Search

CN-121998361-A - Electric network-traffic energy flow coupling research method under electric automobile access oriented to differentiated scene

CN121998361ACN 121998361 ACN121998361 ACN 121998361ACN-121998361-A

Abstract

The invention relates to the field of power grid analysis, in particular to a power grid-traffic energy flow coupling research method under the condition of electric automobile access facing a differentiated scene, which comprises the steps of firstly constructing a branch-level multi-type electric automobile energy flow model, and realizing the accurate simulation of large-scale traffic branch energy flow through vehicle energy consumption base line, branch state correction and space-time weight fusion; and finally, a dynamic scheduling strategy is provided based on constraint particle swarm optimization, and the charging and discharging, energy storage power and load peak clipping of the electric automobile are optimized. The method forms a closed loop system from scene parameter construction, energy flow simulation, coupling mapping and multidimensional load analysis to optimal scheduling, and provides high-precision and extensible decision basis for power grid safe operation and traffic energy management under high-permeability electric automobile access.

Inventors

  • WANG PENGYU
  • Bai Panpan
  • CHEN YUKAI
  • GAO TIANLE
  • QIU YUE

Assignees

  • 南方电网数字电网研究院股份有限公司

Dates

Publication Date
20260508
Application Date
20260128

Claims (6)

  1. 1. The utility model provides a power grid-traffic energy flow coupling research method under electric automobile access facing differentiation scene, which is characterized by comprising the following steps: The method comprises the steps of S1, providing a regional characteristic quantization and differentiation scene construction method, forming regional operation characteristic vectors by carrying out parameterization modeling on key factors of regional load level, new energy permeation proportion, power grid structure complexity and spatial distribution characteristics, realizing scene classification and characteristic characterization of a research region, and providing scene constraint conditions and parameter bases for subsequent electric vehicle energy flow modeling and power grid-traffic coupling analysis; S2, providing an energy flow modeling method of the branch-level multi-type electric automobile, and simulating the running process of the multi-type electric automobile in a traffic network by constructing an energy consumption base line model of private automobiles, network about automobiles and logistics automobiles and combining traffic branch flow, vehicle acceleration and deceleration behaviors and running working condition characteristics to obtain an energy flow distribution result taking traffic branches as granularity, so that the definition and the authenticity of traffic-side energy flow depiction are improved; S3, constructing a scene-adaptive power grid-traffic energy flow coupling model, and adjusting the energy flow transmission intensity by constructing a mapping relation between traffic branches and power grid nodes and combining regional scene characteristics to realize scene mapping of traffic branch energy flow to power grid node loads, thereby constructing a power grid-traffic energy flow coupling model reflecting different regional characteristics and solving the defect that the existing coupling method lacks of scene pertinence; S4, designing a multi-dimensional load influence analysis method of the power grid based on the coupling energy flow, and analyzing the load change, voltage response and operation stability of the power grid under different time scales by introducing the coupled traffic energy flow into a power grid analysis model, so as to realize comprehensive analysis of instantaneous load impact and long-term load evolution characteristics of the power grid under the condition of electric automobile access and improve the accuracy and scene adaptability of load influence assessment; S5, designing a comprehensive output method based on a power grid-traffic energy flow coupling result, and realizing analysis result output oriented to different area scenes by carrying out integrated analysis on traffic branch energy flows, power grid node load changes and operation influence indexes under each scene, so as to provide decision basis for power distribution network planning, electric vehicle charging facility layout and operation scheduling.
  2. 2. The method for researching power grid-traffic energy flow coupling under electric automobile access for differentiated scenes according to claim 1, wherein the step S1 comprises the following steps: s11, collecting an original data set of a research area in a given statistical period, wherein the original data set comprises area peak load, renewable energy capacity in area total assembly machine capacity, total power distribution network nodes and total traffic branches, and normalizing the original data set to form a basic feature vector: Wherein, the Representing the collected raw dataset; Representing vectors Any component, and thus a normalized result is obtained ; S12 normalizing the result For input, respectively calculating a power grid intensity factor and a traffic intensity factor for distinguishing scenes: Wherein, the And (3) with Respectively the peak load and the node number after standardization; 、 Weighting coefficients (dimensionless, preset or obtained by fitting historical data) for grid indexes, and As preliminary scene coordinates for scene classification; S13 is based on scene coordinates Determining scene class by using rectangular dividing method, and setting threshold value And (3) with (Dimensionless) the region is identified as one of four classes of scenes according to the following rule: Wherein, the Determining by quantiles of the historical samples; s14 dividing the region into Each space unit calculates the power density per unit area And the traffic density per unit area And defining the original scoring of the cell as its product: The weight factor is then calculated: Wherein index is an index Weight set Reflecting the spatial distribution of the power-traffic coupling intensity in the space and taking the spatial distribution as the spatial weighted input of the energy flow of the subsequent branch, S15 calculating a time adjustment factor based on the historical hour-level time sequence data to describe the peak and off-peak scene differences and define the time The time adjustment factor of (2) is: S16, identifying scene categories Scene coordinates Space weight set Time adjustment function acquired with S15 Combining to form a differential scene parameter set It is defined as: S17 pair for construction Performing integrity check on the basic data of the data points, and complementing the linear interpolation of the missing data points according to adjacent moments, e.g. for missing moments By way of example of (a) using adjacent moments The linear interpolation of (2) gives the estimate: Wherein, the Representing any missing time series data, the complemented data is again used for calculation S18 pairs of generated scene parameter sets Version management is implemented, and updating is carried out daily or hourly according to latest acquired data And recording the parameter change log for subsequent backtracking and model retraining.
  3. 3. The method for researching power grid-traffic energy flow coupling under electric automobile access for differentiated scenes according to claim 1, wherein the step S2 comprises the following steps that S21 respectively collects historical speed-energy consumption samples for three electric automobiles, namely private automobiles, network about automobiles and logistics automobiles, and fits a vehicle unit driving distance energy consumption function through polynomial regression, wherein the method is defined as: Wherein, the Is the first Vehicle-like running speed The energy consumption per unit driving distance under the condition; indexing for polynomial coefficients obtained by fitting historical data Corresponding to the private car of the person, Corresponding to the order of the net, S22, aiming at the influence of factors such as traffic jams, frequent start and stop and the like on energy consumption, introducing a branch state correction factor: Wherein, the Is a branch circuit Is used for correcting the energy consumption coefficient; the standard deviation of the average speed of the branch is used for reflecting the speed fluctuation; Is the average speed of the branch; Is an empirical scale factor; calculating the instantaneous energy consumption of each type of vehicle on each branch by combining the S21 baseline energy consumption function with the S22 branch correction coefficient: Wherein, the Is the first Class vehicle passing branch Energy consumption of (2); is a branch circuit Is determined based on the estimated average speed of (a); is the length of the branch, S24 is carried out on each branch The throughput of the vehicles of various types is overlapped to obtain the total energy flow of the branch circuit: Wherein, the Is the first Class vehicle passing through branch in statistical period Is used in the number of (a) and (b), For total energy consumption of the branch, S25 takes the spatial weight factor calculated in S1 And (3) corresponding to the branch energy flow, carrying out weighted integration: Wherein, the For weighted branch energy flows; is a branch circuit The set of the space units is located, S26 considers the time change of the hour level or the minute level, and adjusts the time of S15 by a factor And (3) introducing branch energy flow calculation: Wherein, the To be at the moment Is of (2) And the energy flow realizes the simulation of dynamic scenes such as the peak of the holidays or the daytime.
  4. 4. The method for studying power grid-traffic energy flow coupling under electric automobile access for differentiated scenes according to claim 3, wherein said step S3 comprises the steps of S31 collecting traffic branches in the area Aggregate with grid nodes Establishing spatial association: Wherein, the Representing branches Corresponding power grid node Is used for the direct load input of (a), Indicating that there is no direct mapping; For the total number of branches, The method is a power grid node set; S32, introducing energy flow guide coefficients aiming at different scenes and branch-node association strengths Adjusting the mapping weight: Wherein, the Is a branch circuit The weight of the space unit; For scene intensity function, the power grid intensity factor And traffic intensity factor Fusion, defined as: The method is used for describing the influence degree of the branch energy flow on the power grid node under different scenes such as strong electricity-strong traffic; s33 through mapping matrix Coefficient of conduction with energy flow Calculating the moment of the power grid node Is a load increment of (a): Wherein, the Is a power grid node Due to the instantaneous load of the traffic branch energy flow input; the branch energy flow is output for S2; s34, introducing scene coupling coefficients for further enhancing coupling accuracy in different scenes Dynamically correcting the node load: Wherein, the According to historical load peak-valley, node power supply capacity and traffic flow fluctuation dynamic update, the method can calculate by an exponential smoothing method: Is a smoothing coefficient; Normalizing the power supply capacity of the node; s35, the load contributed by the traffic branch and the original load of the power grid are combined Adding to obtain the total load of the nodes: ; s36, carrying out total load on all nodes Branch-node mapping matrix And energy flow conductivity coefficients Combining to form a coupling model: 。
  5. 5. the method for researching power grid-traffic energy flow coupling under electric automobile access for differentiated scenes according to claim 4, wherein the step S4 comprises the following steps: S41 utilizing a coupling model Total load of nodes in (3) Constructing an instantaneous load vector of each node: Wherein, the The total number of the grid nodes; s42, calculating node voltage by adopting a three-phase power flow calculation method according to the total load of the nodes and the topological relation of the power distribution network Deviation from voltage : Wherein, the Is a node Jacobian matrix of (a); In order to increase the load caused by the energy flow access of the electric automobile, the power quality index can be comprehensively calculated through voltage deviation and frequency deviation: Wherein, the For the rated voltage to be the rated voltage, For the nominal frequency to be a set value, Is the frequency deviation; S43 load per node Performing multi-time scale decomposition to define minute scale Grade of hours And daily/quaternary level Is a load sequence of (a): inputting the sequence into an LSTM load prediction model, and constructing a prediction function: Wherein the method comprises the steps of For predicted future loads Is the input window length; and S44, defining an instantaneous load increment index for quantifying the instantaneous impact of the electric automobile access to the power grid: and calculate normalized impact strength: Wherein the method comprises the steps of As a baseline load average value, the load average value, Standard deviation for baseline load; s45 combined node load prediction And (3) calculating the long-term load trend by adopting a weighted moving average method and time sequence analysis: Wherein the method comprises the steps of Is a weight coefficient , Long-term trend for sliding window length The method can be used for planning and scheduling strategies of the power distribution network; S46, instantaneous load of the node Voltage and power quality Load impact index With long-term load trend Combining to form a multi-dimensional load analysis result set: 。
  6. 6. the method for researching power grid-traffic energy flow coupling under electric automobile access for differentiated scenes according to claim 5, wherein the step S5 comprises the following steps: s51, setting a comprehensive scheduling target, minimizing total power fluctuation and energy consumption of a system, and simultaneously considering node voltage deviation and power quality indexes: Wherein, the As a result of the fact that the target weight coefficient, Respectively derived from S4 instantaneous load impact, voltage deviation and power quality indexes; s52 introduces adjustable decision variables, including: 1) Charging and discharging power of electric automobile The constraints are: Wherein the method comprises the steps of In the case of a vehicle type, Is a branch index; 2) Adjustable energy storage unit power The constraints are: 3) Load peak clipping/shifting scheduling The node load is used for fine adjustment; s53 node power balancing constraint: Wherein, the Representing branches Corresponding power grid node Is used for the direct load input of (a), Representing the energy flow conductivity coefficient, node voltage constraints: the constraint of the charging/discharging power and the energy storage capacity of the electric automobile is that the rated capacity is not exceeded and the dynamic update of the SOC (State of Charge) is met; S54, solving a scheduling problem by adopting a constraint particle swarm optimization algorithm, and adopting an iterative formula: Wherein the method comprises the steps of Is the first The number of the decision variable vectors of the particles, As a velocity vector of the velocity vector, For the particle history to be optimal, For a global optimum of the device, the device is, As a parameter of the algorithm, Is a random number; S55 sets S1 scene parameters And S2 branch energy flow Integrating optimization iteration, and adjusting the charging and discharging strategies of the electric automobile according to different scenes in each iteration: Wherein the method comprises the steps of The charge-discharge strategy is dynamically adjusted to accord with scenes such as strong current-strong traffic or weak current-weak traffic; The S56 output includes the post-scheduling load per node Electric automobile on branch optimizes charge-discharge power Energy storage unit power Optimizing objective function values The outputs can be directly used for power grid dispatching, load peak clipping, power quality optimization and new energy consumption strategy execution in a differentiated scene.

Description

Electric network-traffic energy flow coupling research method under electric automobile access oriented to differentiated scene Technical Field The invention relates to the field of power grid analysis, in particular to a power grid-traffic energy flow coupling research method under electric automobile access facing to a differentiation scene. Background In recent years, with the rapid popularization of electric vehicles and the increase of urban traffic energy demands, the load characteristics of a power grid gradually show high time-space non-uniformity, and the influence of electric vehicle access on a regional power distribution network is increasingly remarkable. The traditional power grid load analysis method mainly comprises the steps of firstly, enabling dynamic characteristics and charging behavior differences of different types of electric vehicles (private vehicles, network about vehicles and logistics vehicles) to be obvious, enabling the existing energy flow model to assume that vehicles are homogeneous and difficult to reflect dynamic energy consumption distribution of actual traffic branches, so that load assessment deviation is large, secondly, enabling the electric vehicle charging mode to be various and influenced by factors such as time, scenes and prices, enabling the traditional static load analysis method to be difficult to quantify instantaneous load impact and long-term load trend, lacking in multidimensional and space-time resolution capability, thirdly, enabling the power grid structure, power capacity and traffic flow characteristic differences of different areas to be obvious, enabling the existing coupling model to fully consider scene suitability, enabling the power grid-traffic energy coupling relation to be inaccurate, enabling safe operation and optimal scheduling under the differentiated scenes to be difficult to be supported, and enabling the centralized load calculation and scheduling strategy to face calculation efficiency and expandability, and meanwhile enabling traffic-power data dispersion and privacy protection requirements to increase regional cooperative difficulty analysis along with continuous expansion of the electric vehicle access scale. Therefore, a closed-loop method capable of integrating multi-type electric vehicle energy flow modeling, power grid-traffic coupling mapping, multi-dimensional load analysis and optimal scheduling is urgently needed. The method is used for realizing the whole process from traffic branch energy flow simulation, scene adaptation coupling and multidimensional load analysis to optimal scheduling under a unified frame, improving the power grid load analysis precision, robustness and calculation efficiency, and providing scientific basis for the distribution network safe operation, traffic energy management and decision making under the condition of high-permeability electric vehicles. . Disclosure of Invention The invention aims to solve the problems that the branch energy flow of the existing power grid is difficult to accurately simulate and insufficient in coupling analysis and the multidimensional load calculation and optimal scheduling efficiency are low under the condition of large-scale electric vehicle access, and discloses a power grid-traffic energy flow coupling research method under the condition of electric vehicle access for a differentiation scene. The energy consumption baseline model of the branch-level multi-type electric automobile is built, the power grid-traffic coupling mapping model of scene adaptation is built, and a closed-loop analysis system from energy flow simulation, coupling mapping and multidimensional load analysis to optimal scheduling is provided for the technical scheme with high reliability and popularization for power grid safe operation, traffic energy management and expandable scheduling under high-permeability electric automobile access. The invention discloses a differentiation scene-oriented electric automobile access lower power grid-traffic energy flow coupling research method, which comprises the following steps: S1, aiming at the problem that different areas have significant differences in the aspects of power grid bearing capacity, traffic running state, new energy access level and the like, so that the conventional analysis method is difficult to uniformly describe different application scenes, a regional characteristic quantization and differentiation scene construction method is provided, and a regional running characteristic vector is formed by carrying out parametric modeling on key factors such as regional load level, new energy penetration ratio, power grid structure complexity, space distribution characteristics and the like, so that scene classification and characteristic characterization of a research region are realized, and scene constraint conditions and parameter bases are provided for subsequent electric vehicle energy flow modeling and power grid-traffic coupling anal