CN-121984009-A - Three-phase four-wire system power distribution network tide simulation method and system based on power system simulation library

CN121984009ACN 121984009 ACN121984009 ACN 121984009ACN-121984009-A

Abstract

The invention provides a three-phase four-wire system power distribution network tide simulation method and system based on a power system simulation library, wherein the method comprises the steps of constructing a three-phase four-wire system power distribution network topological structure, including a distribution box power supply end, a main line and branch nodes; the method comprises the steps of establishing a multi-type user load model, generating a time sequence electricity consumption curve containing multiple layers of random disturbance, simulating electricity stealing behaviors on the time sequence electricity consumption curve, generating actual electricity consumption inconsistent with an electricity meter metering value, calculating and outputting actual electric parameters of a user by means of power flow calculation based on a power curve of the actual electricity consumption, and processing the actual electric parameters of the user according to the electricity stealing behaviors to output a multi-dimensional simulation data set containing the electricity meter metering electric parameters and electricity stealing behavior labels. According to the invention, by constructing a radial network topology and based on a user type differential load curve, noise and electricity stealing behavior data are superimposed, a real electricity utilization scene is simulated, a multidimensional data set is output, and a technical support is provided for safe and stable operation of an electric power system.

Inventors

WANG CHENG
MAO ZHEN
YIN JIANFENG
SUN YUE
ZHU DESHENG

Assignees

江苏林洋能源股份有限公司

Dates

Publication Date: 20260505
Application Date: 20251230

Claims (10)

1. The power flow simulation data generation method of the three-phase four-wire system power distribution network based on the power system simulation library is characterized by comprising the following steps of: s1, constructing a three-phase four-wire system distribution network topological structure, wherein the topological structure comprises a distribution box power supply end, a main line and branch nodes; s2, establishing a multi-type user load model, and generating a time sequence electricity consumption curve containing multiple layers of random disturbance; s3, simulating electricity stealing behavior on a time sequence electricity consumption curve to generate actual electricity consumption inconsistent with the metering value of the ammeter; s4, based on a power curve of the actual power consumption, carrying out tide calculation to output actual electrical parameters of a user; s5, processing actual electrical parameters of the user according to electricity stealing behaviors, and outputting a multi-dimensional simulation data set which comprises the electric parameters measured by the ammeter and the electricity stealing behavior labels.
2. The method of claim 1, wherein S1 comprises: s11, randomly generating at least one main line by taking a power end of a distribution box as a starting point, and forming a radial network topology; s12, generating branch nodes on the main line according to a preset distance, wherein each branch node is connected with a corresponding user ammeter box through a line, and each ammeter in the user ammeter box is connected to a user terminal through a household wiring.
3. The method of claim 1, wherein S2 comprises: s21, defining user types and constructing a reference daily load curve for each user type, wherein the user types comprise an outgoing business worker family, a free professional, a office worker family and a retired old man family; s22, superposing a seasonal scale fluctuation coefficient and a daily scale fluctuation coefficient on a reference daily load curve; S23, injecting daily noise into different users with preset probability, and generating differentiated time sequence electricity consumption curves for the users.
4. A method according to claim 3, wherein S22 is specifically configured to set a seasonal scale fluctuation coefficient and a daily scale fluctuation coefficient for each user according to seasons and different dates, respectively, to reflect the electricity consumption level of the user.
5. The method of claim 3, wherein in S23, the intra-day noise comprises micropulse events for simulating high power electricity usage scenarios and perturbation events for simulating differential electricity usage scenarios generated by outgoing, sick, and gathering event scenarios.
6. The method of claim 1, wherein S3 comprises: S31, randomly selecting at least one non-zero electricity utilization user as an electricity stealing user in a simulation period, and setting a continuous electricity stealing time window in an electricity utilization peak period; S32, in the electricity stealing time window, performing proportion hiding operation on the normal electricity consumption to generate a reduced ammeter metering value; S33, generating a power stealing behavior label corresponding to the power stealing type, the power stealing period and the power stealing intensity, wherein the power stealing intensity comprises a power stealing proportion and additional power consumption.
7. The method of claim 6, wherein, The execution proportion hiding operation is ammeter data tamper type electricity larceny, and the electricity larceny proportion is applied to recorded values in the time-consuming electricity curve to be adjusted, so that an ammeter metering value is obtained; And the operation of superposing the extra power consumption is the behavior of stealing electricity around a meter and pulling the electric wire privately.
8. The method of claim 1, wherein S4 comprises: s41, distributing the power supply terminal voltage of the distribution box at the current moment to each branch node as an initial voltage; s42, calculating a current value according to a power curve of the actual power consumption of each user, accumulating and returning all the user currents to a power supply end, and obtaining current values among branch nodes; s43, pushing the new voltage value of each branch node forward according to the line impedance and the current value among the branch nodes; s44, comparing the difference value between the new voltage value and the initial voltage of each branch node, confirming whether the new voltage value is within a preset convergence threshold value, and repeating tide iteration if the new voltage value exceeds the threshold value until convergence conditions are met, and outputting actual electrical parameters of a user, including voltage, current, power, active power, reactive power and electricity consumption.
9. The method of claim 1, wherein S5 comprises: when the electricity stealing behavior is a proportion hiding operation, multiplying the electricity consumption in the actual electrical parameters of the user by the electricity stealing proportion to obtain a meter measurement value; When the electricity stealing behavior is the operation of overlapping the additional electricity consumption, subtracting the overlapped additional electricity consumption from the electricity consumption in the actual electrical parameters of the user to obtain a meter measurement value.
10. A three-phase four-wire system power distribution network flow simulation data generation system based on a power system simulation library, characterized in that the system is configured to perform the method of any of claims 1-9.

Description

Three-phase four-wire system power distribution network tide simulation method and system based on power system simulation library Technical Field The invention belongs to the technical field of power system simulation and data analysis, and particularly relates to a power distribution network simulation method and system construction and power data generation method based on Pandapower. Background With the deep advancement of smart grid construction, requirements for fine management and economic operation of an electric power system are increasingly increased. The power distribution network is used as a final link of the power system facing users, and the accurate grasp of the running state of the power distribution network directly relates to the power supply quality, the energy utilization efficiency and the economic running level of the system. Traditional power distribution network operation analysis methods mainly rely on historical data statistics and simplified model calculation. The methods have obvious limitations in the aspects of reflecting the real-time running state of the system, simulating complex electricity utilization behaviors, analyzing line loss causes and the like, and are difficult to meet the requirements of the modern power distribution network on fine analysis and intelligent management. With the wide application of big data and artificial intelligence technology in the electric power field, advanced applications such as power distribution network running state analysis based on data driving, line loss accurate calculation, abnormal electricity utilization identification and the like have become research emphasis. However, the development and application of algorithms for line loss, load prediction, and topology identification face a key bottleneck, namely, difficulty in acquiring high-quality training data. On one hand, the complete system operation data for model training and verification is deficient, and due to factors such as data security, privacy protection and the like, a data set containing detailed power grid parameters and user behavior characteristics is difficult to obtain, and on the other hand, the massive data generated by the power consumption information acquisition system often have defects in the aspects of integrity, accuracy and labeling information. The quality and quantity problems of data sources severely hamper the research, testing and performance evaluation of high-level analysis algorithms of power distribution networks. In order to break through the bottleneck, the generation of realistic distribution network operation data by using a simulation technology becomes an effective solution. Although the existing simulation tools of the electric power system can perform basic calculation, the simulation tools have obvious defects in the aspects of simulating the behavior diversity of real users, the power consumption randomness and the complex operation characteristics of the system. Therefore, a high-fidelity simulation method and a high-fidelity simulation system which can accurately simulate the topology and the operation characteristics of a power distribution network, truly reflect the diversity of electricity consumption behaviors of users and completely reproduce various operation states of a system are urgently needed, so that the blank of the existing data is filled, and a comprehensive data support and test platform is provided for advanced applications such as operation analysis, line loss accurate calculation, electric energy quality assessment, abnormal electricity consumption detection and the like of the power distribution network. Disclosure of Invention The invention provides a power distribution network multidimensional electricity utilization data generation method and an abnormal electricity stealing simulation system based on power system simulation and user behavior modeling, and aims to solve the problem that the existing power distribution area lacks high-precision, controllable and reusable electricity utilization behavior data and electricity stealing abnormal data, thereby providing data support for development and verification of key algorithms such as electricity stealing detection, line loss calculation, electricity utilization behavior analysis, topology identification and the like. The technical scheme of the invention is as follows: in a first aspect, the invention provides a power flow simulation data generation method for a three-phase four-wire system power distribution network based on a power system simulation library, which comprises the following steps: s1, constructing a three-phase four-wire system distribution network topological structure, wherein the topological structure comprises a distribution box power supply end, a main line and branch nodes; s2, establishing a multi-type user load model, and generating a time sequence electricity consumption curve containing multiple layers of random disturbance; s3, simul