CN-122021349-A - Power system time sequence deduction method based on improved OPA model

CN122021349ACN 122021349 ACN122021349 ACN 122021349ACN-122021349-A

Abstract

The invention discloses a power system time sequence deduction method based on an improved OPA model, and belongs to the technical field of power system safety simulation. The method comprises the following technical steps of 1) constructing a three-layer deduction model of a cascade fault inner layer, a daily time sequence middle layer and an energy blocking outer layer, 2) constructing an input data set based on power grid topology, power generation information, a load curve and a new energy curve, 3) sequentially simulating dynamic behaviors of a power system under the multiple time scales of short-time faults, daily operation, long-term energy blocking and the like, and realizing full-time deduction of the power system from a small time level to a month level by fusing multi-source information of the power grid topology, the load curve, new energy output, primary energy storage, a supply strategy and the like, and 4) outputting load loss, load loss time period and power deduction results. The invention can effectively evaluate the load loss risk, the energy consumption and the supply demand of the power system in extreme scenes and provides scientific basis for emergency planning and toughness improvement of the power system.

Inventors

LI JIANKE
HAO JIANXIN
LUO SHAN
CHEN JINGJING
YANG YUAN
XU QIWEI
ZHANG HAITAO

Assignees

中国人民解放军陆军工程大学

Dates

Publication Date: 20260512
Application Date: 20260410

Claims (10)

1. The power system time sequence deduction method based on the improved OPA model is characterized by comprising the following steps of: s1, constructing a three-layer deduction model of a cascade fault inner layer, a solar time sequence middle layer and an energy blocking outer layer; S2, constructing an input data set based on power grid topology, power generation information, a load curve and a new energy curve; S3, determining input parameters, inputting the input parameters and an input data set into the three-layer deduction model, and sequentially executing outer layer, middle layer and inner layer circular deduction to realize multi-time scale power system behavior simulation and obtain a power system time sequence deduction result; And S4, outputting a power system time sequence deduction result, wherein the power system time sequence deduction result comprises a load loss time sequence curve, total load loss and energy storage change condition of each node.
2. The power system time sequence deduction method based on the improved OPA model according to claim 1, wherein in step S1, a three-layer deduction model of a cascade fault inner layer, a day time sequence middle layer and an energy blocking outer layer is constructed, comprising: S1-1, constructing a CASCADE fault inner layer deduction model, wherein the CASCADE fault inner layer deduction model considers an electric power breaking strategy, takes a hit line and a hit transformer substation as initial faults, and simulates the CASCADE fault condition of an electric power network within an hour time scale by using an improved CASCADE model; S1-2, constructing an intra-day time sequence middle layer deduction model, wherein the intra-day time sequence middle layer deduction model considers the intra-day load demand and the time sequence change of new energy, simulates the operation trend of an electric power network in an intra-day hour time scale, and calculates the consumption condition of primary energy reserves of a thermal power plant, and the new energy comprises wind power and photovoltaic; and S1-3, constructing an energy blocking outer layer deduction model, wherein the energy blocking outer layer deduction model simulates the process of energy supply limitation, storage capacity reduction and strategic energy transfer of the next time of a multi-day and month time scale.
3. The method according to claim 2, wherein in step S1-3, the primary energy supply limitation includes a scenario of implementing blockage of a coal, gas, and fuel inlet, and the strategic energy transfer includes a strategy of supplying primary energy for transportation of a train or tank truck via a transportation network.
4. The method for power system time sequence deduction based on the improved OPA model according to claim 3, wherein in step S2, an input data set is constructed based on the power grid topology, the power generation information, the load curve and the new energy curve, comprising: S2-1, acquiring and inputting power network model information comprising a transformer substation, a power transmission line and a station-line topological relation; S2-2, obtaining and inputting maximum load power information of each load transformer substation; S2-3, acquiring and inputting typical daily load curves of the whole network load of different months; and S2-4, obtaining and inputting typical active output curves of wind power and photovoltaics in different months.
5. The method for power system time sequence deduction based on the improved OPA model according to claim 4, wherein in step S3, input parameters are determined, the input parameters and the input data set are input into the three-layer deduction model, and loop deduction of an outer layer, a middle layer and an inner layer is sequentially executed, so as to realize behavior simulation of a multi-time scale power system, and obtain a power system time sequence deduction result, which comprises the following steps: S3-1, determining input parameters including total days of time sequence deduction simulation, primary energy import lockout strategy, primary energy initial reserve of each thermal power plant and power attack events appointed by users; Step S3-2, initializing a cyclic variable d=1 of the energy blocking outer layer; step S3-3, initializing a cyclic variable h=0 of the middle layer in the time sequence in the day; S3-4, calculating the power flow of each branch in the power network at the current moment according to an improved maximum flow algorithm, and counting the load loss at the current moment; step S3-5, checking whether a broken event exists at the current moment, if yes, jumping to step S3-6 to conduct cascade fault deduction, and if not, jumping to step S3-8 to calculate primary energy reserve consumption; s3-6, calculating cascading faults which will occur to the power network under the initial attack event according to the improved CASCADE model; S3-7, according to the deduced cascading failure set, stopping the failure branch or the transformer substation, updating a power network model, recalculating power network flow after topology structure change, and counting the load loss after cascading failure; S3-8, calculating the primary energy consumption in the current period according to the active output state of each thermal power plant at the current moment, and subtracting the consumption from the current primary energy reserve to update the primary energy reserve state of each thermal power plant; step S3-9, updating a circulating variable of a middle layer in a daily time sequence, setting h=h+1, judging whether h is equal to 24, if yes, jumping to step S3-10 to update an energy blocking outer layer deduction model variable, otherwise, jumping to step S3-4 to calculate the power network flow at the next moment; step S3-10, updating a circulation variable of the energy blocking outer layer, and setting d=d+1; Step S3-11, judging whether the current days d is larger than the total days of the time sequence deduction simulation, if so, outputting a time sequence deduction simulation result which comprises a time sequence curve of each node without load, total load loss and energy reserve change condition, otherwise, jumping to step S3-12 to judge whether to implement primary energy supply; Step S3-12, judging whether to implement primary energy supply, if yes, jumping to step S3-13 to update primary energy storage capacity, otherwise, returning to step S3-3 to initialize a circulating variable of the middle layer in the time sequence in the day; and S3-13, distributing primary energy inlet quantity according to the proportion of the residual storage capacity of primary energy of the thermal power plant, updating the primary energy storage quantity of each thermal power plant, and returning to the step S3-3 to initialize the circulating variable of the middle layer in the time sequence in the day after updating.
6. The method according to claim 5, wherein in step S3-4, calculating the load flow of each branch in the power network at the current time according to the modified maximum flow algorithm, and counting the load loss at the current time comprises: S3-4-1, constructing a flow network model of a power network, abstracting a power plant node as a source point, abstracting a load transformer station node as a sink point, abstracting a power transmission line as a side, and using a thermal stability limit of the line as the capacity of the side to form a multi-source multi-sink network; Step S3-4-2 of introducing a virtual source point S and a virtual sink point t, converting the multi-source point multi-sink network into a single-source point single-sink network by adding a directed edge (S, g) from the virtual source point S to each power plant node g, the capacity c (S, g) of which is set as the available power generation capacity of the power plant g Adding a directed edge (l, t) from each load node l to the virtual sink t, wherein the capacity c (l, t) is set as the actual active load demand of the load node l at the current moment ; Step S3-4-3, executing a maximum flow algorithm to solve the maximum feasible flow from the virtual source point S to the virtual sink point t The maximum feasible flow is equal to the minimum cut, and the maximum flow algorithm solves for the maximum feasible flow that meets the following flow conservation and capacity constraints: ; ; Where f (i, j) means traffic on edge (i, j) from node i to node j; Means that the feasible flow from the virtual source point s to the virtual sink point t, s means the virtual source point representing the sum of all power plants, t means the virtual sink point representing the sum of all loads, and the flow conservation equation means that the net flow flowing out of the virtual source point s is equal to the net flow flowing into the virtual sink point t, and the flow of the intermediate node is kept balanced; c (i, j) means the capacity of the edge (i, j), E means the set of all edges in the network, and the capacity constraint equation indicates that the flow of any edge cannot exceed the capacity of the edge; The flow of the edge between the load transformer substation and the virtual sink is the actual power supply at the moment, and the difference between the actual power supply and the actual active load demand is the load loss of the load transformer substation; the flow of the edge between the virtual source point and the power plant node is the active output of the power plant at the moment; S3-4-4, identifying the position of the minimum cut, and determining the system state by judging the type of the minimum cut, wherein the system state is determined to be insufficient in power supply when the minimum cut is in an edge set between a virtual source point and a power plant node; step S3-4-5, counting the load losing quantity of each node according to the maximum flow calculation result, wherein the actual power supply of the load node l is equal to the flow f (l, t) on the side (l, t), and the load losing quantity of the load node l Calculated as follows: ; Wherein, the Meaning the amount of load lost by load node l at time h; f (l, t) means the flow of the edge from the load node l to the virtual sink t, namely the actual power supply; Step S3-4-6, calculating total load loss of the system : ; Wherein, the Meaning the total load loss of the system at time h; Meaning a set of all load nodes in the system; S3-4-7, outputting the flow f (i, j) of each branch, the load loss of each load node, the total load loss of the system and the actual output of each power plant Minimal cut position information.
7. The power system time sequence deduction method based on the improved OPA model according to claim 6, wherein in the step S3-4-3, the maximum flow algorithm is the Ford-Fulkerson algorithm.
8. The power system timing deduction method based on the improved OPA model according to claim 7, wherein in step S3, the multi-time scale power system behavior simulation includes: Simulating a cascading failure propagation process of a time scale in an hour in a cascading failure inner layer; simulating the operation trend of the power network in the time sequence of the day in the time scale of the hour in the day, wherein the time scale of the hour in the day is 24 hours; Simulating an energy supply limited process of a time scale of a plurality of days and a month in an energy blocking outer layer.
9. The power system time sequence deduction method based on the improved OPA model according to claim 5, wherein in step S3-8, the primary energy reserve consumption condition of the thermal power plant is calculated, comprising: according to the classification of the coal-fired power plant, the gas power plant and the fuel oil power plant, respectively calculating the fuel coal consumption, the gas consumption or the fuel oil consumption corresponding to the active output of each type of power plant in the current period; the primary energy consumption and the active power output are obtained by conversion based on the unit energy consumption coefficient of the power plant.
10. The power system timing deduction method based on the improved OPA model according to claim 9, wherein in step S3-13, updating the primary energy storage amount of each thermal power plant includes: Determining the proportion of the primary energy remaining storage capacity of each thermal power plant to the total primary energy remaining storage capacity of all thermal power plants; Distributing the total amount of the distributable primary energy inlets to each thermal power plant according to the proportion; And adding the primary energy inlet amount obtained by distribution to the current primary energy storage amount of each thermal power plant to finish updating.

Description

Power system time sequence deduction method based on improved OPA model Technical Field The invention belongs to the technical field of safety analysis and simulation of power systems, and particularly relates to a power system time sequence deduction method based on an improved OPA model. Background The inter-regional interconnected power grid has been developed into one of the most complex artificial industrial networks at present, and in recent years, a plurality of blackouts which occur at home and abroad are all caused by power grid cascading failures, so that domestic and foreign scholars increasingly pay attention to the study of the power grid cascading failures and the blackout propagation mechanism, and various power grid cascading failure models are proposed, wherein the most attention is not paid to an OPA model, and the OPA (self-organizing critical theory) model is called as a 'blackout model' or a 'self-organizing critical power system model'. The model is based on the theory of self-organization criticality. This theory suggests that a complex power system (e.g., a power grid) spontaneously evolves to a critical state under external continuous, slow "driving" (e.g., load increase) and internal local, fast "dissipating" (e.g., line overload trip) interactions. In this critical state, a small disturbance (such as a line fault) may trigger a series of chain reactions, resulting in a large-scale power outage. The OPA model is essentially a theoretical framework and simulation tool that reveals the risk of spontaneous breeding blackouts in the grid due to its inherent complexity and continued growth. The traditional OPA model is based on a self-organizing critical theory, and is used for simulating line outage and linkage overload through a fast dynamic process and simulating system capacity and load increase through a slow dynamic process. However, the model does not fully consider new energy time sequence fluctuation, primary energy reserve and supply limitation and energy blocking and transferring strategies under a long time scale, and is difficult to truly reflect the time sequence evolution behavior of the power system in a complex environment. The OPA model divides the grid evolution process into two interacting time scales, namely "fast dynamics" (simulating fault propagation) and "slow dynamics" (simulating system upgrades). The related processes of line outage and linkage overload can be known through a fast dynamic process, and the increase of line capacity, the increase of load level and the improvement of system power generation capacity can be simulated to a certain extent through a slow dynamic process. The combination of "fast dynamics" and "slow dynamics" will explain the risk of a grid blackout. The OPA model simulates the long term dynamics of the grid through one cycle, and this process reveals why there is always the risk of a large-scale outage occurring, driven by complexity and continuous growth, regardless of how the grid is reinforced. The solving core of the OPA model is a direct current power flow optimization problem, and the limitations of the traditional OPA model are as follows: Static assumption-it is generally assumed that the load level and the generated power are fixed during fast dynamics, ignoring daily and seasonal timing fluctuations (especially wind power, photovoltaic randomness). Neglecting energy constraints, namely only paying attention to physical constraints (line capacity and generator output) of the power grid, and not considering storage, consumption and replenishment limits of primary energy (coal, gas and oil) of the power plant. The model is lack of active strategy simulation, more reflects the self-organized evolution of the system, and is difficult to simulate artificial active attack (electric power attack) or macroscopic energy blocking strategies and long-term influence thereof. The time scale coupling is insufficient, and although the speed and the speed are differentiated, the coupling relation between the intra-hour-level daily operation and the month-level strategic reserve change is characterized as weaker. Therefore, a time sequence deduction method capable of integrating multiple time scales, multiple energy types and multiple fault modes is needed to improve the simulation reliability and decision support capability of the power system in extreme scenes. Disclosure of Invention Aiming at the defects of the conventional OPA model, the invention provides a power system time sequence deduction method based on an improved OPA model, and the three-layer deduction structure of a cascade fault inner layer, a daily time sequence middle layer and an energy blocking outer layer is constructed, so that the behavior simulation of the power system from a small time scale to a month scale is realized, and the simulation reliability and the decision support capability of the power system under extreme scenes can be improved. In order to achieve t