CN-121745632-B - Sustainable development regulation and control method for yellow river basin cascade reservoir group system

CN121745632BCN 121745632 BCN121745632 BCN 121745632BCN-121745632-B

Abstract

The invention discloses a sustainable development regulation and control method of a yellow river basin cascade reservoir group system, which comprises the steps of dividing a basin system into water sand, ecology and society subsystems, obtaining multidimensional parameter data and identifying sequential state variable evolved by a driving system, calculating the same vibration degree of the sequential state variable, representing the cooperative level of variable evolution, calculating the integral harmonic contract degree of the system based on the same vibration degree, selecting water sand regulation and control factors, constructing a quantitative regulation and control model containing the mapping relation of the same vibration degree of the water sand regulation and control factors and the sequential state variable and the system harmonic contract degree, setting a target harmonic contract degree interval of the system, and reversely calculating a water sand regulation and control factor regulation threshold value required for realizing the target based on the quantitative regulation and control model. By introducing evolution collaborative analysis and a reverse regulation mechanism, the method solves the problems that the traditional method is difficult to quantify time sequence synchronism and can not reversely push a regulation boundary, and realizes targeted accurate regulation of a river basin system.

Inventors

JIAO JIAN
HU YING
QU BO
JIANG ENHUI
Tang Bixuan
YANG XIAOYU
WANG YIFEI
CHEN BEN
DING LEI
HUANG YUMING
DOU XIPING
MA AIXING

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: 20260508
Application Date: 20260211

Claims (7)

1. A sustainable development regulation and control method for a yellow river basin cascade reservoir group system is characterized by comprising the following steps: Dividing a yellow river basin cascade reservoir group system into water and sand, ecological and social subsystems, and acquiring multidimensional parameter data representing the running state of each subsystem; identifying an order state flow variable which drives the subsystem to evolve from the multidimensional parameter data; calculating the degree of co-vibration of each sequence flow variable, and representing the cooperative level of variable evolution; calculating the system harmonic moment of the yellow river basin cascade reservoir group system based on the same vibration moment; Selecting water sand regulating factors, and constructing a quantitative regulation model containing the water sand regulating factors, the same vibration degree of sequence flow variables and a mapping relation of system harmonic deepness; Setting a target harmonic contractual degree interval of the yellow river basin cascade reservoir group system, and reversely solving a water sand regulating factor regulating threshold value meeting the target harmonic contractual degree interval based on a quantitative regulating model; calculating the degree of co-vibration of each sequence flow variable, comprising: normalizing historical sequence data of the sequence state flow variable to obtain a normalized level value, and representing the static state development level of the sequence state flow variable; Calculating evolution co-coefficients between the sequence flow variable and other parameters in the subsystem, and representing the time sequence synchronization degree of the sequence flow variable and the overall evolution of the subsystem; Calculating the co-vibration degree, namely U (P xy ) = U*(P xy )×η xy ; wherein U (P xy ) is the co-vibration degree, U (P xy ) is a normalized level value, and eta xy is an evolution co-coefficient; The method for constructing the quantitative regulation model comprising the mapping relation of the co-vibration degree of the water sand regulation factors and the sequence flow variables and the system harmonic contract degree comprises the following steps: Selecting corresponding water and sand regulating factors for each sequential state flow variable, and establishing a first quantitative relation between the water and sand regulating factors and the equivalent vibration degree of the sequential state flow variable by using regression analysis; based on the weight of each sequential state flow variable, establishing a second quantitative relation between the same vibration degree of the sequential state flow variable and the system harmonic contract degree; Cascading the first quantitative relation and the second quantitative relation to obtain a quantitative regulation model taking the water sand regulation factor as input and taking the system harmonic degree as output; before establishing the first quantitative relation between the water sand regulating factor and the equivalent vibration degree of the sequential state flow variable, the method further comprises the following steps: performing accumulated effect treatment on the water sand regulating factors, and constructing accumulated effect regulating factors reflecting the hysteresis superposition influence of regulating measures; identifying optimal accumulation parameters by taking maximization of correlation coefficients of the water and sand regulation factors and the same vibration degree of the sequence flow variable as targets; and taking the cumulative effect regulating factor generated based on the optimal cumulative parameter as an independent variable, taking the degree of co-vibration of the sequence flow variable as the dependent variable, and performing fitting calculation to construct a first quantitative relation.
2. The method of claim 1, wherein identifying from the multi-dimensional parametric data an order state variable that drives the subsystem evolution, comprises: classifying the multidimensional parameter data, and carrying out arithmetic average processing on the parameter data of the same class to obtain a simplified parameter set; Based on the simplified parameter set, performing preliminary identification by using a pre-configured sequence state flow variable identification model to obtain a macroscopic sequence state flow variable; extracting an original parameter corresponding to the macroscopic state flow variable from the multidimensional parameter data, and performing secondary identification by using the state flow variable identification model to obtain a final state flow variable.
3. The method of claim 2, wherein the state flow variable identification model identifies state flow variables by calculating similarity coefficients of the value parameter structure and the ideal value parameter structure; The similarity coefficient is calculated by a modified cosine similarity formula: r ki = (∑ j=1 p (w ij ×w* ij )) / (sqrt(∑ j=1 p (w ij ) 2 )×sqrt(∑ j=1 p (w* ij ) 2 )); Wherein r ki is a similarity coefficient in the kth dominant mode, w ij is the jth value parameter component of the ith parameter in the kth dominant mode, w ij is a corresponding component in the ideal value parameter structure, and p is the dimension of the value parameter; And selecting a dominant mode with the largest similarity coefficient as an evolution direction of the system, and determining a first parameter of the dominant mode as an order state flow variable.
4. The method of claim 1, wherein calculating the co-coefficients of evolution between the sequence flow variable and other parameters within the subsystem comprises: Constructing a reference parameter set, wherein the reference parameter set is composed of the rest parameters except for the sequence state variable in the subsystem; Calculating evolution direction indication functions of all parameters in the sequence state flow variable and reference parameter set, and marking the change states of the parameters in adjacent years, wherein the change states comprise ascending, leveling or descending; Based on the evolution direction indication function, calculating the direction consistency of any parameter in the sequence flow variable and the reference parameter set at the same moment, and calculating the average direction consistency of the sequence flow variable in the statistical period; Based on the average directional consistency of all parameters in the reference parameter set, the evolution synergy coefficient eta xy is calculated by using the following formula: ; Wherein |Ω x | is the number of parameters in the reference parameter set Ω x , and C xy,i is the average direction consistency of the sequence flow variable and the i-th parameter in the reference parameter set.
5. The method of claim 1, wherein constructing a quantitative regulation model comprising a mapping relationship of the co-vibration degree and the system harmonic degree of the water-sand regulation factor and the sequence flow variable, further comprises: Taking a yellow river basin cascade reservoir group system as a dynamic control system, and constructing a discrete time state space model which takes a sequence state flow variable as a state variable and a candidate water sand regulating factor as a control variable; calculating a corresponding controllability Gramian matrix based on a discrete time state space model aiming at different combinations of candidate water and sand regulating factors; calculating a harmonic contract controllability index according to a controllability Gramian matrix, and representing the capability of the combination of the regulating factors to drive the sequential state variable to reach a preset target state; And selecting an optimal regulation factor combination when the harmonic wedge controllability index meets a preset threshold value, and constructing a quantitative regulation model based on the optimal regulation factor combination.
6. The method of claim 5, wherein the discrete-time state space model is represented as: x(t+1) = A×x(t) + B×u(t) + w(t); Wherein x (t) is an order state variable vector at the moment t, u (t) is a water sand regulating factor vector at the moment t, w (t) is a random disturbance vector, A is a state matrix reflecting the natural evolution of the system, and B is an input matrix reflecting the influence of the regulating factor; the discrete time state space model is built, and specifically comprises a state matrix A and an input matrix B which are obtained by utilizing a least square method based on prestored historical data in a fitting mode.
7. The method of claim 5, wherein calculating the corresponding controllability Gramian matrix and harmonic controllability indexes comprises: for any subset S of candidate water and sand control factors, a controllability Gramian matrix Wc (S) in the finite domain is calculated using the following formula: Wc(S)=∑ k=0 N-1 (A k ×Bs×Bs T ×(A T ) k ); Wherein Bs is a submatrix composed of column vectors corresponding to the subset S in the input matrix B, T is a transpose, N is a time window length, k is a discrete time step index in a finite field; The harmonic contract controllability index C ctrl (S) is calculated using the following formula: C ctrl (S) = log(det(Wc(S) +εI)); where det represents the matrix determinant and εI is the regularization term to prevent singularities; if the value of C ctrl (S) is larger, the better the harmonic contract controllability of the water and sand regulating factor combination corresponding to the subset S is judged.

Description

Sustainable development regulation and control method for yellow river basin cascade reservoir group system Technical Field The invention belongs to the field of hydraulic engineering scheduling, and particularly relates to a sustainable development regulation and control method for a yellow river basin cascade reservoir group system. Background The cascade reservoir group system forms a complex coupling body of water sand power, ecological maintenance and water resource allocation. In the regulation and control of the large-scale system, the core technical difficulty is to quantify the coupling relation between multidimensional parameters and establish a closed-loop control mechanism from engineering dispatching to system state. The method is used for accurately identifying the sequential state variable of the evolution of the driving system, quantitatively evaluating the cooperative level of the internal parameters of the system, and is a technical premise for determining the operation strategy of the reservoir group and guaranteeing the evolution stability of the river basin system. At present, regulation and control research on a cascade reservoir group usually adopts an sequential state flow variable identification technology to characterize the evolution characteristics of the system. The existing mainstream technical path is that historical multidimensional parameter data of a water-sand and ecological equal-dividing system are collected, the data are normalized by means of a method such as extremely poor standardization and the like, static co-vibration degree reflecting the parameter development level is calculated, water-sand regulating factors (such as reservoir outlet flow, sand content and the like) which are relatively high in sequence state flow variable correlation are screened out through a statistical method such as Pearson (Pearson) correlation analysis and the like, and a forward evaluation model of the system is built to evaluate the system harmony state under a specific scheduling scheme. However, the prior art has the deep technical problems of disjointing static numerical evaluation and dynamic evolution characteristics and mismatch of statistical correlation and physical controllability. Specifically, the co-vibration degree calculation based on static normalization only reflects the level of parameters, ignores the synchronicity of evolution trend among parameters, generates false co-vibration erroneous judgment when the parameter level is higher but the evolution direction is opposite (such as one rise and one fall), and cannot truly represent the co-evolution of the system. In addition, screening for regulatory factors based on statistical correlation lacks verification of controllability, and high correlation does not represent that the regulatory factors have enough power to drive the system state into the target interval in a limited time, resulting in difficulty in establishing an accurate inverse solving path from the target tuning degree to a specific engineering regulatory threshold. Disclosure of Invention The invention aims to provide a sustainable development regulation and control method for a yellow river basin cascade reservoir group system, so as to solve the problems in the prior art. The technical scheme is that the sustainable development regulation and control method of the yellow river basin cascade reservoir group system comprises the following steps: Dividing a yellow river basin cascade reservoir group system into water and sand, ecological and social subsystems, and acquiring multidimensional parameter data representing the running state of each subsystem; identifying an order state flow variable which drives the subsystem to evolve from the multidimensional parameter data; calculating the degree of co-vibration of each sequence flow variable, and representing the cooperative level of variable evolution; calculating the system harmonic moment of the yellow river basin cascade reservoir group system based on the same vibration moment; Selecting water sand regulating factors, and constructing a quantitative regulation model containing the water sand regulating factors, the same vibration degree of sequence flow variables and a mapping relation of system harmonic deepness; setting a target harmonic contractual degree interval of the yellow river basin cascade reservoir group system, and reversely solving a water and sand regulating factor regulating threshold value meeting the target harmonic contractual degree interval based on a quantitative regulating model. The method has the beneficial effects that by introducing evolution collaborative analysis and a reverse regulation mechanism, the problems that the time sequence synchronism is difficult to quantify and the regulation boundary cannot be reversely pushed in the traditional method are solved, and the targeted accurate regulation of the river basin system is realized. Drawings Fig. 1 is a flow chart of st