Intelligent identification and control method for distribution network dispatching operation based on cloud platform

Abstract

This invention relates to the field of power grid operation control technology, and in particular to a cloud platform-based intelligent identification and control method for distribution network dispatching and operation. The method includes: calculating representative electricity consumption values ​​for power-consuming areas in each time interval based on electricity consumption data stored on the cloud platform; determining interval characteristic values ​​for each time interval based on these representative values; determining the data properties of the representative electricity consumption values ​​based on these interval characteristic values; determining parallel regions based on these data properties; calculating the combined path values ​​within the parallel regions to obtain the optimal combined region; and automatically adjusting route configurations based on the optimal combined region to achieve a uniform distribution of electricity load, avoid local overload or underload, and enhance the reliability and economy of the power grid.

Description

Technical Field

[0001] This invention relates to the field of power grid operation control technology, and in particular to a cloud-based intelligent identification and control method for distribution network dispatching and operation. Background Technology

[0002] In modern power systems, distribution network dispatching and operation are key links to ensure reliable power supply and efficient utilization.

[0003] The prior art CN110474334A discloses a method for controlling the entire process of distribution network dispatching and operation, including: setting up a closed-loop control process for fault handling in the distribution automation system that covers all types of faults in the distribution network and runs through the entire fault handling process; automatically analyzing and identifying faults and locating tripping points and fault sections; generating fault isolation and transfer schemes and providing corresponding operation sequences; establishing an information interaction channel with the OMS and publishing power outage information; providing a one-click sequential control function for the scheme; manually resetting the fault point and regenerating the fault isolation and transfer scheme; merging the secondary fault handling process with the primary fault handling process; recording fault and emergency repair content and synchronizing it to the OMS dispatch log, supporting the selection of safety measures information; identifying remote control failures and missed or false alarms, and pushing the list of primary and secondary equipment involved to the defect control process; generating a post-event analysis report; and providing a historical data query tool.

[0004] However, as distribution automation continues to improve, distribution dispatching still relies on human experience, which makes it impossible for distribution dispatching to respond to changes in user demand in real time, resulting in uneven distribution of power resources and affecting the stability and economy of the power grid. Summary of the Invention

[0005] The purpose of this invention is to solve the problems in the background art by proposing a cloud platform-based intelligent identification and control method for power distribution network scheduling and operation.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A cloud-based intelligent identification and control method for power distribution network scheduling and operation, comprising the following steps:

[0008] Step 1: Determine the overall area of ​​the power distribution network and the power supply path of the power grid circuits within the overall area. At the same time, divide the overall area of ​​the power distribution network into several power consumption areas based on the power supply path. Collect the power consumption data of the power consumption areas in real time using power acquisition equipment and upload it to the cloud platform for storage.

[0009] Step 2: Set time intervals, calculate the representative power consumption value of the power-consuming area in each time interval based on the stored data of the cloud platform, and then calculate the standard deviation of the data for each time interval based on the representative power consumption value. Use the standard deviation of the data to determine the stable data in the representative power consumption value, and then take the mean of the stable data to obtain the interval feature value. Finally, determine the convex and concave data in the representative power consumption value based on the interval feature value.

[0010] Step 3: Sequentially use each time interval as a calibration interval. Based on the convex and concave data and interval feature values ​​in the calibration interval, determine the parallel region. The parallel region includes the first sub-region and the second sub-region. Then, based on the combined path value of the parallel region, obtain the optimal combined region. Then, merge the two power-consuming regions in the optimal combined region to obtain the combined power supply region. The distribution network system automatically adjusts the line configuration according to the lines of the combined power supply region corresponding to the real-time time interval.

[0011] Step 4: Obtain the power supply and power consumption of the distribution network to the entire region in each time interval. Then, based on the power transmission path value of the distribution network, determine the simulation model of the unit loss coefficient based on the neural network model. Then, identify the actual time interval at this time, determine the power consumption value of the entire region based on the representative value of electricity consumption, and determine the actual power supply in this time interval based on the power consumption value and the unit loss coefficient. Finally, the distribution network supplies power to the power-consuming area according to the actual power supply.

[0012] As a further aspect of the present invention, the method for calculating the electricity consumption representative value includes:

[0013] S1: Use one day as the cycle time point, and divide the day into several time intervals according to the unit time.

[0014] S2: Randomly select a power consumption area as the target analysis area, and extract the power consumption data of the target analysis area in the previous period from the storage information of the cloud platform, with the period as the threshold;

[0015] Select a time interval as the calibration interval, and obtain the representative electricity consumption value of the target analysis area within the calibration interval. Specifically, based on the cyclic time points in the cycle time, the electricity consumption data of the calibration area at each cyclic time point is obtained, and the average value of all electricity consumption data is taken as the representative electricity consumption value of the target analysis area within the calibration interval.

[0016] As a further aspect of the present invention, the method for determining stable data in electricity consumption representative values ​​includes:

[0017] Calculate the representative electricity consumption values ​​of all power-consuming areas in different time intervals. At the same time, select a calibration interval in the time interval and mark the representative electricity consumption values ​​of the power-consuming areas in this calibration interval as DYi, i∈[1,I], indicating that there are I power-consuming areas.

[0018] based on The standard deviation u of the data in the calibration interval is obtained, and Dp is the mean of the representative electricity consumption value DYi;

[0019] According to the normal distribution algorithm, a distribution interval Q1 is taken, Q1 = [Dp-2u, Dp+2u]. The electricity consumption representative value DYi of the calibration interval is compared with the distribution interval. If DYi ∈ Q1, the corresponding electricity consumption representative value is marked as stable data; otherwise, if... The corresponding electricity consumption value will then be marked as peak data.

[0020] As a further aspect of the present invention, the method for determining convexity data and concaveness data includes:

[0021] After comparing the representative electricity consumption value DYi with the distribution interval, all stable data in the calibration interval are obtained, and the mean of the stable data is calculated. The result of the mean calculation is marked as the interval characteristic value Tq. Then, the peak data of the calibration interval is marked as Zf. The peak data Zf is compared with the interval characteristic value Tq. If Zf > Tq, the representative electricity consumption value corresponding to the peak data Zf is marked as convex data. If Zf < Tq, the representative electricity consumption value corresponding to the peak data Zf is marked as concave data.

[0022] As a further aspect of the present invention, the method for determining the parallel region includes:

[0023] Each time interval is used as a calibration interval in turn. The electricity consumption representative value of each power-consuming user in the calibration interval is obtained, and the interval characteristic value Tq of the calibration interval is obtained at the same time.

[0024] Choose data from convex and concave data, and make inequality B1 based on the chosen data: Established, of which X k Let represent convex and concave data, k1 and k2 belong to k, and k1≠k2, n is a positive integer, and b1 is a threshold constant;

[0025] Based on inequality B1, X that makes inequality B1 true k1 and X k2 The corresponding power-consuming region is marked as the parallel region, and X is also marked as the parallel region. k1 The corresponding power-consuming area is marked as the first sub-region, and X is... k2 The corresponding power-consuming region is marked as the second sub-region, and the parallel region consists of the first sub-region and the second sub-region.

[0026] As a further aspect of the present invention, when determining the parallel region according to inequality B1, the data properties in the parallel region can only be a combination of concave data and concave data, or a combination of concave data and convex data. If there exists convex data divided by Tq∈[n, n+b1], then the power consumption region corresponding to this convex data is directly marked as the parallel region.

[0027] As a further aspect of the present invention, the method for obtaining the combined power supply area includes:

[0028] Obtain all parallel regions within the calibration interval, and simultaneously obtain the combined path value LHc of the first and second sub-regions within each parallel region, based on... Determine the optimal combination region, where c represents different parallel regions, and c∈[1,f1], f1 represents the number of parallel regions obtained after operating on inequality B1. Indicates taking The minimum value of the operation;

[0029] Based on the optimal combination region, the first and second sub-regions in the optimal combination region are merged to obtain the combined power supply region. At the same time, based on inequality B1, the n value of each optimal combination region is determined.

[0030] The distribution network identifies real-time time intervals. The distribution network system automatically adjusts the line configuration based on the lines of the combined power supply area corresponding to the real-time time interval and the n value, and supplies power to the power consumption area corresponding to the combined power supply area and the stable data.

[0031] As a further aspect of the present invention, when determining the optimal combination region, in a set of During the operation, the first and second sub-regions of any parallel region do not overlap with the first and second sub-regions of other parallel regions. If there is overlap, the combined path value LHc that overlaps is randomly deleted.

[0032] As a further aspect of the present invention, the method for determining the unit loss coefficient includes:

[0033] SS1: Based on the cloud platform's storage information within a cycle time, collect the power supply Goj of the distribution network in different time intervals at each cycle time point. At the same time, collect the power consumption value of the overall area of ​​the distribution network in different time intervals and mark it as Dzj, where j represents different time intervals.

[0034] Among them, power supply refers to the total power provided by the distribution network to the user side, including the power consumed by the user and the power lost during line transmission; power consumption value refers to the sum of power consumption data of all power-consuming areas.

[0035] SS2: Measures the power transport distance Ld of the distribution network. The power transport distance Ld refers to the sum of the power movement paths between the distribution network and all power-consuming areas.

[0036] The power supply (Goj) and energy consumption (Dzj) in each time interval are used as data sets. Combined with the power transport distance (Ld), the data sets for each time interval within the period are used as experimental data and input into the simulation model for simulation training. The simulation model M1 is then obtained based on the training results. Where Hd is the unit loss coefficient, Go is the power supply, Dz is the power consumption value, Go and Dz correspond to each other, a1, a2 and a3 are fixed factors, and λ is the exponential factor.

[0037] As a further aspect of the present invention, the method for determining the actual power supply during a time interval includes:

[0038] Identify the actual time interval at this time, obtain the representative power consumption value of all power-consuming areas within this time interval, calculate the sum of the representative power consumption values ​​of all power-consuming areas, and mark it as DXb;

[0039] according to The optimal solution for the power supply Go is obtained, and this optimal solution is taken as the actual power supply for the current time interval. Finally, the distribution network supplies power to the power-consuming area based on the actual power supply. Here, Hdmin refers to the minimum value of the unit loss coefficient Hd within a defined range. The defined range refers to... The defined interval is [Ry1, Ry2], where Ry1 and Ry2 are both proportional thresholds, and Ry1≤Ry2.

[0040] Compared with existing technologies, the advantages of this invention are:

[0041] This invention calculates the representative electricity consumption value of the power-consuming area in each time interval, determines the interval characteristic value of each time interval based on the representative electricity consumption value, determines the data nature of the representative electricity consumption value based on the interval characteristic value, and then takes convex and concave data according to the data nature and performs further analysis to obtain parallel regions. Then, it calculates the combined path values ​​in the parallel regions to determine the optimal combined region. Then, the distribution network automatically adjusts the route configuration according to the optimal combined region, thereby achieving a uniform distribution of electricity load, avoiding local overload or light load, and enhancing the reliability and economy of the power grid.

[0042] This invention determines a simulation model of the unit loss coefficient of the power distribution network by combining historical power supply data and user energy consumption values ​​stored in the cloud platform with the power transmission distance of the power distribution network. Then, based on the sum of representative power consumption values ​​of the power consumption area in the actual time interval, the actual power supply in this time interval is determined, and power is supplied to the power consumption area according to the actual power supply, thereby improving the efficiency of power resource utilization and reducing transmission losses. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the process method structure of the present invention. Detailed Implementation

[0044] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0045] Example 1, refer to Figure 1 A cloud-based intelligent identification and control method for power distribution network scheduling and operation, which specifically includes the following steps:

[0046] Step 1: Based on the power distribution information, determine the overall power supply area of ​​the power distribution network. At the same time, based on the power supply path of the power grid circuit, divide the overall power supply area of ​​the power distribution network into several power consumption areas. Each branch of the power supply path in the power grid circuit corresponds to a power consumption area. Furthermore, there will be multiple power users in a power consumption area.

[0047] Then, a power consumption data acquisition device is installed on each power user. The power consumption data acquisition device is used to collect the power consumption data of the power user in real time and upload it to the cloud platform. The cloud platform is used to store the real-time collected power consumption data. In this embodiment, the power consumption data acquisition device is a smart meter.

[0048] Step Two: Based on the storage information of the cloud platform, analyze the data properties of electricity consumption data in different time intervals for the power-consuming region. Data properties include stable data, convex data, and concave data. Specifically, the methods for determining the data properties of electricity consumption data in different time intervals for the power-consuming region include:

[0049] S1: Using one day as the cycle time point, and dividing one day into several time intervals according to the unit time. In this embodiment, the unit time is set to 1 hour, so there are 24 time intervals in one cycle time point.

[0050] S2: Select any power consumption area as the target analysis area, and extract the power consumption data of the target analysis area in the previous period from the storage information of the cloud platform, where the period is a threshold. In this embodiment, the period is set to 3 months.

[0051] Select a time interval as the calibration interval, and obtain the representative value of electricity consumption in the target analysis area within the calibration interval. Specifically, based on the cycle time points in the cycle time, obtain the electricity consumption data of the calibration area at each cycle time point, and take the average of all electricity consumption data as the representative value of electricity consumption.

[0052] It should be further explained that when taking the average of the electricity consumption data of the calibrated area within the period, it is necessary to identify abnormal data, delete the identified abnormal data, and then perform the average calculation.

[0053] The time intervals of the target analysis area are successively used as calibration intervals, and the representative value of electricity consumption for each time interval is calculated according to the above method.

[0054] S3: Following the method in step S2, calculate the representative power consumption values ​​of all power-consuming areas in different time intervals. At the same time, select a calibration interval in the time interval and mark the representative power consumption values ​​of the power-consuming areas in this calibration interval as DYi, i∈[1,I], indicating that there are I power-consuming areas.

[0055] based on The standard deviation u of the data in the calibration interval is obtained, and Dp is the mean of the representative electricity consumption value DYi;

[0056] Then, according to the normal distribution algorithm, a distribution interval Q1 is taken, Q1 = [Dp-2u, Dp+2u]. The electricity consumption representative value DYi of the calibration interval is compared with the distribution interval. If DYi ∈ Q1, the corresponding electricity consumption representative value is marked as stable data; otherwise, if... The corresponding electricity consumption value will then be marked as peak data;

[0057] After comparing the electricity consumption representative value DYi with the distribution interval, all stable data in the calibration interval are obtained, and the mean of the stable data is calculated. The result of the mean calculation is marked as the interval characteristic value Tq. Then, the peak data of the calibration interval is marked as Zf. The peak data Zf is compared with the interval characteristic value Tq. If Zf > Tq, the electricity consumption representative value corresponding to this peak data Zf is marked as convex data. If Zf < Tq, the electricity consumption representative value corresponding to this peak data Zf is marked as concave data. It should be further noted that there is no case where Zf = Tq. If Zf = Tq, it indicates that there is an abnormality in the front-end data processing. In this case, a processing abnormality signal is generated and the relevant technical personnel are reminded through the terminal device.

[0058] Using the method described above, the data properties of representative electricity consumption values ​​for all power-consuming areas in different time intervals are marked.

[0059] Step 3: Based on the data characteristics of the representative electricity consumption values ​​of the power consumption areas, set the optimal combination areas. Specific setting methods include:

[0060] Each time interval is used as a calibration interval in turn. The representative value of electricity consumption of each power-consuming user in the calibration interval is obtained. At the same time, based on the data properties, the mean value of stable data in the calibration interval is obtained, that is, the interval characteristic value Tq.

[0061] Choose data from convex and concave data, and make inequality B1 based on the chosen data: Established, of which X k The data represents convex and concave data, k1 and k2 belong to k, and k1≠k2, n is a positive integer, and b1 is a threshold constant. In this embodiment, b1 is 0.253.

[0062] Then, based on inequality B1, X that makes inequality B1 hold will be... k1 and X k2 The corresponding power-consuming region is marked as the parallel region, and X is also marked as the parallel region. k1 The corresponding power-consuming area is marked as the first sub-region, and X is... k2 The corresponding power-consuming area is marked as the second sub-region. At this time, the parallel region consists of the first sub-region and the second sub-region.

[0063] In another embodiment of this scheme, when determining the parallel region according to inequality B1, the data properties in the parallel region can only be a combination of concave data and concave data, or a combination of concave data and convex data, i.e., X k1 and X k2 Data cannot be convex at the same time. If there is convex data divided by Tq∈[n, n+b1], then the power consumption region corresponding to this convex data is directly marked as the parallel region.

[0064] Obtain all parallel regions within the calibration interval, and simultaneously obtain the combined path value LHc of the first and second sub-regions within each parallel region, based on... Determine the optimal combination region, where c represents different parallel regions, and c∈[1,f1], f1 represents the number of parallel regions obtained after operating on inequality B1. Indicates taking The minimum value of the operation;

[0065] It should be further explained that, when determining the optimal combination region, in a set of During the operation, the first and second sub-regions of any parallel region do not overlap with the first and second sub-regions of other parallel regions. If there is overlap, the combined path value LHc that overlaps is randomly deleted.

[0066] Then, based on the optimal combination region, the first and second sub-regions in the optimal combination region are merged to obtain the combined power supply region. At the same time, based on inequality B1, the n value of each optimal combination region is determined.

[0067] The power supply combination area for each time interval is determined according to the above method. Then, the distribution network identifies the real-time time interval. The distribution network system automatically adjusts the line configuration according to the line of the power supply combination area corresponding to the real-time time interval and the n value, and supplies power to the power consumption area corresponding to the power supply combination area and the stable data, so as to make the power load of the distribution network evenly distributed.

[0068] Example 2, based on Example 1, differs from Example 1 in that it further includes:

[0069] Based on the power transmission path in the power-consuming area, the unit loss value of the distribution network is determined, and based on the unit loss value, the real-time power supply of the distribution network is determined. Specific methods include:

[0070] SS1: Based on the cloud platform's storage information within a cycle time period, collect the power supply of the distribution network Goj in different time intervals at each cycle time point;

[0071] At the same time, the energy consumption values ​​of the entire power distribution network area in different time intervals are collected and marked as Dzj, where j represents different time intervals;

[0072] It should be further explained that the power supply refers to the total power provided by the distribution network to the user side, including the power consumed by the user and the power lost during line transmission. The power consumption value refers to the sum of the power consumption data of all power-consuming areas.

[0073] SS2: Measures the power transport distance Ld of the distribution network. The power transport distance Ld refers to the sum of the power movement paths between the distribution network and all power-consuming areas.

[0074] The power supply (Goj) and energy consumption (Dzj) in each time interval are used as data sets. Combined with the power transport distance, the data sets for each time interval within the cycle are used as experimental data and input into the simulation model for simulation training. The simulation model M1 is then obtained based on the training results. Where Hd is the unit loss coefficient, Go is the power supply, Dz is the power consumption value, Go and Dz correspond to each other, a1, a2 and a3 are fixed factors, and λ is an exponential factor. The specific a1, a2, a3 and λ are obtained by the simulation model after multiple iterations of training. In this embodiment, the simulation model uses a neural network model algorithm. The specific algorithm process is existing technology and will not be described in detail here.

[0075] Identify the actual time interval at this time, obtain the representative power consumption value of all power-consuming areas within this time interval, calculate the sum of the representative power consumption values ​​of all power-consuming areas, and mark it as DXb;

[0076] Then according to The optimal solution for the power supply Go is obtained, and this optimal solution is taken as the actual power supply for the current time interval. Finally, the distribution network supplies power to the power-consuming area based on the actual power supply. Here, Hdmin refers to the minimum value of the unit loss coefficient Hd within a defined range. The defined range refers to... The defined interval is [Ry1, Ry2], where Ry1 and Ry2 are both proportional thresholds, and Ry1≤Ry2. The specific values ​​of Ry1 and Ry2 are obtained by those skilled in the art through big data calculations.

[0077] Example 3 is based on Example 1 and Example 2, and is used to integrate Example 1 and Example 2 and implement them.

[0078] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A cloud-based intelligent identification and control method for power distribution network scheduling and operation, characterized in that, Specifically, the following steps are included: Step 1: Determine the overall area of ​​the power distribution network and the power supply path of the power grid circuits within the overall area. At the same time, divide the overall area of ​​the power distribution network into several power consumption areas based on the power supply path. Collect the power consumption data of the power consumption areas in real time using power acquisition equipment and upload it to the cloud platform for storage. Step 2: Set time intervals, calculate the representative power consumption value of the power-consuming area in each time interval based on the stored data of the cloud platform, and then calculate the standard deviation of the data for each time interval based on the representative power consumption value. Use the standard deviation of the data to determine the stable data in the representative power consumption value, and then take the mean of the stable data to obtain the interval feature value. Finally, determine the convex and concave data in the representative power consumption value based on the interval feature value. The methods for determining convex and concave data include: S1: Use one day as the cycle time point, and divide the day into several time intervals according to the unit time. S2: Randomly select a power consumption area as the target analysis area, and extract the power consumption data of the target analysis area in the previous period from the storage information of the cloud platform, with the period as the threshold; Select a time interval as the calibration interval, and obtain the representative value of electricity consumption of the target analysis area in the calibration interval. Specifically, according to the cycle time points in the cycle time, the electricity consumption data of the calibration area at each cycle time point is obtained, and the average value of all electricity consumption data is taken as the representative value of electricity consumption of the target analysis area in the calibration interval. S3: Calculate the representative power consumption values ​​of all power-consuming areas in different time intervals, and select a calibration interval in the time interval. Mark the representative power consumption values ​​of the power-consuming areas in this calibration interval as DYi, i∈[1,I], indicating that there are I power-consuming areas. based on The standard deviation u of the data in the calibration interval is obtained, and Dp is the mean of the representative electricity consumption value DYi; Based on the normal distribution algorithm, the distribution interval Q1 is selected. The representative electricity consumption value DYi of the calibration interval is compared with the distribution interval. If DYi ∈ Q1, the corresponding representative electricity consumption value is marked as stable data; otherwise, if DYi ∈ Q1, the corresponding representative electricity consumption value is marked as stable data. Q1, then the corresponding electricity consumption value will be marked as peak data; After comparing the electricity consumption representative value DYi with the distribution interval, all stable data in the calibration interval are obtained, and the mean of the stable data is calculated. The mean calculation result is marked as the interval characteristic value Tq. Then, the peak data of the calibration interval is marked as Zf. The peak data Zf is compared with the interval characteristic value Tq. If Zf > Tq, the electricity consumption representative value corresponding to this peak data Zf is marked as convex data. If Zf < Tq, the electricity consumption representative value corresponding to this peak data Zf is marked as concave data. Step 3: Sequentially use each time interval as a calibration interval. Based on the convex and concave data and interval feature values ​​in the calibration interval, determine the parallel region. The parallel region includes the first sub-region and the second sub-region. Then, based on the combined path value of the parallel region, obtain the optimal combined region. Then, merge the two power-consuming regions in the optimal combined region to obtain the combined power supply region. The distribution network system automatically adjusts the line configuration according to the lines of the combined power supply region corresponding to the real-time time interval. The methods for determining the parallel region include: Each time interval is used as a calibration interval in turn. The electricity consumption representative value of each power-consuming user in the calibration interval is obtained, and the interval characteristic value Tq of the calibration interval is obtained at the same time. Choose data from convex and concave data, and make inequality B1 based on the chosen data: Established, among which Let represent convex and concave data, k1 and k2 belong to k, and k1≠k2, n is a positive integer, and b1 is a threshold constant; Based on inequality B1, the following will make inequality B1 true: and The corresponding power-consuming region is marked as the parallel region, and at the same time... The corresponding power-consuming area is marked as the first sub-region. The corresponding power-consuming region is marked as the second sub-region, and the parallel region consists of the first and second sub-regions; Step 4: Obtain the power supply and power consumption of the distribution network to the entire region in each time interval. Then, based on the power transmission path value of the distribution network, determine the simulation model of the unit loss coefficient based on the neural network model. Then, identify the actual time interval at this time, determine the power consumption value of the entire region based on the representative power consumption value, and determine the actual power supply in this time interval based on the power consumption value and the unit loss coefficient. Finally, the distribution network supplies power to the power-consuming area according to the actual power supply.

2. The intelligent identification and control method for distribution network scheduling and operation based on a cloud platform according to claim 1, characterized in that, When determining the parallel region according to inequality B1, the data properties in the parallel region can only be combined by concave data and concave data, and concave data and convex data. If there exists convex data divided by Tq∈[n, n+b1], then the power consumption region corresponding to this convex data is directly marked as the parallel region.

3. The intelligent identification and control method for distribution network scheduling and operation based on a cloud platform according to claim 1, characterized in that, Methods for obtaining combined power supply areas include: Obtain all parallel regions within the calibration interval, and simultaneously obtain the combined path value LHc of the first and second sub-regions within each parallel region, based on... Determine the optimal combination region, where c represents different parallel regions, and c∈[1,f1], f1 represents the number of parallel regions obtained after operating on inequality B1. Indicates taking The minimum value of the operation; Based on the optimal combination region, the first and second sub-regions in the optimal combination region are merged to obtain the combined power supply region. At the same time, based on inequality B1, the n value of each optimal combination region is determined. The distribution network identifies real-time time intervals. The distribution network system automatically adjusts the line configuration based on the lines of the combined power supply area corresponding to the real-time time interval and the n value, and supplies power to the power consumption area corresponding to the combined power supply area and the stable data.

4. The intelligent identification and control method for distribution network scheduling and operation based on a cloud platform according to claim 3, characterized in that, When determining the optimal combination region, in a set of During the operation, the first and second sub-regions of any parallel region do not overlap with the first and second sub-regions of other parallel regions. If there is overlap, the combined path value LHc that overlaps is randomly deleted.

5. The intelligent identification and control method for distribution network scheduling and operation based on a cloud platform according to claim 1, characterized in that, Methods for determining the unit loss coefficient include: SS1: Based on the cloud platform's storage information within a cycle time, collect the power supply Goj of the distribution network in different time intervals at each cycle time point. At the same time, collect the power consumption value of the overall area of ​​the distribution network in different time intervals and mark it as Dzj, where j represents different time intervals. Among them, power supply refers to the total power provided by the distribution network to the user side, including the power consumed by the user and the power lost during line transmission; power consumption value refers to the sum of power consumption data of all power-consuming areas. SS2: Measures the power transport distance Ld of the distribution network. The power transport distance Ld refers to the sum of the power movement paths between the distribution network and all power-consuming areas. The power supply (Goj) and energy consumption (Dzj) in each time interval are used as data sets. Combined with the power transport distance (Ld), the data sets for each time interval within the period are used as experimental data and input into the simulation model for simulation training. The simulation model M1 is then obtained based on the training results. Where Hd is the unit loss coefficient, Go is the power supplied, Dz is the energy consumption value, Go and Dz correspond to each other, and a1, a2 and a3 are fixed factors. It is an exponential factor.

6. The intelligent identification and control method for distribution network scheduling and operation based on a cloud platform according to claim 5, characterized in that, The methods for determining the actual power supply during a time interval include: Identify the actual time interval at this time, obtain the representative power consumption value of all power-consuming areas within this time interval, calculate the sum of the representative power consumption values ​​of all power-consuming areas, and mark it as DXb; according to The optimal solution for the power supply Go is obtained, and this optimal solution is taken as the actual power supply for the current time interval. Finally, the distribution network supplies power to the power-consuming area based on the actual power supply. Here, Hdmin refers to the minimum value of the unit loss coefficient Hd within a defined range. The defined range refers to... The defined interval is [Ry1, Ry2], where Ry1 and Ry2 are both proportional thresholds, and Ry1≤Ry2.

Images