CN109765789B - Crude oil blending double-blending head coordination optimization method considering tank bottom oil property - Google Patents

Abstract

The invention discloses a crude oil blending double-blending head coordination optimization method considering tank bottom oil properties. The method comprises the following steps: firstly, initializing blending task parameters; secondly, setting an optimization period, an objective function weight, and setting upper and lower limits of each attribute index of the blended crude oil in each blending tank, the storage amount of each component oil used by each blending head, the maximum flow of each component oil of each blending line, the maximum flow of a common component and the unit mass cost of each component oil; thirdly, acquiring the blended component oil and the attribute data of the bottom oil of each tank according to the optimization period, and updating the oil dipstick at the bottom of each tank, the reserve capacity of the component oil and the blending residual time of the batch; and finally, solving the optimal formula of each blending component in the current optimization period, and sending the optimal formula to a blending control system for execution.

Description

Coordination optimization method of crude oil blending double blending head considering the properties of tank bottom oil

technical field

The invention belongs to the field of crude oil blending, and in particular relates to a crude oil blending double blending head coordination optimization method considering the properties of tank bottom oil.

Background technique

In actual crude oil blending, there are often cases where two blending heads share one component oil, and the flow rate of the shared component oil is limited, and sometimes the needs of the two blending heads cannot be met at the same time. Each blending head is coordinated, so as to achieve the optimal overall economic index and meet the target requirements of the properties of the crude oil blended in each blending buffer tank.

For the multi-constraint and nonlinear characteristics of coordination problems, traditional methods are difficult to directly solve such problems due to the limitation of algorithm performance. Usually, the two blending heads are optimized and solved separately. When the actual usage of common components is greater than their total When the flow is limited, the upper bound of the common component formula of each blending head is reduced to meet the requirements by combining the current optimization value, and then the methods of re-optimizing the two blending heads are coordinated respectively. This method of reducing the upper bound actually limits the search range of the algorithm, so that a complete search of the entire feasible region cannot be achieved, and the global optimal solution is often excluded from the new range, so the global optimal cannot be achieved. And because each blending head needs to be calculated twice, a total of 4 calculations are required, which leads to a long solution time, which limits the control period of the blending process and is not conducive to further improvement of the production level.

Therefore, there is a need in the art for a coordination optimization method that can achieve sufficient and effective coordination of the two blending heads in the entire range and reduce the solution time by only one optimization solution.

SUMMARY OF THE INVENTION

The invention aims to directly solve the situation of the double blending head, and applies the intelligent evolutionary algorithm to the coordination problem of the double blending head of crude oil to help further improve the blending level of the crude oil.

The invention provides a crude oil blending double blending head coordination optimization method considering the properties of tank bottom oil, and the method comprises the following steps:

First, initialize the parameters of the blending task;

Secondly, set the optimization period, the weight of the objective function, and set the upper and lower limits of each attribute index of the blended crude oil in each blending tank, the oil reserves of each component used in each blending head, the maximum flow rate of each component oil in each refining line, and the common components. Maximum flow and unit mass cost of each component oil;

Thirdly, obtain the attribute data of each blended component oil and each tank bottom oil according to the optimization cycle, and update each tank bottom oil dipstick, component oil reserve, and the remaining blending time of this batch; and

Finally, the optimal formula of each blending component in the current optimization cycle is obtained and sent to the blending control system for execution.

In another preferred example, the blending task parameters include the corresponding storage tank number of the component oil and the blending tank number corresponding to each blending head, and a common component is set; the formula of the component oil in each blending head The upper and lower limits of the blending head, the flow rate of the blending head, the oil dipstick at the bottom of the tank, the target blending oil dipstick, the target blending density value and the blending time of this batch.

In another preferred example, the optimized objective function is

Among them, k (k=1,2,...,n) represents the blending head label;

i(i=1,2,...,n) represents the oil product label of the blending component;

r _i,k represents the formula of the i-th component oil of the k-th blending head to be optimized;

l(l=1,2,...,n) represents the crude oil attribute index;

c _i,k represents the unit mass cost of the i-th component oil of the k-th blending head;

w _c , w _g represent the minimum weight of the blending cost, and the minimum weight of the deviation between the crude oil density and the target density after blending;

Ρ _m,k represents the target density blended by the kth blending head;

P _ρ,k represents the instantaneous predicted value of the blending density of the component oil in the kth blending head;

TP _k (P _ρ,k ) represents the predicted value of crude oil density in the k-th blending tank;

t represents the remaining blending time of this batch;

Bc _i,k represents the maximum flow rate of the i-th component mixing line of the k-th blending head;

B _k represents the flow of the k-th blending head in this batch;

B _{com_max} represents the maximum flow of common components;

r _comk represents the formula of the common component at the kth blending head;

Vi _,k represents the reserves of the i-th component oil of the k-th blending head;

r _{i,k_min} , r _{i,k_max} respectively represent the upper and lower limits of the i-th component oil formula of the k-th blending head;

s _{l, k_min} , s _{l, k_max} respectively represent the upper and lower limits of the attribute value l of the k-th blending tank;

P _l,k represents the blending prediction value of the component oil property l of the kth blending head;

TP _k (P _l,k ) represents the predicted value of crude oil property l in the k-th blending tank.

In another preferred example, formulas (2) and (3) are used to calculate the crude oil property index of the whole tank to obtain the predicted values TP _k (P _ρ,k ), TP _k of the crude oil density and other properties in each blending tank. (P _l,k ):

Among them, TVOL _k represents the blending target quality of the kth blending tank;

HVOL _k represents the quality of the bottom oil of the k-th blending tank;

T _ρ,k represents the tank bottom oil density value of the kth blending tank;

T _l,k represents the value of the bottom oil property l of the kth blending tank.

In another preferred embodiment, a mathematical model for the coordination problem of crude oil blending dual-blending heads considering the properties of tank bottom oil is established, and a composite differential evolution (CoDE, Composite Differential Evolution) optimization method is used to solve the coordination optimization problem.

In another preferred embodiment, the hybrid differential evolution optimization method includes the steps:

(1) Initialize the population, and the population size is NP;

(2) The loop starts, and when the algorithm termination condition has not been met, proceed as follows:

(i) For each individual _xi in the population, use three test vector generation strategies, and randomly select a set of parameters for each strategy to generate three test vectors u _i1 , u _i2 and u _i3 corresponding to _xi ;

(ii) evaluating the fitness function values of u _i1 , u _i2 and u _i3 ;

(iii) Select the test vector with the optimal fitness function value among u _i1 , u _i2 and u _i3 , denoted as u _i ;

(iv) Compare the fitness function values of _xi and _ui , if _xi is better than _ui , then _xi enters the next generation population; if u _i is better than _xi , then _ui replaces _xi and enters the next generation population;

(3) The algorithm stops and the final population NP is obtained, and the individual with the best fitness value in the population is the solution of the optimization problem.

In another preferred example, the hybrid differential evolution optimization method is implemented using three test vector generation strategies of formula (4), (5) and formula (6):

(1) "rand/1/bin" method

(2) "rand/2/bin" method

(3) "current-to-rand/1" method

u _i,G = _xi,G +rand·(x _r1,G -xi _,G )+F·(x _r2,G -x _r3,G ) (6)

Among them, x _{r1, G} , x _{r2, G} , x _{r3, G} , x _{r4, G} , x _{r5, G} are 5 different individuals randomly selected in the G generation, which are different from the current individual x _{ri, G} Also not equal.

Accordingly, the present invention provides a coordination optimization method that can achieve sufficient and effective coordination of the two blending heads in the entire range and reduce the solution time by only one optimization solution.

Description of drawings

Figure 1 is a process diagram of a crude oil blending process with dual blending heads and blending tanks.

Fig. 2 is a flow chart of the intelligent coordination optimization method of the dual-tuning head.

Figure 3 is a flowchart of the hybrid differential evolution optimization algorithm.

Detailed ways

Aiming at the problems of inability to perform global optimization and low calculation efficiency caused by the traditional method of reducing the upper limit while adopting the double-blending head coordination problem in the crude oil blending production of oil refining enterprises, the present invention proposes a method considering tank bottom oil. Attributed crude oil blending dual blending head intelligent coordination optimization method. This method takes the lowest cost and the smallest deviation between the crude oil density and the target density after blending as the optimization objective. At the same time, various crude oil quality index constraints (sulfur content, acid value, naphtha yield, nitrogen content, etc.), inventory restrictions of each component oil, formulation restrictions and flow restrictions of each component oil are considered. The hybrid differential evolution optimization algorithm is introduced to improve the calculation efficiency, and the optimization and coordination of the dual mixing heads can be realized only through one optimization solution. , improve production efficiency.

The technical scheme of the present invention is:

The intelligent coordination optimization method of crude oil blending double blending head considering the properties of tank bottom oil includes the following steps:

First, initialize the parameters of the blending task, select the corresponding storage tank number and blending tank number of the component oil corresponding to each blending head, and set the common components. Set the upper limit and lower limit of the component oil formula in each blending head, the flow rate of each blending head, the oil dipstick at the bottom of the tank, the target blending oil dipstick, the target blending density value and the blending time of this batch; secondly, set the optimization Period and weight of objective function, set the upper and lower limits of each attribute index of the blended crude oil in each blending tank, the reserves of each component oil used in each blending head, the maximum flow rate of each component oil in each reference refining line, the maximum flow rate of common components, and the maximum flow rate of each component. The unit mass cost of component oil; thirdly, according to the optimization cycle, obtain the attribute data of each blended component oil and each tank bottom oil, and update each tank bottom oil dipstick, component oil reserve, and the remaining blending time of this batch; Finally, under certain constraints, the optimal formula of each blending component in the current optimization cycle is obtained by using the hybrid differential evolution optimization algorithm, and sent to the blending control system for execution.

In the present invention, the method aims to achieve the lowest cost and the smallest deviation between the crude oil density and the target density after blending under certain constraints. For a single optimization cycle, the optimal blending formula is obtained by solving equations (1), (2) and (3):

Among them, k(k=1,2,...,n) represents the blending head label; i(i=1,2,...,n) represents the blending component oil product label; ri _,k represents the first to be optimized The formula of the i-th component oil of k blending heads; l (l=1,2,...,n) represents the crude oil attribute index (sulfur content, acid value, naphtha yield, nitrogen content, etc.); c _i,k represents the unit mass cost of the i-th component oil of the k-th blending head; w _c , w _g represent the minimum weight of the blending cost, the minimum weight of the deviation between the crude oil density and the target density after blending; P _{m ,k} represents the target density for blending of the kth blender; P _ρ,k represents the instantaneous predicted value of the blending density of the component oil in the kth blender; TP _k (P _ρ,k ) represents the kth blending density The predicted value of crude oil density in the blending tank; t represents the remaining blending time of the batch; Bc _i,k represents the maximum flow rate of the i-th component of the k-th blending head; B _k represents the k-th component The flow rate of the blending head in this batch; B _{com_max} represents the maximum flow rate of the common component; r _comk represents the formula of the common component in the k-th blending head; Vi _,k represents the i-th group of the k-th blending head The reserve of oil separation; ri _{,k_min} , ri _{,k_max} respectively represent the upper and lower limits of the ith component oil formula of the kth blending head; s _{l,k_min} , sl _{,k_max} respectively represent the kth blending tank The upper and lower limits of the attribute value l; P _l,k represents the blending prediction value of the oil attribute l of the kth blending head; TP _k (P _l,k ) represents the crude oil attribute l in the kth blending tank. Predictive value.

The predicted values P _ρ,k and P _l,k of the blending density and properties of each blending head component oil are calculated according to the principle of linear superposition, and then use the density and properties to compensate the density and properties of the blended tank bottom oil deviation, so that the density of the entire blending tank reaches the target value and the properties are qualified. The predicted values TP _k (P _ρ,k ) and TP _k (P _l,k ) of the crude oil density and other properties in each blending tank can be calculated according to the following formulas (2) and (3) according to the crude oil attribute index of the whole tank. calculate:

Among them, TVOL _k represents the blending target quality of the k-th blending tank; HVOL _k represents the tank bottom oil quality of the k-th blending tank;

T _ρ,k represents the density value of the bottom oil of the kth blending tank; T _l,k represents the value of the bottom oil property l of the kth blending tank.

Finally, the hybrid differential evolution optimization method is used to coordinate this problem, and the steps are:

1) Initialize the population, and the population size is NP;

2) The loop starts, and when the algorithm termination condition has not been met, proceed:

i) For each individual _xi in the population, use three test vector generation strategies, and randomly select a set of parameters for each strategy to generate three test vectors u _i1 , u _i2 and u _i3 corresponding to _xi ;

ii) evaluating the fitness function values of u _i1 , u _i2 and u _i3 ;

iii) Select the test vector with the optimal fitness function value among u _i1 , u _i2 and u _i3 , denoted as u _i ;

iv) Comparing the fitness function values of _xi and _ui , if _xi is better than _ui , then _xi enters the next generation population;

If u _i is better than _xi , then u _i replaces _xi and enters the next generation population.

3) The algorithm stops and the final population NP is obtained, and the individual with the best fitness value in the population is the solution of the optimization problem.

The three test vector generation strategies used are:

(1) "rand/1/bin" method

(2) "rand/2/bin" method

(3) "current-to-rand/1" method

u _i,G = _xi,G +rand·(x _r1,G -xi _,G )+F·(x _r2,G -x _r3,G ) (6)

where x _i,G represents the i-th individual in the G-th generation, x _i,j,G represents its j-th component, and u _i,G represents the cross-variant individual corresponding to the i-th individual in the G-th generation, u _i,j,G represents its jth component, r ₁ , r ₂ , r ₃ , r ₄ , r ₅ are the numbers of five individuals randomly selected from the G generation individuals, and are different from the corresponding i in the formula , and F is the scale factor, which is a list of constant vectors.

The above features mentioned in the present invention or the features mentioned in the embodiments can be combined arbitrarily. All the features disclosed in this specification can be used in combination with any composition, and each feature disclosed in the specification can be replaced by any alternative features that serve the same, equivalent or similar purpose. Therefore, unless otherwise stated, the disclosed features are only general examples of equivalent or similar features.

The main advantages of the present invention are:

The invention proposes an intelligent coordination optimization method for crude oil blending double blending heads considering the properties of tank bottom oil. And the hybrid differential evolution optimization algorithm is used to optimize the solution, which is faster than the traditional method and truly realizes the global optimization. Under certain constraints, the full coordination of the dual blending heads is achieved, which not only meets the lowest economic cost, but also achieves the smallest deviation between the crude oil density and the target density after blending, thus achieving the optimal overall benefit.

The obvious defects of the existing coordination algorithms are: (1) It is necessary to judge whether the coordination conditions are triggered, and then reduce the flow of each blending head by an equal proportion to make it meet the current constraints; (2) The objective function is only a single The cost is the lowest (and linear), and in fact there are other objectives such as stationary rate control, which will greatly increase the computational difficulty for multi-objective and nonlinear situations. The intelligent evolutionary difference algorithm adopted in the present invention has the function of one-time solution, and can make up for the aforementioned defects.

The present invention will be further described below with reference to the accompanying drawings and embodiments, but the present invention is not limited to this embodiment.

Embodiment 1 is solved by the method involved in the present invention, and implemented on the premise of the technical solution of the present invention,

Embodiment 2 is based on the traditional method of solving each blending head separately, and then reducing the upper limit.

Example 1

This embodiment is implemented on the premise of the technical solution of the present invention, but the protection scope of the present invention is not limited to the following embodiments. The technical process of the crude oil blending process with dual blending heads and blending tanks involved in this embodiment is shown in FIG. 1 , and the blending head 1 selects groups of storage tanks with tank numbers of 1#, 2# and 3# storage tanks. Separate the oil for blending, and select the corresponding blending tank 1. The blending head 2 selects the component oils whose storage tank numbers are 3#, 4# and 5# for blending, and also selects the corresponding blending tank 2. It can be seen from the above. 3# component oil will be the common component.

As shown in Figure 2, the workflow of the intelligent coordination optimization method of the dual harmonic head mainly includes the following steps:

Step 1: Initialize the parameters of the blending task.

Set the component oil corresponding to each blending head, select the corresponding tank number and blending tank number, and set the 3# component oil as the common component.

In blending head 1, set the lower limit of the formula of 1# component oil to 0, and the upper limit of the formula to 0.35; the lower limit of the formula of 2# component oil is 0.2, and the upper limit of the formula is 0.72; the lower limit of the formula of 3# component oil is 0, and the upper limit of the formula is 0.4 .

In blending head 2, set the lower limit of the 3# component oil formula to 0, and the upper limit of the formula to 0.4; the lower limit of the 4# component oil formula is 0, and the upper limit of the formula is 0.4; the lower limit of the 5# component oil formula is 0, and the upper limit of the formula is 0.5 .

Set the blending flow rate in blending head 1 to 1000t/h, the quality of tank bottom oil to 1000t, the target blending volume of this batch to be 11000t, and the target density of crude oil after blending to be 830kg/m ³ .

Set the blending flow rate in blending head 2 to 1000t/h, the quality of tank bottom oil to be 1000t, and the target blending volume of this batch to be 9000t.

The target density of the crude oil after blending is 830kg/m ³ .

The blending time of this batch is 10h.

Step 2: Set the optimization period, the weight of the objective function, the upper and lower limits of the optimization, the common components, the maximum flow rate of the common components, the oil reserves of the components used by each blending head, and the unit mass cost of each component oil.

The optimization period is set to 5min, the minimum weight of the blending cost is 0.4, and the minimum weight of the deviation between the crude oil density and the target density after blending is 0.6. The lower limit of sulfur content in the blending tank after blending in blending head 1 is set to 0%, and the upper limit is 2.5%; the lower limit of acid value is 0mgKOH, and the upper limit is 0.5mgKOH; the lower limit of naphtha yield is 0.19%, and the upper limit is 0.21%.

The lower limit of sulfur content in the blending tank after blending in blending head 2 is set to 0%, and the upper limit is 2.5%; the lower limit of acid value is 0mgKOH, and the upper limit is 0.5mgKOH; the lower limit of naphtha yield is 0.17%, and the upper limit is 0.19%.

The stocks of components 1#, 2# and 3# are set to be 4000t, 7200t and 4500t respectively, and the stocks of component 4# and 5# oil are set to be 4000t and 5000t respectively.

The maximum flow rate and unit mass cost of each component oil ginseng refining line are shown in Table 1. Since the 3# component oil is a common component, the maximum flow rate of the common component is equal to the maximum flow rate of the 3# component oil ginseng refining line, which is 410t. /h.

Step 3: Obtain the attribute data of each blended component oil and tank bottom oil according to the optimization cycle, and update the tank bottom oil dipstick, component oil reserves, and the remaining blending time of this batch.

The property data of each blending component oil and each tank bottom oil are obtained as shown in Table 1 below:

Table 1

The quality of the bottom oil in the obtained blending tank 1 is 1000t, the quality of the bottom oil in the blending tank 2 is 1000t, and the stocks of the obtained 1#, 2#, 3#, 4# and 5# component oils are 4000t and 7200t respectively. , 4500t, 4000t and 5000t. The remaining time for blending in this batch is 10h.

Step 4: Calculate the optimal formula.

Using the above data, after establishing the crude oil blending double blending head coordination model (such as formula (1)), the hybrid differential evolution optimization algorithm is called to solve, and the solution process is shown in Figure 3.

It can be calculated that the formula of 1# component oil corresponding to blending head 1 is 0.23178, the 2# component oil formula is 0.60156, the 3# component oil formula is 0.16666, and the cost is US$86.9522/barrel. The final density of the oil in the obtained blending tank is 0.23178. It is 830.1262kg/m ³ , the final properties of the oil in the blending tank are 1.9915% sulfur content, 0.37815 mg KOH acidity, and 0.20613% naphtha yield.

The formula of 3# component oil corresponding to blending head 2 is 0.3041, the formula of 4# component oil is 0.19604, the formula of 5# component oil is 0.49986, the cost is 83.2672 US dollars/barrel, and the final density of the oil in the obtained blending tank is 830.072kg/ m ³ , the final properties of the oil in the blending tank are 2.2064% sulfur content, 0.44496 mg KOH acidity, and 0.17598% naphtha yield. The total flow of common components is 409.9427t/h, and the constraint is 410t/h. The final solution time is 6.8125 seconds, and the objective function value is 68.1004.

Step 5: Send the current optimal formula to the blending control system for execution.

Step 6: Determine whether the blending is completed. If the blending is not completed, wait for an optimization cycle, and then return to Step 3.

The methods not involved in the present invention are the same as those in the prior art, and can be implemented by adopting the prior art.

Example 2

This embodiment is implemented on the premise of the traditional technical solution.

The traditional method performs the optimization and solution of the above objective function for a single blending head respectively, and the optimal formula value of the common components of blending head 1 is rc _{1, best} , and the optimal formula value of the common components of blending head 2 is rc _{2, best} , if the usage of the common component is greater than the maximum flow rate of the common component, that is, B1×rc _{1, best} +B2×rc _{2, and best} >B _{com_max} , make the common components in blending head 1 and blending head 2 The upper bounds of the recipes become

rc_max ₁ =rc _{1, best} ×B _{com_max} /(B1×rc _{1, best} +B2×rc _{2, best} ) (7)

rc_max ₂ =rc _{2, best} ×B _{com_max} /(B1×rc _{1, best} +B2×rc _{2, best} ) (8)

Then, the two tuning heads are optimized and solved respectively.

Using the data in Example 1 to solve, the first solution obtains the formula of 1# component oil in blending head 1 0.27935, the 2# component oil formula 0.32065, the 3# component oil formula 0.4, and the 3# component oil formula in blending head 2 Component oil formula 0.30601, 4# component oil formula 0.19399, 5# component oil formula 0.5.

The flow of common components in blending head 1 is 400, the flow of common components in blending head 2 is 244.8094, and the actual total flow of common components is 644.8094, which is greater than the maximum flow limit value of 3# component oil of 410. Change the formula of 3# component oil in the two blending heads according to formulas (7) and (8), and obtain rc_max ₁ =0.25434, rc_max ₂ =0.19458.

Resolve the two blending heads separately, and obtain the formula of 1# component oil in blending head 1 0.25022, the 2# component oil formula 0.49544, the 3# component oil formula 0.25434, and the 3# component in blending head 2. Oil formula 0.19458, 4# component oil formula 0.30542, 5# component oil formula 0.5. Computation time 14.4907 seconds. The results of comparing the two examples are shown in Table 2:

Table 2

It can be seen from the above table that the value of the objective function obtained by the method involved in the present invention in Example 1 is smaller, more sufficient global coordination is achieved, and computing time is significantly saved. In terms of cost target and density index, the method involved in this invention also shows better performance than the traditional method.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the scope of the essential technical content of the present invention. The essential technical content of the present invention is broadly defined within the scope of the claims of the application, and any technical entity completed by others Or method, if it is exactly the same as that defined in the scope of the claims of the application, or an equivalent change, it will be deemed to be covered by the scope of the claims.

Claims (4)

1. A crude oil blending double blending head coordination optimization method considering tank bottom oil properties is characterized by comprising the following steps:

firstly, initializing blending task parameters;

secondly, setting an optimization period, an objective function weight, and setting upper and lower limits of each attribute index of the blended crude oil in each blending tank, the storage amount of each component oil used by each blending head, the maximum flow of each component oil of each blending line, the maximum flow of a common component and the unit mass cost of each component oil;

thirdly, acquiring the blended component oil and the attribute data of the bottom oil of each tank according to the optimization period, and updating the oil dipstick at the bottom of each tank, the reserve capacity of the component oil and the blending residual time of the batch; and

finally, the optimal formula of each blending component in the current optimization period is obtained and sent to a blending control system for execution;

the blending task parameters comprise the corresponding storage tank numbers of the component oil and the blending tank numbers corresponding to the blending heads, and public components are set; the upper limit and the lower limit of the component oil formula in each blending head, the flow of the blending head, the oil dipstick at the bottom of the tank, the target blending oil dipstick, the target blending density value and the blending time of the batch;

the optimized objective function is

Wherein k (k ═ 1,2, …, n) denotes a blend header number;

i (i ═ 1,2, …, n) denotes the blending component oil designation;

r_i,krepresenting the formulation of the i component oil of the k blending head to be optimized;

l (l ═ 1,2, …, n) represents a crude oil property index;

c_i,kexpressing the unit mass cost of the ith component oil of the kth blending head;

w_c，w_gexpressing the lowest weight of blending cost, and the minimum weight of deviation between the density of the blended crude oil and the target density;

Ρ_m,krepresenting a target density for the kth blending head blending;

P_ρ,krepresenting the blending density instantaneous predicted value of the component oil in the kth blending head;

TP_k(P_ρ,k) Representing a predicted value of crude oil density in the kth blending tank;

t represents the remaining blending time of the batch;

Bc_i,kthe maximum flow of the ith component participating in the mixing line of the kth mixing head is shown;

B_krepresents the flow of the kth blending header batch;

B_{com_max}represents the maximum flow rate of the common component;

r_comkrepresenting the formula of the common component at the kth blending head;

V_i,krepresenting the reserve of the ith component oil of the kth blending head;

r_{i,k_min}，r_{i,k_max}respectively representing the upper limit and the lower limit of the formulation of the ith component oil of the kth blending head;

s_{l,k_min}，s_{l,k_max}respectively representing the upper limit and the lower limit of the k-th blending tank attribute value l;

P_l,ka blending predicted value representing the property l of the kth blending head component oil;

TP_k(P_l,k) Representing the predicted value of crude oil property l in the kth blending tank.

2. The method of claim 1, wherein the property indicator for the whole crude oil calculated using equations (2) and (3) requires a predicted value TP for the density and other properties of the crude oil in each blending tank_k(P_ρ,k)、TP_k(P_l,k)：

Wherein, TVOL_kIndicating the blending target quality of the kth blending tank;

HVOL_kthe bottom oil quality of the kth blending tank is represented;

T_ρ,kthe tank bottom oil density value of the kth blending tank is shown;

T_l,krepresents the value of the tank bottom oil property l of the kth blending tank.

3. The method of claim 1 or 2, wherein a mathematical model of the crude oil blending dual blending head coordination problem is established that takes into account tank bottoms properties, and the coordination optimization problem is solved using a hybrid Differential Evolution (CoDE) optimization method.

4. The method of claim 3, wherein the hybrid differential evolution optimization method comprises the steps of:

(1) initializing a population, wherein the size of the population is NP;

(2) the loop starts, when the algorithm termination condition is not yet met, the following is carried out:

(i) for each individual x in the population_iGenerating strategies by using three test vectors, randomly selecting a group of parameters aiming at each strategy, and generating x_iCorresponding three test vectors u_i1，u_i2And u_i3；

(ii) Evaluation of u_i1，u_i2And u_i3A fitness function value of;

(iii) selection u_i1，u_i2And u_i3The test vector with the optimal function value of the medium fitness degree is marked as u_i；

(iv) Comparison x_iAnd u_iThe fitness function value of (a), if x_iIs superior to u_iThen x_iEntering a next generation population; if u_iIs superior to x_iThen u is_iReplacement of x_iEntering a next generation population;

(3) stopping the algorithm to obtain a final population NP, wherein the individual with the optimal fitness value in the population is the solution of the optimization problem;

the hybrid differential evolution optimization method is implemented by adopting three test vector generation strategies of an equation (4), an equation (5) and an equation (6):

(1) "rand/1/bin" method

(2) "rand/2/bin" method

(3) Current-to-rand/1 method

u_i,G＝x_i,G+rand·(x_r1,G-x_i,G)+F·(x_r2,G-x_r3,G) (6)

Wherein x is_r1,G，x_r2,G，x_r3,G，x_r4,G，x_r5,GIs 5 individuals randomly selected in the G generation, which are different from the current individual x_ri,GAnd are not equal.

Images