CN-121981466-A - Reservoir scheduling scheme making method based on multi-objective decision preference

CN121981466ACN 121981466 ACN121981466 ACN 121981466ACN-121981466-A

Abstract

The invention discloses a reservoir dispatching scheme making method based on multi-objective decision preference, and belongs to the field of water resource management and reservoir dispatching decision. According to historical data, collecting typical hydrologic annual warehouse-in flow data, reservoir characteristic parameters and reservoir characteristic curve data. And then constructing constraint conditions and setting an objective function calculation module according to actual demand conditions and capacity limitations of the unit, and setting up the calculation module based on a water balance principle and a unit water-electricity conversion relation in a period of time so as to calculate the dam front water level and the unit output. Finally, a multi-objective optimal scheduling and preference recommendation model is obtained, and the pareto optimal solution set and different preference recommendation schemes are obtained through solving by a non-dominant genetic algorithm. The invention integrates the competitive coordination relationship among all water conservancy targets, and converts the fuzzy preference into an accurate and executable scheduling scheme.

Inventors

ZHOU TAO
SHAO LINLIN
XU JIJUN
RAN QIHUA
WANG YONGQIANG
WANG DONG
REN YUFENG

Assignees

长江水利委员会长江科学院
河海大学
中国长江电力股份有限公司

Dates

Publication Date: 20260505
Application Date: 20260122

Claims (10)

1. The reservoir scheduling scheme making method based on the multi-objective decision preference is characterized by comprising the following steps of: based on historical data, collecting characteristic parameters and characteristic curves of the reservoir, selecting typical hydrologic years and extracting corresponding runoff data; Calculating a scheduling period based on a water balance principle Water level in front of dam Calculating the time period according to the water-electricity conversion relation of the unit Is of (2) Setting fixed constraint conditions in combination with actual demands and dynamic constraint conditions based on actual capacity of the reservoir to obtain a reservoir multi-objective optimal scheduling model; solving the multi-objective optimal scheduling model by adopting a non-dominant genetic algorithm to obtain a pareto optimal solution set; comparing different schemes in the pareto optimal solution set, and analyzing the influence rule of the preference change of the scheduling target on the scheduling scheme; and fitting a utility function based on the marginal substitution rate, and recommending an optimal scheduling scheme in the pareto optimal solution set according to the preference degrees of different scheduling targets.
2. The reservoir scheduling scheme formulation method based on multi-objective decision preference according to claim 1, wherein the typical hydrologic year comprises a cover of a plentiful year, a plain year and a withered year, and the selection process of the typical hydrologic year is as follows: collecting actual measurement data of the annual runoff of the reservoir in the historical annual intervals, and sequencing the actual measurement data according to the annual runoff from large to small to obtain an ordered annual runoff sequence; based on the ordered annual runoff sequence, calculating the cumulative probability corresponding to each annual runoff , Representing year, respectively selecting accumulated frequencies according to frequency co-occurrence principle Cumulative frequency And cumulative frequency The corresponding annual runoff quantity forms a characteristic runoff threshold set; and the years of which the runoff distribution accords with the long-term hydrologic law of the river basin in the flood season and the non-flood season are sequentially used as the typical high-water years, the typical flat water years and the typical dead water years.
3. A reservoir scheduling scheme formulation method based on multi-objective decision preference as claimed in claim 1 wherein said scheduling period The calculation process of the internal dam front water level comprises the following steps: The normal water level of the reservoir is used as the water level Based on the water balance principle, the time period is calculated by the following formula Storage volume in : ; Wherein, the For a period of time Is used for the storage capacity of the container, For a period of time Is used for controlling the flow rate of the warehouse entry, For the period of time to be long, For a period of time The ex-warehouse flow of the internal power station; Based on time period Storage volume in Obtaining the time period by using the curve function of the reservoir water level and the reservoir capacity Last dam front water level 。
4. The reservoir scheduling scheme formulation method based on multi-objective decision preference according to claim 1, wherein the fixed constraint conditions comprise a dam front water level constraint condition, a unit output constraint condition and a delivery flow constraint condition; the dynamic constraint conditions comprise a power generation flow constraint condition, a storage capacity guarantee constraint condition and a time period maximum ex-warehouse flow constraint condition.
5. The reservoir scheduling scheme formulation method based on multi-objective decision preference according to claim 1, wherein the scheduling objectives comprise maximum annual total power generation, maximum annual water shortage rate and maximum end-of-life water storage; correspondingly, the objective function comprises an annual energy generation total amount objective function Objective function of annual water shortage rate And end of period water storage objective function 。
6. The reservoir dispatching scheme making method based on multi-objective decision preference as claimed in claim 1, wherein the dispatching process line graph comprises a delivery flow process line Dam front water level process line Unit output process line The generation process is as follows: Taking the runoff data of the typical hydrology as a core initial population, randomly generating other populations, and setting iteration parameters; In the genetic iteration process, checking whether each candidate solution meets a fixed constraint condition and a dynamic constraint condition in real time, counting the violation degree of the solution which does not meet the constraint, and adopting a preset penalty coefficient to perform penalty processing on the solution so as to eliminate the candidate solution which seriously violates the constraint; genetic operation is carried out according to the set crossover probability and mutation probability, and continuous iteration is carried out by combining the reservoir characteristic curve until the preset genetic algebra is reached, and then evolution is terminated; screening and obtaining a pareto optimal solution set from the final evolutionary population, and fitting a multi-target pareto optimal curved surface according to the pareto optimal solution set And pareto optimal curve P between two targets 、、 , wherein, Is the total annual power generation amount, Is a annual water shortage rate, Water storage capacity is reserved for the end of the period; drawing ex-warehouse flow process lines corresponding to a plurality of pareto optimal solutions before drawing Dam front water level process line Unit output process line 。
7. The reservoir dispatching scheme making method based on multi-objective decision preference according to claim 1, wherein the influence rule is obtained based on objective function values corresponding to the pareto optimal solution set and dispatching process line analysis, and comprises a power generation objective preference degree, a water supply objective preference degree and a water storage objective preference degree.
8. The reservoir scheduling scheme formulation method based on multi-objective decision preference as claimed in claim 1, wherein the recommendation process of the optimal scheduling scheme comprises the following steps: Setting the total power generation amount according to the preference degree of a decision maker for each target Rate of annual water shortage And end of period water storage Expected marginal substitution rates between each other; Construction of utility functions The expression form is as follows: , wherein, 、 And Respectively the total amount of power generation Rate of annual water shortage And end of period water storage Weight and of (2) ; 、 And Respectively the total amount of power generation Rate of annual water shortage And end of period water storage A derivative of the sub-utility function representing a marginal utility of the target value; establishing the derivative and the expected marginal substitution rate of the weight and the sub utility function Corresponding relation of (2) and simultaneous equation system solution 、 And Determining a final utility function; Normalizing the objective function values of all schemes in the pareto optimal solution set, substituting the normalized objective function values into utility functions, and calculating the utility value U of each pareto optimal solution; And selecting a scheme with the maximum utility value U as an optimal scheduling scheme under the corresponding preference.
9. A reservoir scheduling scheme formulation method based on multi-objective decision preference as claimed in claim 8 wherein the weight, derivative of sub utility function and expected marginal substitution rate The correspondence of (2) is expressed in the following form: , ; In the formula, And Are all preset expected marginal substitution rates, 、 And The derivatives of the sub utility functions are respectively; corresponding, weight 、 And The solving process of (2) is as follows: In the total amount of power generation Rate of annual water shortage And end of period water storage On the premise of the importance balance of the sub utility functions, the derivatives of the sub utility functions on the respective targets are assumed to be the same and constant, namely , Is a constant; expressing the form of the corresponding relation, And Combining and solving to obtain 、 And 。
10. The reservoir scheduling scheme formulation method based on multi-objective decision preference as recited in claim 8, further comprising: based on the optimal scheduling scheme, a corresponding ex-warehouse flow process line is drawn Dam front water level process line Unit output process line 。

Description

Reservoir scheduling scheme making method based on multi-objective decision preference Technical Field The invention belongs to the field of water resource management and reservoir scheduling decision-making, and particularly relates to a reservoir scheduling scheme making method based on multi-objective decision-making preference. Background Reservoir scheduling is a key link of water resource system management, and particularly for a large comprehensive reservoir bearing multiple tasks, scheduling decisions of the reservoir scheduling are directly related to flood control safety, water resource utilization efficiency, energy supply and downstream water guarantee of a river basin. The common scheduling methods are mainly rule-based scheduling and optimization-based scheduling. The rule-based scheduling method relies on rules formulated according to historical experience and simplified scenes, is simple to operate and easy to implement, but lacks flexibility, is difficult to adapt to complex and changeable hydrologic conditions and diversified scheduling target requirements, and cannot guarantee that overall optimization is achieved in the event of multi-target conflict. The scheduling method based on optimization is to construct a mathematical optimization model and solve the optimal solution of the objective function to formulate a scheduling scheme. Often one target is considered as the primary target and the other targets are considered in the form of constraints. The multi-objective collaborative optimization cannot be embodied, the preference integration mechanism is absent, and a decision maker still needs to subjectively select a final scheme after generating the pareto front. The selection process lacks systematic preference quantification and integration mechanism, few researches attempt to introduce weights, but the common direct weight distribution method is too simplified, and cannot capture the subtle characteristics of the marginal substitution rate changing along with the target value, so that the generated scheduling scheme has deviation from the true preference of a decision maker. Therefore, how to scientifically quantify multi-objective preference when facing multiple objectives, to obtain reasonable, non-subjective weight distribution, and further to select the scheme of the most fitting decision maker preference from a plurality of pareto optimal solutions is a technical problem to be solved at present. Disclosure of Invention The invention provides a reservoir dispatching scheme making method based on multi-objective decision preference for solving the technical problems in the background technology. The reservoir dispatching scheme making method based on the multi-objective decision preference comprises the following steps: based on historical data, collecting characteristic parameters and characteristic curves of the reservoir, selecting typical hydrologic years and extracting corresponding runoff data; Calculating a scheduling period based on a water balance principle Water level in front of damCalculating the time period according to the water-electricity conversion relation of the unitIs of (2)Setting fixed constraint conditions in combination with actual demands and dynamic constraint conditions based on actual capacity of the reservoir to obtain a reservoir multi-objective optimal scheduling model; solving the multi-objective optimal scheduling model by adopting a non-dominant genetic algorithm to obtain a pareto optimal solution set; comparing different schemes in the pareto optimal solution set, and analyzing the influence rule of the preference change of the scheduling target on the scheduling scheme; and fitting a utility function based on the marginal substitution rate, and recommending an optimal scheduling scheme in the pareto optimal solution set according to the preference degrees of different scheduling targets. In a further embodiment, the representative hydrologic year includes a cover of a full water year, a flat water year and a dead water year, and the selection process of the representative hydrologic year is as follows: collecting actual measurement data of the annual runoff of the reservoir in the historical annual intervals, and sequencing the actual measurement data according to the annual runoff from large to small to obtain an ordered annual runoff sequence; based on the ordered annual runoff sequence, calculating the cumulative probability corresponding to each annual runoff ,Representing year, respectively selecting accumulated frequencies according to frequency co-occurrence principleCumulative frequencyAnd cumulative frequencyThe corresponding annual runoff quantity forms a characteristic runoff threshold set; and the years of which the runoff distribution accords with the long-term hydrologic law of the river basin in the flood season and the non-flood season are sequentially used as the typical high-water years, the typical flat water years and the t