CN-121983977-A - Smart grid-based electric power energy consumption early warning method and system

CN121983977ACN 121983977 ACN121983977 ACN 121983977ACN-121983977-A

Abstract

The invention discloses an electric power energy consumption early warning method based on a smart grid, which comprises the following steps of collecting power grid operation data, updating a power grid state variable in a digital twin model by utilizing a digital twin model and a state estimation algorithm based on the power grid operation data to enable the digital twin model to be consistent with a physical power grid, analyzing the power grid state variable in the digital twin model in real time, identifying a power unbalance composite risk of coincidence of rapid fluctuation of a power supply side and sudden impact of a demand side, starting simulation of a preset emergency plan aiming at the identified power unbalance composite risk, evaluating the influence of the emergency plan on power grid stability, power gap compensation capability and operation side effect, selecting an optimal emergency plan according to an evaluation result, generating an executable load adjustment or power scheduling instruction, and transmitting the executable load adjustment or power scheduling instruction to intelligent execution equipment through high-speed communication.

Inventors

Mo Wangyi
HU RUI
WEI SHILEI
CHENG XIAOFEI
Ji Peichen
LI JIAN
LI YUAN
QIN LIUYUN
ZENG JINFU

Assignees

南方电网互联网服务有限公司

Dates

Publication Date: 20260505
Application Date: 20251210

Claims (10)

1. The electric power energy consumption early warning method based on the intelligent power grid is characterized by comprising the following steps of: Step A, collecting power grid operation data; Step B, based on the power grid operation data, updating a power grid state variable in the digital twin model by using a digital twin model and a state estimation algorithm, so that the digital twin model is consistent with a physical power grid; Step C, analyzing the state variables of the power grid in the digital twin model in real time, and identifying the power unbalance composite risk of the superposition of the rapid fluctuation of the power supply side and the sudden impact of the demand side; step D, aiming at the identified power unbalance composite risk, starting simulation of a preset emergency plan, and evaluating the influence of the emergency plan on the stability, the power gap compensation capability and the operation side effect of the power grid; and E, selecting an optimal emergency plan according to the evaluation result, generating an executable load adjustment or power supply scheduling instruction, and transmitting the executable load adjustment or power supply scheduling instruction to intelligent execution equipment through high-speed communication.
2. The smart grid-based power consumption pre-warning method according to claim 1, wherein the step D includes: the production line local control system monitors the operation parameters of production equipment, and generates and broadcasts a process criticality label and a criticality effective time window; The digital twin system receives the process criticality label and the criticality effective time window, and maintains a load criticality state table; The digital twin system adjusts the production influence weight of the load in the emergency plan simulation according to the load criticality state table; The digital twin system selects the emergency plan based on the adjusted production influence weight and evaluates the influence of the emergency plan on the stability, the power gap compensation capability and the operation side effect of the power grid.
3. The smart grid-based power consumption pre-warning method according to claim 1, wherein the step C includes: Carrying out real-time high-frequency sampling and feature extraction on output power data flow of a distributed power supply and consumption power data flow of an instantaneous load, calculating instantaneous power change rate, change direction, duration and change amplitude, analyzing instantaneous waveforms of voltage and current of key nodes of a power grid, extracting dynamic features of frequency, amplitude and phase angle, and identifying subcritical transient events; when the subcritical transient event is detected, calculating a risk contribution value according to the instantaneous power change rate, the duration and the influence degree on the local grid voltage and current of the subcritical transient event, and accumulating the risk contribution value into a risk accumulation meter; monitoring the value of the risk accumulation measuring device, and judging that the power unbalance composite risk occurs when the value exceeds an aggregation risk threshold value within a preset time, wherein the aggregation risk threshold value is dynamically adjusted according to the whole load level, the available standby capacity and the local power grid topological structure of the power grid; Analyzing the geographical position and the grid connection relation of the subcritical event causing the risk accumulation, identifying the contribution of the distributed power supply output reduction and the transient load power increase to the risk accumulation, and positioning a local area or feeder line where the composite risk occurs.
4. The smart grid-based power consumption pre-warning method according to claim 1, wherein the step C includes: Collecting instantaneous waveform data of distributed power supply output power, instantaneous load consumption power, power grid key node voltage and current in the digital twin model in real time, and extracting dynamic characteristics of the instantaneous waveform data, wherein the dynamic characteristics comprise instantaneous power change rate, change direction, duration, change amplitude, frequency, amplitude and phase angle dynamic characteristics; comparing the extracted dynamic characteristics with a currently known composite risk characteristic mode library, and marking the dynamic characteristics as potential novel risk characteristics if the matching degree is lower than a preset threshold value; performing cluster analysis on the potential novel risk features to identify a new composite risk feature pattern; adding the new composite risk feature pattern into a composite risk feature pattern library, and distributing initial risk weights to the composite risk feature patterns according to potential influences of the composite risk feature patterns on the stability of the power grid; Continuously monitoring power grid operation data, and dynamically adjusting the risk weight of the new composite risk feature mode according to the frequency and the severity of a power unbalance event caused by the new composite risk feature mode in an actual power grid when the new composite risk feature mode is identified again; and identifying the power unbalance composite risk of the power supply side rapid fluctuation and the demand side sudden impact coincidence in real time based on the updated composite risk characteristic mode library and the risk weights corresponding to the modes.
5. The smart grid-based power consumption pre-warning method according to claim 1, wherein the step C includes: Extracting dynamic characteristics of the instantaneous waveform data, wherein the dynamic characteristics comprise instantaneous power change rate, change direction, duration, change amplitude, frequency, amplitude and phase angle dynamic characteristics; Identifying whether a specific harmonic component, an inter-harmonic component or a high-frequency oscillation mode introduced by power electronic equipment or high-frequency operation load exists in the power grid according to the instantaneous waveform data; Monitoring voltage and current waveforms of key nodes of a power grid, analyzing whether the waveforms have oscillation modes similar to the natural resonant frequency of the power grid or the response frequency of the existing control system, and calculating the damping ratio and the growth rate of the oscillation modes; Comparing the harmonic component, the oscillation mode, the damping ratio, and the growth rate to a library of coupling modes of known device operating characteristics and control strategies; if the comparison result shows that the characteristics matched with the known coupling modes exist, calculating a current risk contribution value; If the comparison result shows that a harmonic component or an oscillation mode which is not matched with the known coupling mode but has a high risk characteristic exists, marking the harmonic component or the oscillation mode as a potential novel coupling risk characteristic; performing cluster analysis on the potential novel coupling risk characteristics to identify a novel coupling risk mode; Adding the new coupling risk mode into the coupling mode library, and distributing initial risk weights to the new coupling risk mode according to potential influences of the new coupling risk mode on the stability of the power grid; continuously monitoring power grid operation data, and dynamically adjusting the risk weight of the new coupling risk mode according to the frequency and the severity of a power unbalance event caused by the new coupling risk mode in an actual power grid when the new coupling risk mode is identified again; And identifying the power unbalance composite risk of the power supply side rapid fluctuation and the sudden impact coincidence of the demand side based on the updated coupling mode library and the risk weight.
6. The smart grid-based power consumption pre-warning method according to claim 1, wherein the step C includes: acquiring instantaneous output data of each distributed power supply, instantaneous consumption power data of each instantaneous high impact load and instantaneous waveform data of voltage, current and frequency of key nodes of a power grid in real time; identifying an output fluctuation mode of the distributed power supply according to the instantaneous output data, wherein the output fluctuation mode comprises a specific output fluctuation mode of the distributed power supply type; identifying an impact pattern of the transient high impact load from the transient power consumption data, the impact pattern comprising an impact pattern specific to a transient high impact load type; Analyzing the coincidence degree of the distributed power supply output fluctuation mode and the transient high impact load impact mode in time, the electrical distance in space and the influence degree of the two on the voltage, the current and the frequency of a local power grid; When the output fluctuation mode of the distributed power supply and the impact mode of the instantaneous high impact load coincide in time and the influence of the output fluctuation mode and the impact mode of the instantaneous high impact load on the local power grid exceeds a preset threshold after being overlapped, judging that the power unbalance composite risk is caused by the complex coupling of heterogeneous equipment; And according to the judged power unbalance composite risk, positioning a main distributed power supply and a transient high impact load which cause the power unbalance composite risk, and evaluating the potential influence of the power unbalance composite risk on the overall stability of the power grid.
7. The smart grid-based power consumption pre-warning method according to claim 1, wherein the step a includes: When a load is started for the first time or a specific transient operation is executed, acquiring original high-frequency voltage and current data streams of high-precision measuring equipment nearby the load; analyzing the original waveform in real time through a characteristic extraction module, and extracting electromagnetic characteristic signatures including high-frequency harmonic component spectrums, rising edge characteristics and falling edge characteristics of transient current impact and influence modes on local voltage; Dynamically storing the electromagnetic characteristic signature in a transient characteristic signature library, and updating along with the load running state; performing similarity matching on transient electromagnetic features in the high-frequency data stream transmitted in real time and known feature signatures in the transient feature signature library; If the matching degree of the real-time features and the stored signature exceeds a preset threshold, identifying the real-time features as background noise caused by specific load and distinguishing the real-time features from the real-time state change of the power grid; Generating a predicted interference waveform in real time according to the identified load electromagnetic characteristic signature; And carrying out differential processing on the measured original waveform and the predicted interference waveform to obtain a pure power grid transient response waveform without equipment self interference.
8. The smart grid-based power consumption pre-warning method according to claim 6, wherein the step of collecting in real time instantaneous output data of each distributed power source, instantaneous consumption power data of each instantaneous high impact load, and instantaneous waveform data of voltage, current, and frequency of a key node of the power grid in the digital twin model comprises: An intelligent power measurement unit is deployed, and the intelligent power measurement unit has a self-calibration function and a fault diagnosis function; Before data acquisition, the intelligent power measurement unit performs state check and precision calibration; when the intelligent power measurement unit finds that the data is abnormal, the intelligent power measurement unit performs cross verification by using the data of the adjacent measurement units; And according to the cross verification result, performing missing data interpolation by utilizing the data of the adjacent measuring units so as to ensure the reliability and the integrity of the instantaneous output data, the instantaneous consumption power data and the instantaneous waveform data.
9. The smart grid-based power consumption pre-warning method according to claim 6, wherein the step of collecting in real time instantaneous output data of each distributed power source, instantaneous consumption power data of each instantaneous high impact load, and instantaneous waveform data of voltage, current, and frequency of a key node of the power grid in the digital twin model comprises: An intelligent sensor is deployed, and the intelligent sensor has multichannel synchronous acquisition capability and high dynamic range; The intelligent sensor synchronously collects instantaneous output data of each distributed power supply, instantaneous consumption power data of each instantaneous high impact load and instantaneous waveform data of voltage, current and frequency of key nodes of the power grid with nanosecond precision; The intelligent sensor monitors the instantaneous output data, the instantaneous consumption power data and the instantaneous waveform data in real time and detects the instantaneous power change rate, the voltage current waveform distortion rate or the frequency change rate; When the instantaneous power change rate, the voltage-current waveform distortion rate or the frequency change rate exceeds a preset dynamic range threshold value, the intelligent sensor is automatically switched to a high-sensitivity acquisition mode and an extended range mode; the intelligent sensor improves sampling frequency and data resolution in the high-sensitivity acquisition mode so as to capture fine changes in the transient process; the intelligent sensor adjusts the measurement range under the extended range mode so as to adapt to the collection of extreme values; the intelligent sensor carries out time stamp marking and event type marking on the collected high-sensitivity wide-range transient data, and transmits the transient data marked by the time stamp marking and the event type marking to a digital twin system.
10. Electric power energy consumption early warning system based on smart power grids, characterized by comprising: The data acquisition module is used for acquiring power grid operation data, and the power grid operation data is acquired at sub-second frequency through high-precision power measurement equipment; The digital twin updating module is used for updating the state variable of the power grid in the digital twin model by utilizing a digital twin model and a state estimation algorithm based on the power grid operation data so as to keep the digital twin model consistent with a physical power grid; The risk identification module is used for analyzing the state variables of the power grid in the digital twin model in real time and identifying the power unbalance composite risk of the superposition of the rapid fluctuation of the power supply side and the sudden impact of the demand side; The plan simulation evaluation module is used for starting microsecond simulation of a preset emergency plan aiming at the identified power unbalance composite risk and evaluating the influence of the emergency plan on the stability of a power grid, the power gap compensation capability and the operation side effect; and the instruction generation issuing module is used for selecting an optimal emergency plan according to the evaluation result, generating an executable load adjustment or power supply scheduling instruction and issuing the executable load adjustment or power supply scheduling instruction to the intelligent execution equipment through high-speed communication.

Description

Smart grid-based electric power energy consumption early warning method and system Technical Field The application relates to the technical field of smart grids, in particular to a power energy consumption early warning method and system based on a smart grid. Background In modern industrial parks, stable supply and efficient utilization of electric power are key to ensuring production continuity and reducing operating costs. To achieve this goal, a campus typically deploys a set of power consumption pre-warning systems. When these systems were initially designed, they were primarily focused on monitoring overall power usage and alerting them to a preset upper power limit to avoid overload operation or additional costs. With the technical progress, some systems also introduce a prediction function based on past electricity consumption conditions and production arrangement, and hope to identify electricity consumption peaks in advance, so as to guide enterprises to adjust production and realize electricity consumption stability. However, with the upgrade of industrial structures in industrial parks, in particular the introduction of new high-energy consumption, instant-start devices, and the large-scale application of distributed renewable energy sources, traditional early warning mechanisms face serious challenges. These new conditions make the supply and demand balance of the grid more complex and difficult to predict, especially in very short time scales, where the grid may suffer from extreme situations where the supply capacity is suddenly reduced and the demand for electricity is suddenly increased. For example, in cloudy weather, the output power of distributed photovoltaic power generation may drop dramatically in tens of seconds, while at the same time, high energy consumption devices may start up instantaneously, creating a huge electricity demand. The superposition of these two events, i.e. the sudden decrease in power supply capacity and the sudden increase in power demand, together form an extremely large and steep power gap, all of which need to be filled instantaneously by the main network. The data analysis and judgment logic of the existing early warning system is generally based on the data change trend of the minute level, and the existing early warning system is completely out of the way for the compound extreme event occurring on the second time scale. When the program in the background of the system is still analyzing data for the last minute, the physical protection device at the front end may have tripped because it cannot withstand a large current surge, causing a power outage in the campus area. The inherent time delay causes that the traditional early warning mechanism cannot respond in time before the physical protection device acts, so that the meaning of early warning is lost, and the power failure accident caused by serious overrun of instantaneous power cannot be effectively avoided. Disclosure of Invention In order to overcome the defects of the prior art, the application discloses a smart grid-based power energy consumption early warning method, which aims to solve the technical problems that when the existing power energy consumption early warning system is applied to industrial parks in large scale with novel high energy consumption, instantaneous starting equipment introduction and distributed renewable energy sources, the composite extreme event which occurs on the second-level time scale and occurs simultaneously with the sudden increase of power consumption requirement cannot be effectively treated, early warning failure is caused, and power failure accidents caused by serious overrun of instantaneous power cannot be avoided. In a first aspect, the application discloses a smart grid-based power energy consumption early warning method, which comprises the following steps: Step A, collecting power grid operation data; Step B, based on the power grid operation data, updating a power grid state variable in the digital twin model by utilizing a digital twin model and a state estimation algorithm, so that the digital twin model is consistent with a physical power grid; Step C, analyzing the state variables of the power grid in the digital twin model in real time, and identifying the power unbalance composite risk of the superposition of the rapid fluctuation of the power supply side and the sudden impact of the demand side; step D, aiming at the identified power unbalance composite risk, starting simulation of a preset emergency plan, and evaluating the influence of the emergency plan on the stability, the power gap compensation capability and the operation side effect of the power grid; And E, selecting an optimal emergency plan according to the evaluation result, generating an executable load adjustment or power supply scheduling instruction, and transmitting the executable load adjustment or power supply scheduling instruction to the intelligent execution equipment throug