CN114254571A - Machine set control rule optimization decision method under extreme working conditions of pumped storage power station - Google Patents

Abstract

The invention discloses a unit control rule optimization decision method under extreme working conditions of a pumped storage power station, and belongs to the technical field of fluid machinery and energy power. The method comprises the following steps: establishing a mathematical model of the extreme working condition transition process of the pumped storage power station, inputting a unit control rule, calculating the unit rotating speed, the volute water attack pressure extreme value and the draft tube water attack pressure extreme value, constructing a multi-target function of the dynamic quality of the unit transition process, and screening a unit control rule Pareto solution set by combining an improved non-dominated genetic algorithm. Further, a multi-criterion decision hierarchy framework of the unit control rule under the extreme working condition of the pumped storage power station is established, qualitative and quantitative indexes are integrated, and an optimal solution is selected in a unit control rule Pareto solution set based on an intuitive fuzzy analytic hierarchy process to serve as the optimal unit control rule under the extreme working condition of the pumped storage power station. The selected pump storage power station extreme working condition unit control rule can effectively improve the dynamic quality of the transition process of the pump storage power station under the extreme working condition, and improve the operation stability and safety of the power station.

Description

An optimization decision-making method for unit control law under extreme conditions of pumped-storage power station

technical field

The invention belongs to the technical field of fluid machinery and energy power, and relates to the technical field of pumped-storage power station control, and more particularly, to a decision-making method for optimizing unit control laws under extreme working conditions of a pumped-storage power station.

Background technique

In recent years, a large number of renewable intermittent energy sources such as wind and solar have been continuously connected to the power grid, and their volatility has seriously threatened the stable operation of the power grid. As an important energy storage technology, pumped storage power station is an effective and indispensable adjustment method for the power system. Its unique operating characteristics of peak regulation and valley filling play an important role in adjusting the load, promoting energy conservation of the power system and maintaining the security and stability of the power grid. important functions to run. However, when the power grid or unit fails, the water flow in the pipeline needs to be cut off in time. Under this extreme working condition, the hydraulic transient will cause great risks to the reliability of the unit and the pressure diversion system, which is not conducive to the safe, efficient and stable operation of the unit. In addition, pumped-storage power plants are developing towards high water head, large unit capacity, complex water-passing systems, and super-long water diversion pipelines. Frequent switching of working conditions also makes it easier for the units to fall into extreme working conditions, which is very important for the safety of pumped-storage power plants. Stable operation poses a great threat. In order to achieve the purpose of safe and stable operation of the unit, it is urgent to improve the control level of the pumped storage power station under extreme conditions.

Studies have shown that the shut-off law of pumped storage units has a great influence on the transition process. A reasonable shut-down law can reduce the water hammer pressure and the peak speed of the rotational speed within a certain range, and obtain a good dynamic quality of the transition process. In recent years, some people have applied optimization algorithms to the optimization of control laws under extreme conditions of pumped-storage power plants, but the results obtained by the optimization algorithms are a set of Pareto solutions, and the pros and cons of the solutions in the solution set cannot be compared. Moreover, the selection of the control law of the pumped-storage power station is essentially a multi-objective and multi-criteria decision-making problem. It is necessary to comprehensively consider various dynamic qualities of the transition process and other qualitative and quantitative factors in addition to the extreme value of the unit’s speed and pressure. Determine the unit shutdown pattern.

SUMMARY OF THE INVENTION

In view of the defects and improvement needs of the existing technology, the present invention proposes a decision-making method for optimizing the control law of the unit under extreme working conditions of the pumped-storage power station. Qualitative and quantitative factors of dynamic quality, multi-criteria decision-making to select the optimal control law of pumped-storage power station, to ensure the safe and stable operation of pumped-storage system.

In order to achieve the above purpose, the present invention provides a decision-making method for optimizing the control law of units under extreme working conditions of a pumped-storage power station, including:

(1) Establish a mathematical model for the transition process of the pumped storage power station under extreme conditions, input the unit control law, calculate the unit speed, the extreme value of the water hammer pressure of the volute, and the extreme value of the water hammer pressure of the draft tube, and use the multi-objective function of the dynamic quality of the unit transition process In order to optimize the objective, the Pareto solution set of the unit control law is screened by the improved non-dominated genetic algorithm, which integrates Latin hypercube sampling and piecewise linear chaotic mapping, and embeds an adaptive penalty strategy to solve the multiple decision space boundary constraints.

(2) Establish a multi-criteria decision-making hierarchy framework for the unit control law of the pumped-storage power station under extreme working conditions, synthesizing qualitative and quantitative indicators, and select the optimal solution from the Pareto solution set of the unit control law as the optimal control of the unit under extreme working conditions of the pumped-storage power station law.

Preferably, in step (1), the mathematical model of the transition process under extreme working conditions of the pumped-storage power station includes a pressurized water flow system, a water pump turbine, a governor and its servo system; the pressurized water flow system includes a pressure pipeline, Surge chamber and ball valve.

The pressure pipeline is described by a continuous equation and a momentum equation, and its flow and pressure transient characteristics are obtained by a characteristic line method. The continuity equation and momentum equation have the following form:

The surge chamber has the following form:

Among them, H _s and Q _s are the pressure and flow rate at the bottom of the surge chamber, respectively, At and A _c are the cross-sectional area of the impedance hole and the cross-sectional area of the surge chamber, respectively, and _{H c} and _HR are the water head and adjustment, respectively. The hydraulic loss head of the pressure chamber, K _R is the hydraulic loss coefficient of the resistance hole, which is related to the flow direction.

The mathematical expression of the ball valve is:

Among them, A and Q are the cross-sectional area at the ball valve, the overcurrent flow at the ball valve, K is the overflow coefficient of the ball valve, and g is the acceleration of gravity.

For the pump-turbine, the pump-turbine full characteristic curve is used as the nonlinear interpolation calculation of the pump-turbine, the improved Suter variation method is introduced to process the unit's full characteristic curve, and the two-variable ternary Lagrangian interpolation method is used to improve the calculation accuracy. The Suter variation and Lagrangian interpolation method have the following form:

The transition process model of the pumped-storage power station is obtained by synthesizing the pressurized water system, the pump turbine, the governor and its servo system.

Preferably, the optimization objective described in step (1) is characterized in that it includes the relative value of the maximum increase in the rotational speed of the unit and the composite extreme value of the water hammer pressure;

The calculation form of the relative value of the maximum increase in the rotational speed of the unit is:

Among them, n is the number of units, x _i is the rotational speed of the i-th unit, and x _r,i is the rated speed of the i-th unit;

The calculation form of the composite extreme value of the water hammer pressure is:

为尾水管水击压力额定值，I_v、I_d分别为蜗壳、尾水管重要度权重系数。Among them, P _vol,i and P _dra,i are the water hammer pressure values of the volute and draft tube, respectively,

is the rated value of the water hammer pressure of the draft tube, and I _v and I _d are the weight coefficients of the importance of the volute and the draft tube, respectively.

Preferably, the closing rule of the unit described in step (1) includes the one-stage straight-line closing rule of the guide vane; the two-stage folding-line closing rule of the guide vane; the two-stage closing rule that the guide-vane ball valve is associated with; the guide-vane delaying one-stage linear ball valve closing rule .

The law of closing a straight line of one section of the guide vane is characterized by: D ₁ =[t _i ]; the law of closing two broken lines of the guide vane is characterized by: D ₂ =[t _i-1 ,t _i-2 ,y _i ]; the guide vane ball valve is associated with two-stage closing law, characterized by: D ₃ =[t _i-1 , t _i-2 , y _i , t _bi-1 , t _bi-2 , θ _i ]; The closing rule of the guide vane delaying one segment of linear ball valve and two segments of broken line is: D ₃ =[t _is , t _id , t _bi-1 , t _bi-2 , θ _i ]. In the formula, i is the number of units in the pumped-storage power station, t _i-1 , t _i-2 , y _i are the closing time of the first stage, the closing time of the second stage and the inflection point in the two-section broken line closing rule of the unit, respectively, t _{bi- 1} t _bi-2 θ _i are the closing time of the first stage, the closing time of the second stage and the inflection point in the closing rule of the two-section broken line of the ball valve of the unit, respectively. t _is t _id is the closing time and delay time for the unit to delay a straight line closing, respectively.

Preferably, the improved non-dominated genetic algorithm described in step (1) includes the following steps:

(1.1) Set the size N of the particle population, the maximum iteration number T, the current iteration number t, the chaotic mutation condition T _chaotic , the decision space dimension D, the cross-recombination probability P _c , and the polynomial mutation probability P _m ;

(1.2) Execute the Latin hypercube sampling strategy to initialize the particle population, divide the D-dimensional decision space into N non-overlapping equally spaced regions, and randomly select a point from each equally spaced region in each dimension as the particle decision variable, Generate N initialized particles as the parent particle swarm P;

(1.3) According to the quantitative relationship between feasible solutions and infeasible solutions in the current iteration loop, implement an adaptive penalty strategy, and calculate the revised optimization target value; the adaptive penalty strategy has the following form:

Among them, f _m (X _i (t)) is the m-th optimization target value of the i-th particle, and p(X _i (t)) is the penalty function added to the particle; the penalty function has two penalty factors M (X _i (t)), N(X _i (t)), the calculation form is:

p(X _i (t))=(1-r _f )M(X _i (t))+r _f N(X _i (t))

为第i个粒子违反约束的总和，

分别为粒子种群中可行解的最大最小优化目标值，r_f为可行解在粒子种群的数量关系；in,

is the sum of constraint violations for the ith particle,

are the maximum and minimum optimization objective values of feasible solutions in the particle population, respectively, and r _f is the quantitative relationship of feasible solutions in the particle population;

(1.4) Execute non-dominated sorting, generate child particle swarm C from parent particle swarm P through strategies such as tournament selection, cross-recombination and polynomial mutation, and calculate the optimal target value of particles in child particle swarm C;

(1.5) Determine whether the current iteration number t is greater than the chaotic mutation condition T _chaotic , if so, execute step (1.7), otherwise go to step (1.7);

(1.6) Execute a piecewise linear chaotic map, the calculation form of the piecewise linear chaotic map is:

Among them, p∈(0, 0.5) is the control parameter, and x _i ∈(0,1) is the digital chaotic pseudo-random sequence;

(1.7) Integrate the parent particle swarm P and the child particle swarm C to form a family particle swarm Ω, and select N preferred particles from the family particle swarm Ω to form an elite file set E based on the non-dominated sorting and crowding degree; The preferred particles are selected according to the following principles: first, they are selected from the particles with the highest non-dominant ranking in the family particle group Ω. When the number of particles in this rank is greater than N, the crowding degree of the particles is used as the preferred basis to select Particles with a smaller crowding degree are put into the elite file set E. If the particle swarm with the highest non-dominant ranking is less than N, it is selected from the particles of the next non-dominant ranking until the elite file set E contains N preferred particles;

(1.8) Determine whether the current number of iterations t is less than the maximum number of iterations T, if so, jump out of the loop and take the current elite archive set R as the Pareto solution set; otherwise, the current iteration number t=t+1, take the current elite archive set R as the Pareto solution set The parent particle swarm P of the next cycle goes to step (1.3).

Preferably, step (2) comprises the following substeps:

(2.1) Take the intuitionistic fuzzy set (x|μ _x , ν _x , π _x ) as the judgment base, compare the decision-making hierarchy layer by layer from bottom to top, and form the intuitionistic fuzzy hierarchical judgment matrix R; where the intuitionistic fuzzy values μ _x , ν _x , π _x are membership degree, non-membership degree and hesitation degree, respectively;

(2.2) Consistency test is carried out on the judgment matrix R of the intuitionistic fuzzy hierarchy, and the judgment matrix satisfying the consistency test is considered to be consistent with the judgment consistency, and then go to step (2.4), otherwise, go to step (2.3);

多次重复上述步骤直至获得满足一致性的判断矩阵；(2.3) Modify the intuitionistic fuzzy hierarchical judgment matrix, calculate the complete consistency matrix R _p of the intuitionistic fuzzy hierarchical judgment matrix, select the control parameter σ, and fuse R and R _p to get

Repeat the above steps several times until a judgment matrix satisfying consistency is obtained;

(2.4) Integrate the intuitionistic fuzzy hierarchical judgment matrix of each layer to obtain the intuitionistic fuzzy set of each scheme, and calculate the optimal weight ρ of the scheme; the calculation form of the optimal weight is: ρ(x)=0.5(1+π _x )( 1-μ _x ); the scheme with a smaller ρ value has a higher superiority level, and the scheme with the smallest preferred weight ρ is selected as the optimal control law of the unit under extreme conditions of the pumped-storage power station.

Preferably, the feasible solution and the infeasible solution in step (1.3) are characterized in that: if the particle violates the multiple decision space boundary constraints, the particle is classified as an infeasible solution, otherwise, it is classified as a feasible solution. The multiple decision space boundary constraints include speed limit, volute and ball valve pressure limit, draft tube vacuum limit, guide vane and ball valve operating speed limit.

The rising speed limit has the following form: (x _i,max -x _i,r )/x _i,r ≤45%, where x _i,max is the maximum speed of the ith unit, and x _i,r is The speed rating of the i-th unit;

其中，P_vol,i为第i台机组蜗壳水击压力值，

为第i台机组蜗壳水击压力初始值，h_n为工作水头，

为蜗壳水击压力限制值；The water pressure limitation of the volute and the ball valve has the following forms:

Among them, P _vol,i is the water hammer pressure value of the volute casing of the i-th unit,

is the initial value of the volute water hammer pressure of the i-th unit, h _n is the working water head,

is the water hammer pressure limit value of the volute;

其中，P_dra,i为第i台机组尾水管水击压力值，

为第i台机组尾水管水击压力初始值，h_n为工作水头，

为尾水管水击压力限制值；The limit of the vacuum degree of the draft tube has the following form:

Among them, P _dra,i is the water hammer pressure value of the draft tube of the i-th unit,

is the initial value of the water hammer pressure of the draft tube of the i-th unit, h _n is the working head,

is the limit value of the water hammer pressure of the draft pipe;

其中，ΔY为导叶关闭角度，T_simu为单位仿真步长。Y_max为导叶最大开度，

为导叶最快关闭时间，Δθ为球阀关闭角度，θ_max为球阀最大开度，

为球阀最快关闭时间。The operating speed limit of the guide vane and the ball valve has the following forms:

Among them, ΔY is the closing angle of the guide vane, and T _simu is the unit simulation step size. Y _max is the maximum opening of the guide vane,

is the fastest closing time of the guide vane, Δθ is the closing angle of the ball valve, θ _max is the maximum opening of the ball valve,

The fastest closing time for the ball valve.

In general, compared with the prior art, the above technical solutions conceived by the present invention can achieve the following beneficial effects:

(1) The optimal decision-making method for extreme operating conditions of the pumped-storage power station provided by the present invention embeds an improved multi-objective non-dominated genetic algorithm, and comprehensively considers multiple extreme values of unit speed and pressure under extreme operating conditions of the pumped-storage power station. The key objective of the transition process is to incorporate multiple criteria of the control law into the decision-making evaluation basis. The method can solve the optimal control law of the unit under extreme working conditions, which can be applied to engineering practice and obtain excellent dynamic quality of the transition process.

(2) The improved non-dominated genetic algorithm provided by the present invention integrates the piecewise linear chaotic map of the Latin hypercube sampling machine, and embeds an adaptive penalty strategy. In the initial stage of the algorithm, the Latin hypercube sampling strategy is used to generate the initialized particle population uniformly and randomly in the decision space; in the later stage of the algorithm iteration, the piecewise linear chaotic mapping mechanism is used to drive the particles to jump out of the local optimal range and search other areas in the decision space; The proposed adaptive penalty strategy makes full use of the proportional relationship between feasible solutions and infeasible solutions under the current iteration and the information contained in the particles, focusing on finding more feasible solutions in the early stage of the algorithm, and focusing on finding better solutions in the later stage of the algorithm. .

(3) The intuitionistic fuzzy AHP is introduced into the multi-criteria decision-making framework. The intuitionistic fuzzy value can fully reflect the decision-maker's important measure of the goal, including the decision-maker's affirmation, negation and hesitation, which is more in line with the decision-making reality.

(4) The multi-criteria decision-making hierarchy framework for extreme working conditions of the pumped-storage power station constructed by the present invention fully considers the complexity and safety of the dynamic quality and control laws in the transition process of the pumped-storage power station, and the obtained results are more comprehensive and comprehensive.

Description of drawings

FIG. 1 is a schematic flowchart of a method for optimizing the control law of units under extreme working conditions of a pumped-storage power station disclosed in an embodiment of the present invention;

2 is a schematic flowchart of an improved non-dominated genetic algorithm disclosed in an embodiment of the present invention;

3 is a schematic diagram of a multi-criteria decision-making process disclosed in an embodiment of the present invention;

FIG. 4 is a multi-criteria decision-making hierarchy framework for extreme working conditions of a pumped-storage power station disclosed in an embodiment of the present invention;

5 is a multi-criteria decision-making criterion layer intuitionistic fuzzy weight disclosed in an embodiment of the present invention;

Fig. 6 is the transition process of the optimal control law of the units under the successive load rejection disclosed in the embodiment of the present invention, (a) is the rotational speed of the first unit, (b) is the rotational speed of the second unit, and (c) is the worm gear of the first unit Shell pressure value, (d) is the volute pressure value of the second unit, (e) is the vacuum degree of the draft tube of the first unit, and (f) is the vacuum degree of the draft tube of the second unit.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention, but not to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

The invention provides a decision-making method for optimizing the control law of units under extreme working conditions of a pumped-storage power station. .

FIG. 1 is a schematic flowchart of a method for optimizing the control law of units under extreme working conditions of a pumped-storage power station disclosed in an embodiment of the present invention. The method shown in FIG. 1 includes the following steps:

(1) Establish a mathematical model of the transition process of the pumped storage power station under extreme working conditions, input the control law of the unit, and calculate the speed of the unit, the extreme value of the water hammer pressure of the volute, and the pressure value of the draft tube water machine. The dynamic quality multi-objective function of the unit transition process is as The optimization objective is combined with the improved non-dominated genetic algorithm to screen the Pareto solution set of the unit control law;

(2) Establish a multi-criteria decision-making hierarchy framework for the unit control law in extreme conditions of the pumped storage power station, and select the optimal solution in the Pareto solution set of the unit control law as the optimal control law of the unit under extreme conditions of the pumped storage station.

FIG. 2 is a schematic flowchart of an improved non-dominated genetic algorithm disclosed in an embodiment of the present invention. The method shown in FIG. 2 includes the following steps:

(1.1) Set the basic parameters of the optimization algorithm, specifically, set the size of the particle population N, the maximum iteration number T, the current iteration number t, the chaotic mutation condition T _chaotic , the decision space dimension D, the cross recombination probability P _c , the polynomial mutation probability P _m ;

(1.2) Use Latin hypercube sampling to initialize the particle parent population. Specifically, the D-dimensional decision space is evenly divided into N non-overlapping equally spaced regions, and a point is randomly selected from each equally spaced region of each dimension as Particle decision variable, generating N initialized particles as the parent particle swarm P;

(1.3) Execute the adaptive penalty function to correct the parent particle optimization target value, specifically, according to the quantitative relationship between feasible solutions and infeasible solutions in the current iteration loop, calculate the penalty factor and apply it to the particle optimization target value;

(1.4) Execute non-dominated sorting, championship selection, cross-recombination, and polynomial mutation to generate a population of offspring particles, and calculate the optimal target value of offspring particles;

(1.5) Chaos judgment, specifically, judge whether the current iteration number t is greater than the chaotic mutation condition T _chaotic , if so, execute step (1.6), otherwise go to step (1.7);

(1.6) Implement piecewise linear chaotic mapping mechanism;

(1.7) Integrate parent and child particle swarms, perform fast non-dominated genetic sorting based on non-dominated sequence and crowding degree distance, and select new parent particles;

(1.8) End judgment, specifically, judge whether the current iteration number t is less than the maximum iteration number T, if so, jump out of the loop, and use the current solution set as the Pareto solution set; otherwise, the current iteration number t=t+1, with the current solution set Set as the parent particle swarm P of the next cycle, and enter step (1.3).

FIG. 3 shows a multi-criteria decision-making process within a control law decision-making framework of a pumped-storage power station under extreme working conditions disclosed in an embodiment of the present invention. The index is decomposed and a multi-level decision-making framework is constructed. Specifically, the decision-making of the unit control law in extreme working conditions of the pumped-storage power station is divided into the target layer, the criterion layer, and the scheme layer. The target layer is the optimal control law for extreme working conditions of the pumped storage power station; the criterion layer includes the maximum speed increase value of the unit, the maximum water pressure value at the inlet of the volute, the vacuum pressure value at the outlet of the draft tube, the maximum water pressure value at the ball valve, and the controller. Quantitative indicators such as the maximum oil velocity of the ball valve and qualitative indicators such as the fluctuation degree of the rotational speed of the unit, the fluctuation degree of the volute draft tube pressure, the fluctuation degree of the water level of the upstream and downstream surge chambers, the operational complexity of the control law, and the operational safety of the control law; The Pareto solution set obtained by the multi-objective optimization algorithm, the multi-level decision-making framework is shown in Figure 4.

In the process shown in Figure 3, the following steps are included:

(2.1) Use the intuitionistic fuzzy set to compare the indicators at the same level. Specifically, the intuitionistic fuzzy set (x|μ _x , ν _x , π _x ) is used as the judgment base, and the decision-making hierarchy is compared layer by layer from bottom to top. Form the intuitionistic fuzzy hierarchical judgment matrix R;

(2.2) Consistency test judgment, the judgment matrix that satisfies the consistency test is considered to be consistent with the judgment, and goes to step (2.4), otherwise it goes to step (2.3);

多次重复上述步骤直至获得满足一致性的判断矩阵；(2.3) Perform consistency correction, specifically, modify the intuitionistic fuzzy hierarchical judgment matrix, calculate the complete consistency matrix R _p of the intuitionistic fuzzy hierarchical judgment matrix, select the control parameter σ, and fuse R and R _p to obtain

Repeat the above steps several times until a judgment matrix satisfying consistency is obtained;

(2.4) Integrate the intuitionistic fuzzy hierarchical judgment matrix of each layer, calculate the optimal weight of the scheme, and select the scheme with the smallest optimal weight as the optimal control law of the units in extreme conditions of the pumped storage power station.

Figure 5 is a schematic diagram of the intuitionistic fuzzy hierarchical weights of the multi-criteria decision-making target layer in the extreme working conditions of the pumped storage power station according to the embodiment of the present invention. The vacuum degree of the draft tube and the operational safety are the quantitative and qualitative indicators that experts are most concerned about, respectively. The importance obtained by the fuzzy hierarchical ranking is also in line with expert cognition.

The control law finally obtained by the optimization solution in this embodiment is applied to the working condition that two units enter the load shedding successively in the form of one-pipe and two-unit arrangement. The results are shown in Figure 6. The elite file set contains 40 individuals, as shown in Table 1. shown. The comparison between the node extreme value of the control law under successive load shedding conditions and the actual measurement of the power station is shown in Table 2. It can be seen from Figure 6 and Table 2 that the optimal control law of the unit obtained by this method meets the requirements of the adjustment guarantee calculation, retains a large margin, and greatly reduces the speed increase of the unit and the volute. At the same time, the vacuum degree at the draft tube is effectively improved, and the dynamic quality of the transition process is good. The verification shows that the multi-objective optimization and multi-criteria decision-making method of the control law of the present invention under extreme working conditions of the pumped-storage power station is correct and effective to optimize the decision-making of the optimal control law of the pumped-storage power station under extreme working conditions.

Table 1 External file set and detailed parameters

Table 2 Optimal control law, field measurement and standard node extreme value of adjustment and maintenance calculation under successive load shedding conditions

Those skilled in the art can easily understand that the above are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present invention, etc., All should be included within the protection scope of the present invention.

Claims (8)

1. an optimal decision-making method for unit control law under extreme working conditions of a pumped-storage power station, is characterized in that, comprises: (1) Establish a mathematical model for the transition process of the pumped storage power station under extreme conditions, input the unit control law, calculate the unit speed, the extreme value of the water hammer pressure of the volute, and the extreme value of the water hammer pressure of the draft tube, and use the multi-objective function of the dynamic quality of the unit transition process In order to optimize the goal, the Pareto solution set of the unit control law is screened in combination with an improved non-dominated genetic algorithm, which integrates Latin hypercube sampling and piecewise linear chaotic mapping, and embeds an adaptive penalty strategy to solve multiple decision space boundary constraints; (2) Establish a multi-criteria decision-making hierarchy framework for the unit control law of the pumped-storage power station under extreme working conditions, synthesizing qualitative and quantitative indicators, and select the optimal solution from the Pareto solution set of the unit control law as the optimal control of the unit under extreme working conditions of the pumped-storage power station law. 2. method according to claim 1, is characterized in that, optimization target comprises unit rotational speed maximum rising relative value and water hammer pressure compound extreme value; The calculation form of the relative value of the maximum increase in the rotational speed of the unit is:

Among them, n is the number of units, x _i is the rotational speed of the i-th unit, and x _r,i is the rated speed of the i-th unit; The calculation form of the composite extreme value of the water hammer pressure is:

Among them, P _vol,i and P _dra,i are the water hammer pressure values of the volute and draft tube, respectively,

is the rated value of the water hammer pressure of the draft tube, and I _v and I _d are the weight coefficients of the importance of the volute and the draft tube, respectively. 3. The method according to claim 2, wherein the improved non-dominated genetic algorithm comprises the following steps: (1.1) Set the size N of the particle population, the maximum iteration number T, the current iteration number t, the chaotic mutation condition T _chaotic , the decision space dimension D, the cross-recombination probability P _c , and the polynomial mutation probability P _m ; (1.2) Execute the Latin hypercube sampling strategy to initialize the particle population, divide the D-dimensional decision space into N non-overlapping equally spaced regions, and randomly select a point from each equally spaced region in each dimension as the particle decision variable, Generate N initialized particles as the parent particle swarm P; (1.3) According to the quantitative relationship between feasible solutions and infeasible solutions in the current iteration loop, implement an adaptive penalty strategy, and calculate the revised optimization target value; the adaptive penalty strategy has the following form:

Among them, f _m (X _i (t)) is the m-th optimization target value of the i-th particle, and p(X _i (t)) is the penalty function added to the particle; the penalty function has two penalty factors M (X _i (t)), N(X _i (t)), the calculation form is: p(X _i (t))=(1-r _f )M(X _i (t))+r _f N(X _i (t))

in,

is the sum of constraint violations for the ith particle,

are the maximum and minimum optimization objective values of feasible solutions in the particle population, respectively, and r _f is the quantitative relationship of feasible solutions in the particle population; (1.4) Execute non-dominated sorting, generate child particle swarm C from parent particle swarm P through strategies such as tournament selection, cross-recombination and polynomial mutation, and calculate the optimal target value of particles in child particle swarm C; (1.5) Determine whether the current iteration number t is greater than the chaotic mutation condition T _chaotic , if so, execute step (1.6), otherwise go to step (1.7); (1.6) Execute a piecewise linear chaotic map, the calculation form of the piecewise linear chaotic map is:

Among them, p∈(0, 0.5) is the control parameter, and x _i ∈(0,1) is the digital chaotic pseudo-random sequence; (1.7) Integrate the parent particle swarm P and the child particle swarm C to form a family particle swarm Ω, and select N preferred particles from the family particle swarm Ω to form an elite file set E based on the non-dominated sorting and crowding degree; The preferred particles are selected according to the following principles: first, they are selected from the particles with the highest non-dominant ranking in the family particle group Ω. When the number of particles in this rank is greater than N, the crowding degree of the particles is used as the preferred basis to select Particles with a smaller crowding degree are put into the elite file set E. If the particle swarm with the highest non-dominant ranking is less than N, it is selected from the particles of the next non-dominant ranking until the elite file set E contains N preferred particles; (1.8) Determine whether the current number of iterations t is less than the maximum number of iterations T, if so, jump out of the loop and take the current elite archive set R as the Pareto solution set; otherwise, the current iteration number t=t+1, take the current elite archive set R as the Pareto solution set The parent particle swarm P of the next cycle goes to step (1.3). 4 . The method according to claim 1 , wherein the multi-criteria decision-making hierarchy framework of the extreme working condition unit control law of the pumped-storage power station comprises a target layer, a criterion layer, and a scheme layer; the target layer is pumped-storage energy storage. 5 . The optimal control law under extreme working conditions of the power station; the criterion layer includes qualitative and quantitative indicators, and the quantitative indicators include the maximum rotational speed rise value of the unit, the maximum water pressure value at the inlet of the volute, the vacuum pressure value at the outlet of the draft tube, and the ball valve. The maximum water pressure value at the location, the maximum oil speed of the controller and the ball valve; the qualitative indicators include the fluctuation degree of the unit speed, the pressure fluctuation degree of the volute draft tube, the fluctuation degree of the water level of the upstream and downstream surge chambers, the operation complexity of the control law, and the operation of the control law. Security; the scheme layer is the Pareto solution set obtained by solving the multi-objective optimization scheme. 5. method according to claim 4 is characterized in that, based on intuitionistic fuzzy analytic hierarchy process optimization decision-making unit control law, the method is specifically: (2.1) Take the intuitionistic fuzzy set (x|μ _x , ν _x , π _x ) as the judgment base, compare the decision-making hierarchy layer by layer from bottom to top, and form the intuitionistic fuzzy hierarchical judgment matrix R; where the intuitionistic fuzzy values μ _x , ν _x , π _x are membership degree, non-membership degree and hesitation degree, respectively; (2.2) Consistency test is carried out on the judgment matrix R of the intuitionistic fuzzy hierarchy, and the judgment matrix satisfying the consistency test is considered to be consistent with the judgment consistency, and then go to step (2.4), otherwise, go to step (2.3); (2.3) Modify the intuitionistic fuzzy hierarchical judgment matrix, calculate the complete consistency matrix R _p of the intuitionistic fuzzy hierarchical judgment matrix, select the control parameter σ, and fuse R and R _p to get

Repeat the above steps several times until a judgment matrix satisfying consistency is obtained; (2.4) Integrate the intuitionistic fuzzy hierarchical judgment matrix of each layer to obtain the intuitionistic fuzzy set of each scheme, and calculate the optimal weight ρ of the scheme; the calculation form of the optimal weight is: ρ(x)=0.5(1+π _x )( 1-μ _x ); the scheme with a smaller ρ value has a higher superiority level, and the scheme with the smallest preferred weight ρ is selected as the optimal control law of the unit under extreme conditions of the pumped-storage power station. 6. The method according to claim 3, wherein the mathematical model of the extreme working condition transition process of the pumped-storage power station is provided with multiple decision space boundary restrictions; Ball valve water pressure limit, draft tube vacuum limit, guide vane and ball valve operating speed limit; if the particle violates any decision space boundary limit, the particle is classified as an infeasible solution, otherwise it is classified as a feasible solution; The rising speed limit has the following form: (x _i,max -x _i,r )/x _i,r ≤45%, where x _i,max is the maximum speed of the ith unit, and x _i,r is The speed rating of the i-th unit; The water pressure limitation of the volute and the ball valve has the following forms:

Among them, P _vol,i is the water hammer pressure value of the volute casing of the i-th unit,

is the initial value of the volute water hammer pressure of the i-th unit, h _n is the working water head,

is the water hammer pressure limit value of the volute; The limit of the vacuum degree of the draft tube has the following form:

Among them, P _dra,i is the water hammer pressure value of the draft tube of the i-th unit,

is the initial value of the water hammer pressure of the draft tube of the i-th unit, h _n is the working head,

is the limit value of the water hammer pressure of the draft pipe; The operating speed limit of the guide vane and the ball valve has the following forms:

Among them, ΔY is the closing angle of the guide vane, T _simu is the unit simulation step size, Y _max is the maximum opening of the guide vane,

is the fastest closing time of the guide vane, Δθ is the closing angle of the ball valve, θ _max is the maximum opening of the ball valve,

The fastest closing time for the ball valve. 7. The method according to claim 1, wherein the mathematical model of the transition process under extreme working conditions of the pumped-storage power station comprises a pressurized water flow system, a pump-turbine full characteristic curve, a governor and a servo system thereof; The pressurized water system includes a pressure pipeline, a pressure regulating chamber and a ball valve. 8 . The method according to claim 1 , wherein the unit control law comprises a straight line closing rule of a guide vane; a guide vane two-segment broken line closing rule; a guide vane ball valve co-connected two-stage closing rule; a guide vane delaying a straight line The two-segment broken line of the ball valve is closed.

Images