network line loss at the next moment is obtained to be 0.1kW.

[0162] Further, obtaining the line loss limit-exceeding probability and performing line loss warning based on the line loss rate of the low-voltage active network at the next moment, the probability distribution of the line loss, the theoretical limit value of the line loss rate, and the fluctuation limit value of the line loss includes:

[0163] The first over-limit probability is obtained based on the probability distribution of the line loss of the low-voltage active network at the next moment and the theoretical limit value of the line loss rate;

[0164] The second over-limit probability is obtained based on the probability distribution of the low-voltage active network line loss at the next moment, the fluctuation limit of the line loss and the line loss value at the current moment;

[0165] A line loss rate threshold is set, and a third over-limit probability is obtained based on the probability distribution of the line loss rate of the low-voltage active network at the next moment and the line loss rate threshold;

[0166] The line loss exceeding limit probability at the next moment is obtained based on the first exceeding limit probability, the second exceeding limit probability and the third exceeding limit probability. If the line loss exceeding limit probability at the next moment exceeds the warning threshold, a line loss warning is issued.

[0167] Specifically, based on the probability distribution of the low-voltage active network line loss at the next moment, it is determined whether the lower limit of the line loss interval exceeds the theoretical limit value of the line loss rate at the next moment, and the probabilities corresponding to the lower limits of the line loss interval that exceed the theoretical limit value of the line loss rate at the next moment are accumulated to obtain a first over-limit probability; the line loss value at the current moment is added to the fluctuation limit value of the low-voltage active network line loss at the next moment to obtain a line loss fluctuation threshold, and based on the probability distribution of the low-voltage active network line loss at the next moment, it is determined whether the lower limit of the line loss interval exceeds the line loss fluctuation threshold, and the probabilities corresponding to the lower limits of the line loss interval that exceed the line loss fluctuation threshold are accumulated to obtain a second over-limit probability; a line loss rate threshold is set, and based on the probability distribution of the low-voltage active network line loss rate at the next moment, it is determined whether the lower limit of the line loss rate interval exceeds the line loss rate threshold, and the probabilities corresponding to the lower limits of the line loss rate interval that exceed the line loss rate threshold are accumulated to obtain a third over-limit probability.

[0168] It can be understood that the present invention obtains the theoretical limit value of the low-voltage active network line loss rate through equivalent network power flow analysis, and considers the theoretical limit of the low-voltage active network line loss when obtaining the probability of line loss exceeding the limit, so that the line loss warning is closer to the actual power grid, thereby improving the accuracy and effectiveness of the low-voltage active network line loss warning. The present invention obtains the fluctuation limit of the low-voltage active network line loss through the amplitude change of the low-voltage side load of the distribution transformer at adjacent moments in the substation area, and considers the fluctuation limit of the low-voltage active network line loss when obtaining the probability of line loss exceeding the limit, realizing the monitoring of the fluctuation amplitude of the low-voltage active network line loss, thereby improving the accuracy and effectiveness of the low-voltage active network line loss warning. The present invention obtains the probability of exceeding the line loss rate threshold through the probability distribution of the low-voltage active network line loss rate at the next moment, and considers the probability of exceeding the line loss rate threshold when obtaining the probability of line loss exceeding the limit, thereby improving the accuracy and effectiveness of the low-voltage active network line loss warning.

[0169] Furthermore, the probability of the line loss exceeding the limit at the next moment is:

[0170] ,

[0171] in, is the probability of line loss exceeding the limit, is the first crossing probability, is the second crossing probability, is the third limit-crossing probability.

[0172] Exemplarily, the probability distribution of the line loss rate and line loss of the low-voltage active network at the next moment is shown in Table 6. It can be seen from Table 6 that no line loss interval lower limit exceeds the theoretical limit value of the line loss rate at the next moment, 2.946kW, so the first over-limit probability is 0. Assuming that the line loss value at the current moment is 1.298kW, 1.298kW is added to the fluctuation limit of the low-voltage active network line loss at the next moment, 0.1kW, to obtain a line loss fluctuation threshold of 1.398kW. It can be seen from Table 6 that only the lower limit of the line loss interval [1.416, 1.465) exceeds the line loss fluctuation threshold of 1.398kW, and the probability corresponding to the lower limit of the line loss interval 1.416 is 0.1, so the second over-limit probability is 0.1. Set the line loss rate threshold to 3kW. It can be seen from Table 6 that no line loss rate interval lower limit exceeds the line loss rate threshold of 3kW, so the third over-limit probability is 0. Based on the above formula for the line loss over-limit probability at the next moment, the line loss over-limit probability at the next moment is 0.1.

[0173] Compared with the prior art, the low-voltage active network line loss early warning method considering the theoretical limit and fluctuation amplitude provided by the present invention has the following beneficial effects:

[0174] 1. The present invention clusters the historical low-voltage side load of the distribution transformer, the historical output of distributed power sources, and the historical line loss values ​​of the low-voltage active network in the substation area to obtain the probability distribution of the line loss rate, and obtains the predicted low-voltage side load of the distribution transformer and the output of distributed power sources through the autoregressive moving average model and the correlation equation, and then obtains the low-voltage active network line loss rate and the probability distribution of line loss at the next moment, thereby improving the accuracy of line loss prediction, providing a basis for improving the accuracy of line loss warning, and thus providing strong technical support for the refined management and efficient operation of the low-voltage distribution network.

[0175] 2. The present invention obtains the theoretical limit value of the low-voltage active network line loss rate through equivalent network power flow analysis, and takes into account the theoretical limit of the low-voltage active network line loss when obtaining the line loss over-limit probability, so that the line loss warning is closer to the actual power grid, thereby improving the accuracy and effectiveness of the low-voltage active network line loss warning.

[0176] 3. The present invention obtains the fluctuation limit of the low-voltage active network line loss through the amplitude change of the low-voltage side load of the distribution transformer at adjacent moments in the substation area. The fluctuation limit of the low-voltage active network line loss is taken into account when obtaining the line loss over-limit probability, thereby realizing the monitoring of the fluctuation amplitude of the low-voltage active network line loss, thereby improving the accuracy and effectiveness of the low-voltage active network line loss early warning.

[0177] 4. The present invention obtains the probability of exceeding the line loss rate threshold through the probability distribution of the line loss rate of the low-voltage active network at the next moment. When obtaining the line loss over-limit probability, the probability of exceeding the line loss rate threshold is taken into account, thereby improving the accuracy and effectiveness of the low-voltage active network line loss warning.

[0178] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0179] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A low voltage active network line loss early warning method considering theoretical limits and fluctuation amplitudes, characterized in that: The method comprises the following steps: Based on the rated capacity of the low-voltage active area distribution transformer and distributed power sources, the number of low-voltage users and the load of each low-voltage user, an approximate equivalent network of the area low-voltage network is constructed; based on the power flow equation, the power flow distribution in the approximate equivalent network is solved, and the load on the low-voltage side of the distribution transformer is obtained according to the power flow distribution result; based on the total load on the low-voltage side of the distribution transformer and the load of the low-voltage users, a theoretical limit value model of the line loss rate of the low-voltage active network is established; based on the equivalent network power flow analysis and the theoretical limit value model of the line loss rate of the low-voltage active network, a comparison table of the load rate on the low-voltage side of the distribution transformer and the theoretical limit value of the line loss rate of the low-voltage active network is obtained; Based on the amplitude variation of the load on the low-voltage side of the distribution transformer at adjacent moments in the low-voltage active area, a fluctuation limit model of the line loss of the low-voltage active network is established; Obtain the historical low-voltage side load of the distribution transformer in the low-voltage active area, the historical output of distributed power sources, and the historical line loss values, and cluster them to obtain multiple cluster categories. The line loss rate of the samples in each cluster category is statistically analyzed to obtain the probability distribution of the line loss rate of the cluster category. Construct a correlation equation and an autoregressive moving average model to obtain the distribution transformer low-voltage side load, distributed power output, temperature data, and sunshine data at the current moment and the previous n moments; input the distribution transformer low-voltage side load into the trained autoregressive moving average model to obtain the predicted distribution transformer low-voltage side load at the next moment, and input the distributed power output, temperature data, and sunshine data into the correlation equation to obtain the predicted distributed power output at the next moment; obtain the corresponding clustering category based on the predicted distribution transformer low-voltage side load and distributed power output, and perform The line loss rate of the low-voltage active network and the probability distribution of the line loss are obtained at the next moment. Based on the predicted load on the low-voltage side of the distribution transformer, the comparison table of the load rate on the low-voltage side of the distribution transformer and the theoretical limit value of the line loss rate of the low-voltage active network and the fluctuation limit value model of the line loss of the low-voltage active network are used to obtain the theoretical limit value of the line loss rate of the low-voltage active network and the fluctuation limit value of the line loss at the next moment. Based on the line loss rate of the low-voltage active network and the probability distribution of the line loss at the next moment, the theoretical limit value of the line loss rate and the fluctuation limit value of the line loss, the probability of the line loss exceeding the limit at the next moment is obtained and a line loss warning is issued. The approximate equivalent network of the low-voltage network in the substation area is constructed based on the rated capacity of the low-voltage active substation area distribution transformer and the distributed power source, the number of low-voltage users and the load of each low-voltage user, including: Construct a low-voltage single-phase line as an approximate equivalent network of the low-voltage network in the substation, and the length of the approximate equivalent network is 1 / 3 of the total length of the low-voltage line in the low-voltage active network; assume that low-voltage users are evenly distributed on the approximate equivalent network, the number of low-voltage users in the approximate equivalent network is 1 / 3 of the number of low-voltage users in the low-voltage active substation, and the rated capacity of the approximate equivalent network line is 1 / 3 of the rated capacity of the distribution transformer in the low-voltage active substation; assume that the distributed power source in the approximate equivalent network is located at 1 / 2 of the total length of its line, and the capacity of its distributed power source is 1 / 3 of the rated capacity of the distributed power source in the low-voltage active substation; The load rate on the low voltage side of the distribution transformer is When the theoretical limit value of the corresponding low-voltage active network line loss rate is calculated according to the following steps, a comparison table of the theoretical limit value of the low-voltage side load rate of the distribution transformer and the low-voltage active network line loss rate is obtained: Set the total load of low-voltage users to the rated capacity of the distribution transformer. ; Starting from the low-voltage user farthest from the distribution transformer, set the load of the low-voltage user to its rated upper limit from far to near until the total load of the low-voltage user reaches the rated capacity of the distribution transformer. ; The approximate equivalent network distribution transformer low-voltage side load is obtained through the following steps S11 to S14; Obtaining a corresponding theoretical limit value of the low-voltage active network line loss rate based on the theoretical limit value model of the low-voltage active network line loss rate; Said include , , , , and sequentially obtain the theoretical limit values ​​of the low-voltage active network line loss rate corresponding to the above four distribution transformer low-voltage side load rates, and establish a comparison table of the theoretical limit values ​​of the distribution transformer low-voltage side load rate and the low-voltage active network line loss rate; The nodes in the equivalent network include load nodes and balance nodes. Each low-voltage user in the approximate equivalent network is taken as a load node, and the low-voltage side of the distribution transformer in the substation is taken as a balance node; the balance node is taken as the starting node, and its voltage amplitude is limited to 1.0 and the phase angle is limited to 0; the load node farthest from the balance node is taken as the end node; the load on the low-voltage side of the distribution transformer in the approximate equivalent network is obtained by the following steps: S11, setting the initial active and reactive power of the balancing node, i.e. the starting node; S12, based on the voltage amplitude, phase angle, active power and reactive power of the starting node, the voltage amplitude and phase angle of the adjacent next node are obtained through the power flow equation, until the voltage amplitude and phase angle of the end node are obtained; S13, based on the active and reactive power of the end node, the voltage amplitude and phase angle of the end node obtained in step S12, the active and reactive power of the adjacent previous node is obtained, until the active and reactive power of the starting node is obtained; S14, judging whether the termination condition is met, if so, obtaining the low-voltage side load of the distribution transformer according to the active and reactive power of the current starting node; otherwise, repeating steps S12 to S14; The corresponding clustering category is obtained based on the predicted low-voltage side load of the distribution transformer and the output of the distributed power source, and then the line loss rate of the low-voltage active network and the probability distribution of the line loss at the next moment are obtained, including: Calculate the Euclidean distance between the predicted distribution transformer low-voltage side load and the distributed power output and the cluster centers of all cluster categories, obtain the corresponding cluster category based on the minimum Euclidean distance, and thus obtain the probability distribution of the corresponding cluster category line loss rate, use the probability distribution of the corresponding cluster category line loss rate as the probability distribution of the low-voltage active network line loss rate at the next moment, and obtain the probability distribution of the low-voltage active network line loss at the next moment based on the predicted distribution transformer low-voltage side load and the probability distribution of the low-voltage active network line loss rate at the next moment; The theoretical limit value model of the low voltage active network line loss rate is: ， in, is the theoretical limit value of the line loss rate of the low voltage active network, It is the load on the low voltage side of the approximately equivalent network distribution transformer. It is the total load of low-voltage users in the approximately equivalent network.

2. The low voltage active network line loss early warning method considering theoretical limit and fluctuation amplitude according to claim 1 is characterized in that: The fluctuation limit model of low voltage active network line loss is established by the following method: ， Where z is the fluctuation limit of the low voltage active network line loss, is the preset minimum fluctuation limit, for The proportionality coefficient between is the load on the low-voltage side of the distribution transformer at the current moment, The load on the low-voltage side of the distribution transformer in the previous moment; The value of is selected in a graded adaptive way, that is, the amplitude change of the low-voltage side load of the distribution transformer at adjacent moments is divided into several grades from small to large, each grade represents a different amplitude change range of the low-voltage side load of the distribution transformer at adjacent moments, and each grade corresponds to different The greater the change in the amplitude of the load on the low-voltage side of the distribution transformer at adjacent moments, the greater the corresponding The smaller the value.

3. The low voltage active network line loss early warning method considering theoretical limit and fluctuation amplitude according to claim 1 is characterized in that: The method of obtaining the line loss exceeding limit probability and performing line loss warning based on the line loss rate of the low voltage active network at the next moment, the probability distribution of the line loss, the theoretical limit value of the line loss rate, and the fluctuation limit value of the line loss includes: The first over-limit probability is obtained based on the probability distribution of the line loss of the low-voltage active network at the next moment and the theoretical limit value of the line loss rate; The second over-limit probability is obtained based on the probability distribution of the low-voltage active network line loss at the next moment, the fluctuation limit of the line loss and the line loss value at the current moment; A line loss rate threshold is set, and a third over-limit probability is obtained based on the probability distribution of the line loss rate of the low-voltage active network at the next moment and the line loss rate threshold; The line loss exceeding limit probability at the next moment is obtained based on the first exceeding limit probability, the second exceeding limit probability and the third exceeding limit probability. If the line loss exceeding limit probability at the next moment exceeds the warning threshold, a line loss warning is issued.

4. The low voltage active network line loss early warning method considering theoretical limit and fluctuation amplitude according to claim 3 is characterized in that: The probability of line loss exceeding the limit at the next moment is: ， in, is the probability of line loss exceeding the limit, is the first crossing probability, is the second crossing probability, is the third limit-crossing probability.

5. The low voltage active network line loss early warning method considering theoretical limit and fluctuation amplitude according to claim 1 is characterized in that: Obtain the historical distribution transformer low-voltage side load curve of the low-voltage active station area within a period of time, divide the historical distribution transformer low-voltage side load curve by hour, and obtain a number of third samples; the third sample corresponding to the distribution transformer low-voltage side load at time t includes to The load on the low-voltage side of the distribution transformer at four moments, the load on the low-voltage side of the distribution transformer at moment t in the previous year, the load on the low-voltage side of the distribution transformer at moment t in the previous month, and the load on the low-voltage side of the distribution transformer at moment t in the previous week; constructing an autoregressive moving average model, and training the autoregressive moving average model based on the third sample to obtain a trained autoregressive moving average model; The autoregressive moving average model is: ， in, is the predicted value of the load on the low-voltage side of the distribution transformer at time t, is the third sample corresponding to the load on the low-voltage side of the distribution transformer at time t, is a constant, , is a coefficient whose value is not 0, is the error term.

6. The low voltage active network line loss early warning method considering theoretical limit and fluctuation amplitude according to claim 5 is characterized in that: Obtain the historical distributed power output curve, historical temperature data curve, and historical sunshine data curve of the low-voltage active area over a period of time; divide the historical distributed power output curve, historical temperature data curve, and historical sunshine data curve by hour to obtain a number of second samples, the i-th second sample includes the distributed power output, temperature data, and sunshine data of the i-th hour of the area, expressed as: ,in, is the distributed power output of the i-th second sample, is the temperature data of the i-th second sample, is the sunshine data of the i-th second sample; constructing a correlation equation, and optimizing the weight coefficient of the correlation equation based on the second sample to obtain a final correlation equation.

7. The low voltage active network line loss early warning method considering theoretical limit and fluctuation amplitude according to claim 1 is characterized in that: Obtain the historical distribution transformer low-voltage side load curve, historical distributed power output curve, and historical low-voltage active network line loss value curve of the low-voltage active area over a period of time; divide the historical distribution transformer load curve, historical distributed power output curve, and historical line loss value curve by hour to obtain a number of first samples, wherein the i-th first sample includes the distribution transformer low-voltage side load, distributed power output, and low-voltage active network line loss value of the i-th hour of the area; Calculate the Euclidean distance between the low-voltage side load of the distribution transformer and the output of the distributed power source of each two first samples as the Euclidean distance between the corresponding two first samples; when the Euclidean distance between the two first samples is less than a threshold, assign the two first samples to a cluster category through clustering; If two cluster categories contain the same first sample, the two cluster categories are merged to obtain a merged cluster category, otherwise, no merging is required; all cluster categories are obtained based on the merged cluster categories and the cluster categories that do not need to be merged, and the average value of all first samples in each cluster category is calculated to obtain the cluster center of each cluster category.

8. The low voltage active network line loss early warning method considering theoretical limit and fluctuation amplitude according to claim 7 is characterized in that: For each cluster: calculate the line loss rate corresponding to each first sample in the cluster and obtain the maximum line loss rate and the minimum line loss rate; Set interval thresholds, and set line loss rate intervals based on the maximum line loss rate, minimum line loss rate, and interval thresholds; The number of first samples of the line loss rate corresponding to each line loss rate interval is counted to obtain the probability distribution of the line loss rate of the cluster category.

Images