CN-122022502-A - Water diversion project risk cascade propagation and risk assessment method and system based on multi-level topological structure

CN122022502ACN 122022502 ACN122022502 ACN 122022502ACN-122022502-A

Abstract

The invention discloses a water diversion project risk cascade propagation and risk assessment method and system based on a multi-level topological structure, wherein the method comprises the steps of obtaining a water diversion project topological network, and generating node comprehensive risk vectors of all nodes based on multi-source operation data; the method comprises the steps of constructing a basic risk propagation matrix according to a physical connection relation, carrying out structural modulation on the basic risk propagation matrix based on node comprehensive risk vectors to generate comprehensive propagation operators, carrying out cascade propagation iterative evolution on the node comprehensive risk vectors based on the comprehensive propagation operators, continuously calculating network dynamics stability characteristic parameters of the comprehensive propagation operators in an evolution process, reconstructing the comprehensive propagation operators when a system is judged to be in a divergent phase state according to the parameters, continuing iterative evolution until convergence, obtaining steady-state risk distribution, and outputting a system comprehensive risk assessment result based on the steady-state risk distribution. The invention can improve the safe control capability of the operation of the water diversion project under the complex operation condition and the external disturbance condition.

Inventors

LI XING
ZHAO YIFU
HU JIANG
GAO CHANGSHENG
ZHU QINGSHUAI
LIU JINGYANG
ZHANG YUHAN
HE LINHUA
HE RENHUI
WANG WENLEI

Assignees

水利部交通运输部国家能源局南京水利科学研究院

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (10)

1. The water diversion project risk cascade propagation and risk assessment method based on the multi-level topological structure is characterized by comprising the following steps of: Acquiring a water diversion project topology network based on trunk line, water diversion ports and branch line structure division, and generating node comprehensive risk vectors of all nodes based on multi-source project operation data; Constructing a basic risk propagation matrix according to a physical connection relation in the water transfer engineering topology network, and carrying out structural modulation on the basic risk propagation matrix based on the node comprehensive risk vector to generate a comprehensive propagation operator; Performing cascade propagation iterative evolution on the node comprehensive risk vector based on the comprehensive propagation operator, and continuously calculating network dynamics stability characteristic parameters of the comprehensive propagation operator in the evolution process to judge a system phase state; When the system is in a divergent phase state according to the network dynamics stability characteristic parameter judgment system, performing self-adaptive triggering reconstruction operation on the comprehensive propagation operator to reduce the propagation intensity of the comprehensive propagation operator, generating a reconstructed comprehensive propagation operator, and continuously performing cascade propagation iterative evolution until convergence based on the reconstructed comprehensive propagation operator to obtain steady state risk distribution and network propagation topological characteristics in the evolution process; Based on steady-state risk distribution, combining with network propagation topological characteristics, and outputting a comprehensive risk assessment result of the system.
2. The method of claim 1, wherein structurally modulating the base risk propagation matrix based on the node composite risk vector generates a composite propagation operator, comprising: performing multi-scale coupling modulation on the basic risk propagation matrix by utilizing a pre-constructed node-level bidirectional mapping relation to obtain a coupling propagation matrix; and performing nonlinear enhancement modulation on the coupled propagation matrix based on the node comprehensive risk vector to generate a comprehensive propagation operator.
3. The method of claim 2, wherein performing multi-scale coupling modulation on the base risk propagation matrix using a pre-constructed node-level bi-directional mapping relationship to obtain a coupling propagation matrix, comprising: Constructing a multi-scale modulation matrix based on a pre-constructed aggregation operator used for representing the mapping from the node to the level and a feedback operator used for representing the mapping from the level to the node and combining with a pre-configured scale coupling strength parameter; Multiplying the multi-scale modulation matrix by the basic risk propagation matrix to obtain a coupling propagation matrix.
4. The method of claim 2, wherein performing nonlinear-enhanced modulation on the coupled propagation matrix based on the node composite risk vector generates a composite propagation operator, comprising: Weighting calculation is carried out on the basis of external disturbance factors of all nodes and the node comprehensive risk vector which are obtained in real time, so that a system-level external disturbance index is obtained; Performing downward-pressing modulation on a preset reference threshold by using a system-level external disturbance index, and determining a dynamic trigger threshold; Extracting overrun nodes which are larger than or equal to a dynamic trigger threshold value in the node comprehensive risk vector, and determining a grading enhancement coefficient according to the risk amplitude of the overrun nodes; And amplifying the out-of-edge weight of the corresponding overrun node in the coupling propagation matrix by using the grading enhancement coefficient to generate a comprehensive propagation operator.
5. The method of claim 1, wherein performing cascading propagation iterative evolution on the node composite risk vector based on the composite propagation operator comprises: Acquiring a preset space physical attenuation factor, scaling the comprehensive propagation operator based on the space physical attenuation factor, and constructing an actual evolution propagation operator; And taking the node comprehensive risk vector as an initial iteration state, and executing the fixed point iteration with limit truncation by using an actual evolution propagation operator.
6. The method of claim 1, wherein the network dynamic stability characteristic is a spectral radius; continuously calculating network dynamics stability characteristic parameters of the comprehensive propagation operator, comprising: and extracting the maximum eigenvalue modulus of the comprehensive propagation operator to obtain the spectrum radius.
7. The method of claim 6, wherein determining that the system is in a divergent phase based on the network dynamic stability characteristics comprises: Comparing the spectrum radius with a preset stability critical value and a preset tolerance parameter; When the spectral radius is greater than the sum of the stability threshold and the tolerance parameter, the system is determined to be in a divergent phase.
8. The method of claim 1, wherein performing an adaptive trigger reconstruction operation on the composite propagation operator to reduce the propagation strength of the composite propagation operator, generating a reconstructed composite propagation operator comprises: Performing hierarchical propagation inhibition processing on the comprehensive propagation operator, and reducing propagation weights corresponding to the high risk hierarchy; Based on the node comprehensive risk vector, performing node risk orientation strengthening treatment on the propagation operator subjected to the hierarchical propagation inhibition treatment, and enhancing the edge entering weight corresponding to the high-risk node; And performing edge structure pruning processing on the propagation operators subjected to node risk orientation strengthening processing by using a preset pruning threshold value, and setting the propagation weight smaller than the pruning threshold value to zero to generate a reconstructed comprehensive propagation operator.
9. The method of claim 1, wherein the adaptively triggered reconfiguration operation is triggered when at least one of the following conditions is met in addition to being triggered when the system is determined to be in a divergent phase state: extracting the maximum node risk amplitude value in the node comprehensive risk vector, and triggering when the maximum node risk amplitude value is greater than a preset node risk warning threshold value; Based on the system-level external disturbance index acquired in real time, triggering is performed when the system-level external disturbance index is larger than a preset external disturbance critical threshold value.
10. The utility model provides a transfer engineering risk cascade propagation and risk assessment system based on multistage topological structure which characterized in that includes: A memory for storing a computer program; processor for implementing the steps of the method according to any of claims 1 to 9 when executing a computer program.

Description

Water diversion project risk cascade propagation and risk assessment method and system based on multi-level topological structure Technical Field The invention relates to the field of hydraulic engineering operation safety guarantee and technical evaluation, in particular to a water diversion engineering risk cascade propagation and risk evaluation method and system based on a multi-level topological structure. Background The water diversion project has multi-level complex topological relations of a trunk line, a water diversion port, a branch line and the like, and has deep nonlinear space-time coupling among structural states, running deviations and external environment disturbance. The method accurately characterizes the cross-level conduction mechanism of the multisource risk factors among heterogeneous engineering units, establishes a dynamic evolution dynamics model with algebraic convergence, and has important technical value for realizing large-scale topological stability prejudgment of a long-distance water delivery system and self-adaptive feedback regulation and control of operation conditions. In the prior art, when the risk assessment of the water diversion network is carried out, a single-node static evaluation model or a linear cascade model based on fixed weight is mainly adopted. Such schemes typically pre-configure a constant risk propagation matrix, calculate a comprehensive risk index by bottom-up data aggregation, and trigger a static alarm or evaluate the security level of a local unit according to a preset constant threshold under a unidirectional information flow framework. The static evaluation method faces deep technical defects of non-closed loop of a propagation mechanism, undefined phase change boundary, lack of physical attenuation constraint and the like, so that the evaluation result is easy to generate dynamic distortion or numerical divergence under a complex working condition. Specifically, the existing constant threshold model is difficult to respond to the erosion of external strong disturbance to the bearing limit of the system, and the fixed propagation weight is adopted, so that the macroscopic situation of the level cannot be reversely mapped to a local node, and the nonlinear transition depicting capability of multi-scale bidirectional coupling is lacked; the method has the advantages that the explicit modeling of the engineering hierarchical structure is lacked, the bidirectional coupling relation between the nodes and the hierarchy is not established, and the influence of the overall situation on the local risk propagation is difficult to reflect; due to the fact that theoretical constraint of a space physical attenuation factor and a network spectrum radius characteristic value is absent, convergence is difficult to ensure by a traditional linear iterative evolution equation, so that a model is difficult to define divergent and convergent phase boundaries of risk propagation in algebraic theory, a cascade propagation model of risks in a topological structure is not built, a diffusion trend of risks evolving along with time is difficult to be described, in addition, a static model lacks an adaptive reconstruction mechanism for actively intervening and cutting off propagation weights according to a phase judgment result, once local extreme degradation occurs, an original model is easy to generate negative weight or physical spurious of reverse propagation, and a remodelling feedback process of engineering disaster prevention scheduling intervention on a network dynamics structure is difficult to be mapped truly. In other words, it is difficult to determine whether the risk is in a decaying state or a diffuse state without a system overall stability determination mechanism. In the emergency dispatch or structure state mutation scene, the dynamic evolution process of the risk diffusion from the local abnormality to the system level is difficult to accurately describe in the prior art. Disclosure of Invention The invention aims to provide a water diversion project risk cascade propagation and risk assessment method and system based on a multi-level topological structure, so as to solve at least one of the problems in the prior art. According to one aspect of the application, a water diversion project risk cascade propagation and risk assessment method based on a multi-level topological structure comprises the following steps: Acquiring a water diversion project topology network based on trunk line, water diversion ports and branch line structure division, and generating node comprehensive risk vectors of all nodes based on multi-source project operation data; Constructing a basic risk propagation matrix according to a physical connection relation in the water transfer engineering topology network, and carrying out structural modulation on the basic risk propagation matrix based on the node comprehensive risk vector to generate a comprehensive propagation operator; Per