Search

CN-121984085-A - Online evaluation method and system for grid-connectable capacity margin and trend of sand-gossypii new energy base

CN121984085ACN 121984085 ACN121984085 ACN 121984085ACN-121984085-A

Abstract

The invention discloses an online evaluation method and an online evaluation system for grid-connectable capacity margin and trend of a sand-gossypii new energy base. According to the method, based on meteorological data, power grid operation data and equipment state data acquired in real time, real-time grid-connection capacity margin of a new energy base is calculated through multi-source data fusion and dynamic modeling, and short-term change trend of the new energy base is predicted by utilizing time sequence analysis and a machine learning algorithm. The system comprises a data acquisition layer, a fusion analysis layer, an evaluation prediction layer and a visual interaction layer, and realizes full-flow automation from data to decision. The invention can improve new energy consumption capability in Sha Gehuang areas, enhance the safety and economy of power grid operation and provide scientific basis for scheduling decisions.

Inventors

  • ZHOU QIANG
  • MA YANHONG
  • ZHANG JIANMEI
  • ZHANG RUIXIAO
  • BAO CHENGJIA
  • ZHANG YONGRUI
  • WANG CHENG
  • ZHANG JIALIN
  • ZHANG ZHIJIE
  • ZHANG WENDA
  • ZHAO HANGHANG
  • GAO PENGFEI
  • ZHAO LONG
  • DING KUN
  • LV QINGQUAN
  • ZHANG ZHENZHEN
  • WU GUODONG
  • ZHANG JINPING
  • LI JIN

Assignees

  • 国网甘肃省电力公司电力科学研究院
  • 国网甘肃省电力公司

Dates

Publication Date
20260505
Application Date
20260126

Claims (9)

  1. 1. The on-line evaluation method for the grid-connectable capacity margin and trend of the sand-gossypii new energy base is characterized by comprising the following steps of 1, multi-source data acquisition, wherein the multi-source data acquisition comprises meteorological monitoring data, power grid monitoring data, new energy station monitoring data, energy storage power station monitoring data and management data; Step 2, data fusion processing is carried out, the multi-source data in the step 1 are subjected to space-time alignment, data are cleaned, abnormal values are identified and removed, missing data are filled, key features are extracted, normalization processing is carried out, and a unified data pool is formed; Step 3, dynamic modeling evaluation, namely establishing an online real-time update model, taking a current power grid operating point as a reference, taking a total output increment delta P of a new energy base as an optimized variable, taking max delta P as an objective function, wherein constraint conditions comprise that f (P, Q, V, theta) = 0;P is dielectric injection active power, Q is node injection reactive power, V is node voltage amplitude, theta is node voltage phase angle, f is a load flow equation vector function, and inequality safety constraint that V min ≤V≤V max ,|P line |≤P line-max ;V min is node allowable minimum voltage amplitude, V max is node allowable maximum voltage amplitude, P line is actual transmission active power of a power transmission line/channel, P line-max is line maximum allowable transmission active power, equipment capacity constraint that 0 is less than or equal to P _new_i ≤P _avail_i ;P _new_i is i new energy equipment actual grid-connected active power, P _avail_i is i new energy equipment current maximum available active power, stable boundary constraint that delta P is less than or equal to P _critical ;P _critical is system maintenance stable maximum bearable active power disturbance threshold, and then solving current maximum grid-connected capacity at the current moment; Step 4, trend prediction analysis, namely, predicting by combining data-driven direct prediction with model-driven rolling prediction, wherein the data-driven direct prediction is based on a historical margin time sequence, and is combined with predicted illumination and wind speed data, and the data is input into a neural network for calculation to obtain a capacity margin predicted value of every 15 minutes in the future 2 hours and a confidence interval thereof, and the model-driven rolling prediction takes predicted weather and load data at each moment in the future as input, and the dynamic modeling evaluation in the step 3 is subjected to rolling calculation to obtain a predicted capacity margin change trend of 15 minutes to 24 hours in the future; And 5, visualizing a diagram related to the capacity margin change through a Web interface, and implementing the alarm exceeding the preset threshold.
  2. 2. The method for online assessment of the grid-connectable capacity margin and trend of the new sand-gorboom energy resource base according to claim 1 is characterized in that in step 1, weather monitoring data comprise collected wind speed, wind direction, illumination intensity, ambient temperature, humidity, air pressure, sand concentration, satellite cloud image data, radar data and numerical weather forecast micro-scale correction data, wherein the grid monitoring data comprise voltage of each node of a power grid, phase angle, active/reactive power flow of a line/transformer, switching state, protection signals and system frequency, the new energy station monitoring data comprise real-time active/reactive power, machine end voltage, current, equipment internal temperature, fault codes, available capacity and control modes, the energy storage station monitoring data comprise real-time charge and discharge power, SOC, health state, maximum charge and discharge capacity and voltage current of an energy storage station and each battery cluster, and the management data comprise planned maintenance information, load prediction curves and market transaction plans.
  3. 3. The method for online evaluation of the grid-connectable capacity margin and trend of the sand-gossypable new energy base according to claim 1, wherein the space-time alignment in the step 2 uses the same time stamp of the PTP synchronous clock, and is unified with the implementation data of the same spatial reference system.
  4. 4. The method for online evaluation of the grid-connectable capacity margin and trend of the sand-gossypii new energy base according to claim 1 is characterized in that in the step 2, cleaning data adopts a state estimation residual analysis model or an isolated forest algorithm, and filling missing data adopts an interpolation method or a regression filling method or a deep learning generation method based on association relations.
  5. 5. The method for online evaluation of the grid-connectable capacity margin and trend of the sand-rich new energy base according to claim 1, wherein the key features in the step 2 include line load rate, voltage deviation rate, new energy permeability instantaneous value, and conversion coefficient of meteorological factor to output force.
  6. 6. The method for online evaluation of the grid-connectable capacity margin and trend of the new sandy-gossypable energy base according to claim 1, wherein the current margin is calculated in the step 3 by adopting a linear programming method, a quadratic programming method or an interior point method.
  7. 7. The method for online evaluation of the capacity margin and trend of the sandy new energy base capable of being connected with the grid as claimed in claim 1, wherein the map related to the capacity margin change in the step 5 comprises real-time capacity margin values, historical and predicted trend curves, power grid key section trend and margin distribution thermodynamic diagrams and weather information superposition diagrams, and the warning mode comprises acousto-optic, system popup windows and short messages.
  8. 8. An online evaluation system for the grid-connectable capacity margin and trend of a sand-gossypii new energy base is characterized in that the online evaluation method for the grid-connectable capacity margin and trend of the Sha Gehuang new energy base disclosed in any one of claims 1 to 7 is implemented and comprises a data acquisition layer, a data fusion and processing layer, an evaluation and prediction layer and an application and display layer, wherein the data acquisition layer acquires multi-source data; the data fusion and processing layer cleans, aligns and fuses the multi-source data and unifies the data; The evaluation and prediction layer comprises a dynamic evaluation module and a trend prediction module, wherein the dynamic evaluation module calculates the current capacity margin according to an online real-time update model, and the trend prediction module predicts the capacity margin change trend; The application and display layer performs visual display and super-threshold range alarming on capacity margin and trend.
  9. 9. The on-line evaluation system for the grid-connectable capacity margin and trend of the sand-gorboom new energy base according to claim 8, wherein the application and display layer is provided with a report generating module and an outward transmission interface, the report generating module automatically generates an evaluation report, and the outward transmission interface is used for pushing the margin evaluation and prediction result to a superior scheduling system, an automatic power generation control system, an energy storage coordination control system and an electric power transaction platform.

Description

Online evaluation method and system for grid-connectable capacity margin and trend of sand-gossypii new energy base Technical Field The invention belongs to the technical field of new energy grid connection, and particularly relates to an online evaluation technology of a new energy base grid-connectable capacity margin and a change trend thereof. Background Under the global energy transformation background, sha Gehuang areas become strategic places for the construction of centralized photovoltaic and wind power projects due to the wide land resources and rich illumination and wind energy resources. The China plans Sha Gehuang large-scale wind power photovoltaic base of hundreds of millions of kilowatts. However, grid-connected operation of these bases faces a series of world-class technical challenges, the core contradiction of which is the conflict between the strong uncertainty and volatility of the new energy output and the limited and rigid load carrying capacity of the grid itself. Traditional grid-connectable capacity assessment methods have revealed serious shortcomings in such scenarios. First, from the characteristics of resources and power grids, the new energy base in Sha Gehuang area has the characteristic of vivid 'two-high two-weak', namely 'two-high', namely high fluctuation of resources and high concentration of equipment. The regional weather conditions are complex and are easily influenced by sand storm, high temperature, strong radiation and severe temperature difference, so that the output power of the photovoltaic module is nonlinear along with illumination, temperature and sand coverage, the output force of the fan is obviously influenced by turbulent wind and extreme wind speed shear, and the power steep rise and fall of minute-level and hour-level become normal. Meanwhile, thousands of power generation equipment are densely arranged in the base, the fluctuation amplitude is amplified by the aggregation effect, and huge uncertainty on the 'source' side is formed. The two weaknesses are the weak power grid structure and the weak regulation capability. Sha Ge barren areas are usually located at the tail ends of power grids, grid structures are sparse, power transmission channels are limited, short-circuit capacity level is low, and voltage supporting capability is poor. Meanwhile, the local load level is low, enough conventional rotary standby and rapid adjustment resources (such as gas turbines and hydropower) are lacked, and the flexibility of the charge side and the storage side of the power grid is seriously insufficient. This coupling of the "strong wave source" to the "weak carrier network" causes the real-time safe operating boundaries of the network to become abnormally narrow and dynamically time-varying. Secondly, the existing assessment method has significant limitations, and is mainly characterized in three aspects of staticizing, off-line and one-sided. Firstly, the traditional method carries out off-line calculation based on typical days, typical scenes or worst conditions, and determines a fixed 'upper limit of installed capacity' or 'guaranteed output curve'. The method can not reflect the minute-level and hour-level dynamic changes of the actual output of the new energy and the real-time changes of the running state of the power grid (such as the withdrawal of a certain line due to overhaul, the random fluctuation of the load and the like). The result is either too conservative, resulting in a large amount of wind and light abandoning, and impaired economic benefit, or too aggressive, and the potential safety hazard of grid out-of-limit or even breakdown is buried. Secondly, depending on manual periodic (such as year, month and day) calculation, the response speed takes a "day" as a unit, and the decision requirement of the power grid real-time scheduling (taking a "minute" and a "second" as a unit) cannot be met. When an emergency meteorological event (such as rapid passing of a sand storm to cause sudden drop of photovoltaic output) is encountered, the offline planning is completely disabled, and a dispatcher can only perform emergency treatment empirically, so that the risk is extremely high. Third, conventional evaluations tend to focus on only a single stability limit, such as a thermal stability limit or voltage static safety constraint. However, sha Gehuang grid-connected capacity margin of the new energy base is a complex index affected by multi-dimensional, multi-time scale factor coupling. The method at least needs comprehensive consideration of a) static safety constraint (circuit/transformer thermal stability, node voltage upper and lower limits), b) transient stability constraint (influence of large-scale new energy off-grid on system frequency and voltage transient process), c) electric energy quality constraint (such as harmonic wave and flicker, particularly broadband oscillation risk caused by access of a large amount of power electronic equipment), d)