CN-121998297-A - Power grid emergency material dispatching double-layer optimization method and system under extreme weather

CN121998297ACN 121998297 ACN121998297 ACN 121998297ACN-121998297-A

Abstract

The invention discloses a double-layer optimization method and a double-layer optimization system for power grid emergency material dispatching under extreme weather. Firstly, acquiring real-time meteorological monitoring data and vulnerability characteristics of power grid equipment, calculating a dynamic time sensitivity coefficient, generating material demand priority parameters by combining disaster point load grades, secondly, constructing a double-layer material scheduling model, wherein the upper layer aims at maximizing key load recovery indexes, introducing constraints to realize degradation supply under resource limitation, the lower layer aims at minimizing comprehensive scheduling cost, realizing time-space synchronization of dynamic risk avoidance and people and goods through the constraints, and finally, adopting a double-layer iteration solving strategy based on interlayer interaction, utilizing an upper-layer topology to construct a mask to execute dimension reduction locking and dynamic interception on lower-layer flow, and accessing real-time meteorological and road condition data to execute online rolling correction in the solving process. The invention solves the problem that the traditional scheduling lacks environment perception and supply and demand flexibility, and improves the real-time adaptability, decision robustness and cooperative efficiency of power grid rush repair under extreme weather.

Inventors

YAN HAIOU
LI GANG
LI XIAO
RAO YUFEI
LI HAIFENG
LIANG GANG
MA LEI
SONG XIAOYAN
LIU YANG
WEI SHICHAO

Assignees

国网河南省电力公司电力科学研究院
华北电力大学（保定）

Dates

Publication Date: 20260508
Application Date: 20251222

Claims (10)

1. The double-layer optimization method for power grid emergency material dispatching under extreme weather is characterized by comprising the following steps: the method comprises the steps of 1, acquiring real-time meteorological monitoring data of a target area and vulnerability characteristic data of power grid equipment, calculating dynamic time sensitivity coefficients of various materials according to the real-time meteorological monitoring data, and generating material demand priority parameters by combining disaster point load levels; Step 2, obtaining road network basic data to construct a road network impedance matrix and constructing a double-layer material scheduling model, wherein the double-layer material scheduling model comprises an upper layer model taking an upper layer topology decision variable as an optimization object and a lower layer model taking a lower layer flow distribution variable as an optimization object, and the upper layer model and the lower layer model are coupled through the material demand priority parameter and the road network impedance matrix; And step 3, solving the double-layer material scheduling model by adopting a double-layer iteration solving strategy, and executing online rolling correction in the solving iteration process, wherein the method comprises the steps of acquiring the latest real-time meteorological monitoring data and road network real-time road condition data, updating the material demand priority parameter and the road network impedance matrix, and executing subsequent iteration based on the updated material demand priority parameter and the road network impedance matrix until an optimal material scheduling scheme is output.
2. The method for double-layer optimization of power grid emergency material scheduling under extreme weather according to claim 1, wherein in step 1: The real-time meteorological monitoring data comprise meteorological strength values aiming at the current meteorological disaster type in a target area, wherein the meteorological strength values comprise wind speed grades, rainfall or icing thickness; The power grid equipment vulnerability characteristic data comprises a pre-constructed equipment vulnerability matrix, wherein the equipment vulnerability matrix comprises vulnerability weight values of different types of emergency materials under different meteorological disaster types.
3. The method for optimizing the emergency material scheduling of the power grid under extreme weather according to claim 2, wherein in the step 1, the specific process of calculating the dynamic time sensitivity coefficient and generating the material demand priority parameter comprises the following steps: acquiring a reference time sensitivity coefficient, a preset meteorological strength safety threshold value, a historical meteorological extremum and a preset regulating factor of various materials; Extracting a current weather intensity value in real-time weather monitoring data, calculating the amplitude of the current weather intensity value exceeding the weather intensity safety threshold value, and calculating the ratio of the amplitude to the historical weather extremum to obtain a weather severity ratio; Carrying out continuous multiplication calculation on the meteorological severity proportion, the adjusting factor and a corresponding vulnerability weight value in the equipment vulnerability matrix to generate a nonlinear drift increment; correcting the reference time sensitivity coefficient by utilizing the nonlinear drift increment to obtain a dynamic time sensitivity coefficient which dynamically changes along with the current meteorological strength value; And determining a corresponding load level coefficient based on the load level of the disaster point, calculating the product of the dynamic time sensitivity coefficient and the load level coefficient, and taking the obtained product as the material demand priority parameter.
4. The method for double-layer optimization of power grid emergency material scheduling under extreme weather according to claim 1, wherein in the step 2, the upper layer model maximizes a key load recovery index as an objective function, and the constraint conditions include: The supply and demand guarantee flexible constraint is that products of guarantee coefficients corresponding to the load levels of the disaster points are calculated for each disaster point and each type of material, so that the minimum guarantee amount is obtained, and the sum of lower-layer flow distribution variables of all reserve points transported to the disaster points is constrained to be not lower than the minimum guarantee amount; The physical constraint of reserve capacity, namely, for each reserve point and each type of material, constraining the sum of lower layer flow distribution variables called by the reserve point to all disaster points to be not more than the physical stock quantity of the reserve point; When the upper topological decision variable indicates that a reserve point is connected with a disaster point, the corresponding actual transportation time in the road network impedance matrix is constrained to not exceed the maximum allowable delay time of the materials; constraint of the transport capacity resource pool, namely dividing lower layer flow distribution variables on all paths by the rated load of a single vehicle and rounding up to obtain the number of vehicles required, wherein the sum of the number of vehicles required by the constraint whole network does not exceed the maximum available vehicle threshold; Establishing a logic locking relation of an upper layer topology decision variable to a lower layer flow distribution variable, and forcing the corresponding lower layer flow distribution variable to be zero when the upper layer topology decision variable is zero.
5. The method for double-layer optimization of power grid emergency material scheduling under extreme weather according to claim 1, wherein in the step 2, the lower layer flow distribution model uses a comprehensive scheduling cost as an objective function, and the comprehensive scheduling cost and the corresponding constraint condition specifically include: The comprehensive scheduling cost is obtained by accumulating material transportation variable cost, vehicle fixed calling cost and weather congestion punishment items, wherein the weather congestion punishment items are continuous products of weather congestion punishment coefficients, delay increment in a road network impedance matrix and the lower layer flow distribution variable; An upper layer decision anchor constraint that restricts the corresponding lower layer flow distribution variable to allow a non-zero value to be taken only when the upper layer topology decision variable indicates that a connection is established; The supply and demand balance base line constraint is that the sum of all lower layer flow distribution variables pointing to each disaster-stricken point is constrained to be not lower than the minimum guarantee amount obtained by the product of the material demand of the disaster-stricken point and the guarantee coefficient corresponding to the load level of the disaster-stricken point; Physical constraint of path capacity, namely, constraint of the maximum traffic capacity of each transport path under the current meteorological condition of the path by using a lower layer flow distribution variable on the path; and (3) material-rush repair cooperative constraint, namely acquiring the expected time of the rush repair team reaching each disaster point, and constraining the absolute value of the difference value between the corresponding actual transportation time in the road network impedance matrix and the arrival time of the rush repair team not to exceed a preset cooperative time threshold.
6. The method for optimizing power grid emergency material scheduling double-layer under extreme weather according to claim 1, wherein in the step 3, the specific implementation process of the double-layer iterative solution strategy comprises the following steps: Inputting each upper-layer topological decision variable generated by the upper-layer model into the lower-layer model as a constant, and constructing a search space mask based on the upper-layer topological decision variables; In the solving process of the lower model, performing dimension-reducing locking and dynamic cutting operation, forcibly locking a lower-layer flow distribution variable corresponding to a path which is not established by the upper-layer topology decision variable indication to be zero by utilizing the search space mask; and obtaining an optimal lower layer flow distribution variable and a corresponding minimum comprehensive scheduling cost output by the lower layer model, feeding back the optimal lower layer flow distribution variable and the corresponding minimum comprehensive scheduling cost to the upper layer model to calculate the fitness value of the current upper layer solution, and driving the upper layer model to carry out next round of iterative updating.
7. The method for optimizing power grid emergency material scheduling double-layer under extreme weather according to claim 1, wherein in the step 3, the specific trigger logic of the online rolling correction comprises: the method comprises the steps of presetting an iteration update period threshold, monitoring the current iteration times in real time in the process of solving the iteration, triggering data acquisition and parameter update operation when the iteration times reach integer multiples of the iteration update period threshold, reserving individual position information of a current population after the update operation is finished, and continuously executing iteration solving of the next stage based on updated material demand priority parameters and a road network impedance matrix.
8. Power grid emergency material scheduling double-layer optimization system under extreme weather, which is characterized by comprising: The parameter dynamic generation module is used for acquiring real-time meteorological monitoring data of a target area and vulnerability characteristic data of power grid equipment, calculating dynamic time sensitivity coefficients of various materials according to the real-time meteorological monitoring data, and generating material demand priority parameters by combining disaster point load grades; The model construction and coupling module is used for acquiring road network basic data to construct a road network impedance matrix and constructing a double-layer material scheduling model, wherein the double-layer material scheduling model comprises an upper layer model taking an upper layer topological decision variable as an optimization object and a lower layer model taking a lower layer flow distribution variable as an optimization object, and the upper layer model and the lower layer model are coupled through the material demand priority parameter and the road network impedance matrix; The strategy solving and correcting module is used for solving the double-layer material scheduling model by adopting a double-layer iteration solving strategy, and executing an online rolling correcting step in the solving iteration process, wherein the method comprises the steps of obtaining latest real-time weather monitoring data and road network real-time road condition data, updating the material demand priority parameter and the road network impedance matrix, and executing subsequent iteration based on the updated material demand priority parameter and the road network impedance matrix until an optimal material scheduling scheme is output.
9. A terminal comprises a processor and a storage medium, and is characterized in that: The storage medium is used for storing instructions; The processor being operative according to the instructions to perform the steps of the method according to any one of claims 1-7.
10. Computer readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the steps of the method according to any of claims 1-7.

Description

Power grid emergency material dispatching double-layer optimization method and system under extreme weather Technical Field The invention relates to the field of logistics management, in particular to a double-layer optimization method and a double-layer optimization system for power grid emergency material scheduling under extreme weather. Background In recent years, extreme weather events such as typhoons, storm and snow storm frequently occur worldwide, and serious threat is caused to the safe and stable operation of a power grid system. These extreme weather often cause damage to critical facilities such as transmission lines and transformers, and cause large-area power failure. The power system is taken as an important component of urban infrastructure, and the post-disaster quick recovery capability of the power system becomes an important index for measuring the toughness of the power system. In the emergency management of power grid disasters, efficient scheduling of materials and emergency repair forces is a key link for recovering power supply. In order to improve the recovery efficiency of the power grid after disaster, scholars at home and abroad conduct a great deal of research on the problem of emergency repair and scheduling of the power grid. According to the emergency material dynamic dispatching-rescheduling algorithm based on classification-dispatching, paths are optimized through k-means classification and genetic algorithm, and dynamic change of demand is adapted. Although the existing research has a certain achievement in the aspect of power grid emergency repair scheduling, the following defects still exist: First, existing models typically statically deal with material demands and path planning, ignoring the real-time evolution characteristics of extreme weather. On one hand, the nonlinear mapping of the meteorological strength and the equipment vulnerability cannot be established, the time sensitivity coefficient of the materials cannot be dynamically adjusted according to the meteorological change, and on the other hand, the path planning lacks a meteorological congestion and risk punishment mechanism. This lack of environmental awareness results in a scheduling scheme that fails to identify high-risk equipment needs and is prone to planning high-risk transportation paths, resulting in delayed or blocked delivery of critical materials. Second, existing scheduling schemes are mostly based on rigid assumptions of supply-demand balance, and only focus on single-sided scheduling of materials. When an extreme disaster causes absolute shortage of resources, the lack of a flexible degradation supply strategy based on load priority easily causes that a critical load cannot be guaranteed by minimum survival, and meanwhile, the strong correlation between the arrival time of materials and the arrival time of a rush repair team is not considered, so that time misplacement of people such as people or objects occurs on site, and the overall progress of rush repair is influenced. Thirdly, in the prior art, single-layer optimization or general heuristic algorithm solution is mostly adopted, a solution space pruning and masking strategy aiming at double-layer coupling characteristics is lacked, convergence speed is low when large-scale power grid nodes are processed, local optimization is easy to fall into, and timeliness requirements of a post-disaster golden rescue period are difficult to meet. In summary, the existing emergency material scheduling scheme has obvious defects in the aspects of time sensitivity, facility priority distinction, dynamic decision robustness and the like, and cannot fully meet the actual requirements of power grid rush-repair tasks under extreme weather disasters. Disclosure of Invention The invention aims to overcome the defects in the existing power grid emergency material scheduling scheme, and provides a power grid emergency material scheduling double-layer optimization method and system under extreme weather aiming at the problems that the traditional model cannot embody a time priority principle, does not distinguish the importance degree of electric power facilities, lacks dynamic environment adaptability, and has insufficient decision robustness. The method comprises the steps of firstly obtaining real-time weather monitoring data and vulnerability characteristics of power grid equipment, calculating a dynamic time sensitivity coefficient, generating material demand priority parameters by combining disaster point load grades, secondly constructing a double-layer material scheduling model, enabling an upper layer to take key load recovery index maximization as a target, introducing constraint to achieve degradation supply under resource limitation, enabling a lower layer to take comprehensive scheduling cost minimization as a target, achieving dynamic risk avoidance and space-time synchronization of people and goods through constraint, and finally adopting a doub