CN106227927B - The optimization method of tower type solar heat and power system light and heat collection subsystem - Google Patents
The optimization method of tower type solar heat and power system light and heat collection subsystem Download PDFInfo
- Publication number
- CN106227927B CN106227927B CN201610565843.5A CN201610565843A CN106227927B CN 106227927 B CN106227927 B CN 106227927B CN 201610565843 A CN201610565843 A CN 201610565843A CN 106227927 B CN106227927 B CN 106227927B
- Authority
- CN
- China
- Prior art keywords
- transferred
- formula
- receiver
- feasible zone
- interpolation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Complex Calculations (AREA)
Abstract
The invention discloses a kind of optimization methods of tower type solar heat and power system light and heat collection subsystem, implementation steps are as follows: (1) by heliostat field piecemeal, using receiver surface focus position as decision variable, temperature of molten salt target a height of as far as possible is exported with receiver, being less than threshold value with receiver surface energy distribution standard deviation is constraint, constitution optimization problem；(2) improvement complex method solving optimization problem is utilized；(3) method solving optimization problem is improved and optimizated using based on MBDFO algorithm.In the present invention; the optimization method of tower type solar heat and power system light and heat collection subsystem; under conditions of guaranteeing that receiver outlet temperature of molten salt is high as far as possible; keep the distribution of receiver surface energy more uniform to avoid hot-spot; be conducive to protect receiver, improve heat exchange efficiency, the operation for tower type solar thermo-power station provides reference.
Description
Technical field
The present invention relates to tower-type solar thermal power generating system fields, poly- more particularly to a kind of tower type solar heat and power system
The optimization method of light collection thermal sub-system.
Background technique
Tower-type solar thermal power generating system is the heliostat using the independently tracked sun in many faces, focuses light rays at one
It is fixed on the receiver of top of tower, and is used in the form of thermal energy, steam turbine, generator is driven to generate electricity.Entire tower
Formula solar energy system is divided into 5 optically focused, thermal-arrest, accumulation of heat, supplementary energy and power generation subsystems, has the sun during energy transmission
The conversion of energy-thermal energy-mechanical energy-electric energy.Wherein light and heat collection subsystem is the critical system that solar energy is converted into thermal energy, research
The Optimization Work of light and heat collection subsystem is conducive to improve generating efficiency, there is important guidance to the security operations of receiver
Effect.
Optimization in relation to tower type solar light and heat collection system, existing research is design optimization mostly；Performance optimization is big
Part is the local optimum for certain single subsystem, lacks optimizing research on the basis of total system.
Summary of the invention
The present invention provides a kind of optimization method of tower type solar heat and power system light and heat collection subsystem, the technologies of use
Scheme is as follows:
A kind of optimization method of tower type solar heat and power system light and heat collection subsystem includes the following steps:
1) by heliostat field piecemeal, using receiver surface focus position as decision variable, fused salt temperature is exported with receiver
Degree target a height of as far as possible in nonvolatile situation, being less than threshold value with receiver surface energy distribution standard deviation is constraint, construction
Optimization problem；
2) method solving optimization problem is improved and optimizated using improvement complex method or based on MBDFO algorithm；According to solution
As a result the optimization of tower type solar heat and power system light and heat collection subsystem is realized in the position for adjusting focus point.
Further, the step 1) specifically:
Jing Chang is divided into two regions, using double focus strategy, i.e., a part of heliostat is irradiated to focus point 1, another
Part heliostat is irradiated to focus point 2, and two focus point coordinates are respectively (0,0, z1), (0,0, z2), fused salt is exported with receiver
Temperature target a height of as far as possible in nonvolatile situation, being less than threshold value with receiver surface energy distribution standard deviation is constraint, structure
It makes shown in optimization problem such as formula (1):
Wherein, TvIt indicates the volatilization temperature value of fused salt, is constant, TfssTemperature of molten salt steady-state value, its table are exported for receiver
It is shown as z1, z2Function fitness (z1,z2), z1And z2Indicate two focus point z-axis coordinates, value must not exceed receiver
Vertical height ht, F is the standard deviation of receiver surface energy values, is expressed as independent variable z1, z2Function g (z1,z2), F0For setting
Maximum standard deviation.
Further, when optimization method is to improve complex method, the process of the solving optimization problem specifically:
Optimized model shown in formula (1) is expressed as min f (x), x ∈ Rn, s.t.g (x) >=0, wherein x=[z1；z2],
Indicate the z coordinate of two focus points, objective function f (x) indicates receiver outlet temperature steady-state value Tfss, constraint function g (x) packet
Include Tv-Tfss、F0- F andThree constraints, the algorithm steps of complex method are as follows:
3.1) intial compound form { x is chosen0,x1,…,xn, reflection coefficient α > 1, spreading coefficient γ > 1, constriction coefficient β ∈
(0,1) and precision ε > 0；
3.2) judge intial compound form whether in feasible zone, if not in feasible zone, according to given scaling criterion tune
Whole initial value position, until intial compound form is in feasible zone；If in feasible zone, by n+1 vertex of complex according to target letter
The size of numerical value renumbers, and the number on vertex is made to meet f (x0)≤f(x1)≤…≤f(xn-1)≤f(xn)；
3.3) it enablesIfStop iteration, exports x0, otherwise turn
Enter step 3.4)；
3.4) x is calculatedn+2=xn+1+α(xn+1-xn), check xn+2Whether in feasible zone, i.e., whether meet g (xn+2) >=0,
If reflection coefficient α is decreased up to x not in feasible zonen+2In feasible zone, f (x is calculatedn+2), if f (xn+2) < f (x0), turn
Enter step 3.5), otherwise as f (xn+2) < f (xn-1) when be transferred to step 3.6), as f (xn+2)≥f(xn-1) it is transferred to step 3.7)；
The minishing method of the reflection coefficient α is the extraction of root or is greater than 0 coefficient less than 1 multiplied by one；
3.5) x is calculatedn+3=xn+1+γ(xn+2-xn+1), check xn+3Whether in feasible zone, if not in feasible zone, it will
Spreading coefficient γ decreases up to xn+3In feasible zone, if f (xn+3) < f (x0), enable xn=xn+3, it is transferred to step 3.2), is otherwise turned
Enter step 3.6)；The spreading coefficient γ minishing method is the extraction of root or is greater than 0 coefficient less than 1 multiplied by one；
3.6) enable xn=xn+2, it is transferred to step 3.2)；
3.7) x is enabledn={ xi|f(xi)=min (f (xn),f(xn+2)), calculate xn+4=xn+1+β(xn-xn+1), check xn+4
Whether in feasible zone, if constriction coefficient β is decreased up to x not in feasible zonen+4In feasible zone, if f (xn+4) < f
(xn), enable xn=xn+4, it is transferred to step 3.2), is otherwise transferred to step 3.8)；The minishing method of the constriction coefficient β be the extraction of root or
Person is greater than 0 coefficient less than 1 multiplied by one；
3.8) x is enabledj=x0+θ(xj-x0), j=0,1 ..., n are transferred to step 3.2).
Further, when optimization method is the solving optimization method based on MBDFO algorithm when improving and optimizating method
Principle are as follows:
Assuming that current iteration position xkThere are a series of sample point Y={ y1,y2,…,yq, wherein
There is no than x in sample YkSmall point is established shown in secondary model such as formula (2):
Wherein, since the gradient information of model is unavailable,WithIt is invalid.So by inserting
It is worth conditional definition scalar c, vector g ∈ Rn, symmetrical matrix G ∈ Rn×n, interpolation condition defines method and discusses below.
mk(yl)=f (yl), l=1,2 ..., q. (3)
Include in model shown in formula (2)A coefficient (i.e. the sum of coefficient of c, g and G, it is contemplated that G's
Symmetry), interpolation condition shown in formula (3) defines model mkUniqueness, as long as q meets formula (4).
In this way, formula (3) can be converted into a quadratic linear equation group.If selecting interpolation point y1,y2,…,yq, then
This linear system is nonsingular, model mkIt is unique.
Model mkOnce establishing, then by approximate solution Trust-region subproblem come material calculation p value.
If xk+ p has sufficiently reappeared objective function, then next iteration is defined as xk+1=xk+ p, and update trusted zones
Radius, new iteration start；Otherwise, the step-length is rejected, interpolation domain Y is improved or reduces trusted zones.
In order to reduce algorithm complexity, every iteration once updates a model mk, rather than recalculate from the beginning.Choosing
The curve that a convenient and simple method fits quadratic polynomial is selected, Lagrange or newton multinomial are generally selected.These
Whether the characteristics of basic method is exactly that may serve to measure interpolation collection Y suitable, and can be improved when needed slotting
Value collection.
In Trust Region Algorithm, whether suitable and Trust Region Radius the more new strategy of step-length is all based on objective function reality
The ratio between border difference and the difference of prediction model, that is,
Wherein, xk +Indicate test point.
If ρ >=η meets, then it represents that obtain the abundant reduction of objective function, this is simplest situation.In such case
Under, test point xk +Total energy deletes an element in Y, makes x as new iteration pointk +Included in Y.
, then may be as caused by two reasons when ρ >=η is unsatisfactory for, one is that interpolated sample collection Y sample is not abundant, separately
One is that trusted zones are too big.First reason typically occur in iteration be limited in one do not include feasible zone lower dimensional spaceIn.The algorithm can converge to this minimal subset.The above situation can be defined linear by monitoring formula (3)
The interpolation condition of system detects.If conditional number is too high, change Y set, usually with a new element replacement one
Geju City element is to keep interplotation system (3) as nonsingular as possible.If the interpolation condition of Y meets condition, can simply subtract
Small Trust Region Radius.
Polynomial interopolation theory is described below and updates the method for interpolation collection:
(1) polynomial interopolation
It is discussed in detail first and how using interpolation method to establish target function model.A linear model is considered, such as formula (7) institute
Show:
mk(xk+ p)=f (xk)+gTp (7)
In order to determine vector matrixUtilize interpolation condition mk(yl)=f (yl), l=1,2 ..., n can be write as
Shown in formula (8):
(sl)TG=f (yl)-f(xk), l=1,2 ..., n (8)
Wherein:
sl=yl-xk, l=1,2 ..., n (9)
Conditional (7) represents a system of linear equations, and the row of coefficient matrix is by vector (sl)TIt provides.It can be seen that model
(7) it is uniquely determined by formula (8), and if only if interpolation point { y1,y2,…,yqIt is included in set { sl: l=1,2 ..., n in and
It is linear independence.If this condition is set up, by point xk,y1,y2,...,ynThe simplex of composition is nondegenerate
(nondegenerate).
The secondary model for considering now how one such as formula (2) form of building, as f=f (xk) when.Rewrite model such as formula
(10) shown in:
Wherein, the element in g and G is the unknown vector of q-1 dimension, and as shown in formula (11), p is the unknown vector of q-1 dimension, such as
Shown in formula (12):
Model (10) is identical as the form of model (7), unknown vectorCoefficient can be calculated under linear case.
Multiple quadratic function can be indicated with different modes.The advantages of monomial base (10), can be simply by setting
Element in G is 0 structure to obtain Hessian matrix.However, other bases can more easily avoid the singular point of model (3).
Pass throughOne group of base in the n dimensional linear space of expression.Function (3) therefore can be expressed as formula (13) institute
Show:
For some factor alphasi, interpolation collection Y={ y1,y2,…,yqα can be uniquely determined by the definition of formula (14)i, and
Determinant defined in formula (14) is non-zero:
(2) interpolation collection Y is updated
When MBDFO algorithm is iterated, determinant δ (Y) may be close to zero, will lead to the difficult even mistake of numerical solution
It loses.Therefore some algorithms include the mechanism for keeping interpolation point correctly to configure.One of mechanism is described below.Ranks are waited until with it
Formula δ (Y) becomes smaller than a threshold value, not as good as one geometry tune of calling when test point can not sufficiently reduce objective function f
Complete machine system, the target of this geometry Regulation mechanism are that one interpolation point of replacement increases the value of determinant (14).Wherein, it uses
The following property of δ (Y), this is stated according to Lagrangian.
For any y ∈ Y, the multinomial of LagrangianL (, y) is defined, degree up to 2, that is, L (y, y)=
1,AndShown in two-dimentional Lagrangian such as formula (15).
Assuming that the update of set Y is by deleting a point y-Increase a new point y+, to obtain new set Y+.That
It can prove formula (16) (under a standardization appropriate and given specified conditions).
|δ(Y+)|≤|L(y+,y-)||δ(Y)| (16)
MBDFO algorithm can make full use of this inequality to update interpolation collection.
The first situation (ρ >=η), test point x are considered first+Objective function f can sufficiently be reduced.Use x+Update the point in Y
y-。
For (16), the point y of deletion-Definition is as shown in formula (17):
Next, considering second situation (ρ >=η), i.e. objective function f reduction is to be not enough.Determine whether Y needs first
It improves, is determined using following rule.In current iteration point xkFor any yi∈ Y meets | | xk-yi| | when≤Δ it is considered that
Y be it is appropriate, in this case, reduce Trust Region Radius simultaneously start next iteration.
If Y is inappropriate, calling geometry Regulation mechanism.Select a point y-∈ Y is replaced, substitution point y+Can
Increase the value of determinant (14).For any yi∈ Y can replace its potential valueIt is provided for formula (18):
The point y that will be deleted-For
As a preference of the present invention, when optimization method is the solution based on MBDFO algorithm when improving and optimizating method
The process of optimization problem specifically:
Optimized model shown in formula (1) is expressed as min f (x), x ∈ Rn, s.t.g (x) >=0, wherein x=[z1；z2],
Indicate the z coordinate of two focus points, objective function f (x) indicates receiver outlet temperature steady-state value Tfss, constraint function g (x) packet
Include Tv-Tfss、F0- F andThree constraints, improved MBDFO algorithm steps are as follows:
5.1) an initial interpolated sample collection Y={ y is selected1,y2,…,yq, wherein yi∈Rn, i=1,2 ..., q, selection
One point x0, it is allowed to for any one yi∈ Y meets f (x0)≤f(yi), select an initial Trust Region Radius Δ0, one often
It measures η ∈ (0,1), wherein for the Optimized model shown in formula (1), yi=[z1 i；z2 i], q can be calculated according to formula (4)
5.2) when meeting the condition of convergence, i.e. when the continuous n times variation of the temperature value of iteration point position is less than convergence precision, i.e.,
Flag=N, N are setup parameter, stop iteration；Otherwise, integrate Y for objective function interpolation as quadratic function m using interpolationk(x), have
Body form is mk([z1；z2])=A+Bz1+Cz2+Dz1 2+Ez1 2+Fz1z2, wherein A, B, C, D, E, F are the ginseng that interpolation method acquires
Number；Constraint function is fitted to quadratic function gk(x)；It is solved in Trust Region Radius as shown in formula (19) with trust region method
Optimization problem obtains a step-length p, obtains model optimal value xp=xk+p；
5.3) judge xpWhether in feasible zone, i.e. g (xp) whether it is greater than 0, if g (xp) >=0, then according to formula (6) calculating ratio
Rate ρ；If g (xp) < 0, then x is adjusted according to given scaling criterionpUntil g (xp) >=0, according to formula (6) calculating ratio ρ；
If 5.4) meet ρ >=η, by a data y in interpolation collection Y-Replace with xp, the data y that is replaced-MeetWherein L (xp, y) and indicate Lagrange interpolation polynomial, shown in calculation method such as formula (15), more
New Trust Region Radius is Δk+1, meet Δk+1≥Δk, update next iteration point xk+1For xp, it is transferred to next iteration, i.e. step
5.2)；If being unsatisfactory for ρ >=η, judge whether interpolation collection Y needs to update, if | | xk-yi||≤Δk,yi∈ Y illustrates interpolation collection Y
It does not need to update, is transferred to step 5.5)；If being unsatisfactory for | | xk-yi||≤Δk,yi∈ Y illustrates that interpolation collection Y needs to update, is transferred to
Step 5.6)；
5.5) Trust Region Radius Δ is updatedk+1, meet Δk+1< Δk, next iteration point xk+1=xk, it is transferred to and changes next time
Generation, i.e. step 5.2)；
5.6) update interpolation collection Y: found out with Trust Region Algorithm meet formula (18) be inserted into valueThen it obtains to be deleted
PointTrusted zones Δk+1=Δk；Selection makes the smallest element of objective function in YAs next
Iteration pointIt is transferred to step 5.2)；
The optimization method of above-mentioned tower type solar heat and power system light and heat collection subsystem can take into account energy and safety two
It is required that there is guiding significance to the operation of practical tower solar energy power plant.
Detailed description of the invention
Fig. 1 is the optimization method flow chart of tower type solar heat and power system light and heat collection subsystem；
Fig. 2 is divided into two-part mirror field schematic diagram；
Fig. 3 is to improve complex method flow chart；
Fig. 4 is to improve MBDFO algorithm flow chart；
Fig. 5 is the focus point coordinate iterative process for improving complex method；
Fig. 6 is the receiver outlet temperature of molten salt iterative process for improving complex method；
Fig. 7 is the imaging X-Y scheme for improving complex method and solving local best points；
Fig. 8 is the imaging three-dimensional figure for improving complex method and solving local best points；
Fig. 9 is the focus point coordinate iterative process for improving MBDFO algorithm；
Figure 10 is the receiver outlet temperature of molten salt iterative process for improving MBDFO algorithm.
Specific embodiment
The application is described further below with reference to embodiment and Figure of description.
As shown in Figure 1, a kind of optimization method of tower type solar heat and power system light and heat collection subsystem, implementation steps are such as
Under:
(1) by heliostat field piecemeal, using receiver surface focus position as decision variable, fused salt temperature is exported with receiver
Target a height of as far as possible is spent, being less than threshold value with receiver surface energy distribution standard deviation is constraint, constitution optimization problem.
In this embodiment example, TvThe volatilization temperature value for indicating fused salt is 600 DEG C, and receiver is highly 3.1m vertically.
Jing Chang is divided into two regions, as shown in Fig. 2, the heliostat that is, filled circles indicate shines using double focus strategy
It is mapped to focus point 1, the heliostat that open circles indicate is irradiated to focus point 2.Two focus point coordinates are respectively (0,0, z1), (0,
0,z2), temperature of molten salt target a height of as far as possible is exported with receiver, being less than threshold value with receiver surface energy distribution standard deviation is about
Beam, constitution optimization problem are shown below:
min|600+273.15-Tfss|
Wherein, TfssIndicate that receiver exports temperature of molten salt steady-state value, it can be expressed as z1, z2Function fitness
(z1,z2), z1And z2Indicate two focus point z-axis coordinates, value must not exceed the vertical height 3.1m of receiver.F is receiver
The standard deviation of surface energy values can be expressed as independent variable z1, z2Function g (z1,z2), F0For the maximum standard deviation of setting.On
The objective function and constraint condition for stating Optimized model can take into account two requirements of energy and safety.
(2) improvement complex method solving optimization problem is utilized.
Optimized model shown in formula (1) is expressed as min f (x), x ∈ Rn, s.t.g (x) >=0, wherein x=[z1；z2],
Indicate the z coordinate of two focus points, objective function f (x) indicates receiver outlet temperature steady-state value Tfss, constraint function g (x) packet
Include 600-Tfss、F0- F and -3.1≤x≤3.1 3 constraint.The algorithm flow chart of complex method is as shown in figure 3, algorithm steps
It is as follows:
2.1) intial compound form { x is chosen0,x1,…,xn, reflection coefficient α > 1, spreading coefficient γ > 1, constriction coefficient β ∈
(0,1) and precision ε > 0；
2.2) judge intial compound form whether in feasible zone, if not in feasible zone, according to given scaling criterion tune
The size of n+1 vertex of complex according to target functional value is renumberd, the number on vertex is made to meet f by whole initial value position
(x0)≤f(x1)≤…≤f(xn-1)≤f(xn)；
Wherein the scaling criterion is as follows:
1) for current initial value x0=[z1；z2], if F0- F > 1000, then z1=z1+ 0.1, z2=z2-0.1；
If 2) 100 < F0<=1000-F, then z1=z1+ 0.01, z2=z2-0.01；
If 3) 10 < F0<=100-F, then z1=z1+ 0.001, z2=z2-0.001；
If 4) 0 < F0<=10-F, then z1=z1+ 0.0001, z2=z2-0.0001；
If 5) F0<=0-F, then do not scale.
2.3) it enablesIfStop iteration and exports x0, otherwise turn
Enter 2.4)；
2.4) x is calculatedn+2=xn+1+α(xn+1-xn), check xn+2Whether in feasible zone, i.e., whether meet g (xn+2) >=0,
If reflection coefficient α is decreased up to x not in feasible zonen+2In feasible zone.Minishing method is for the extraction of root or multiplied by one
Coefficient greater than 0 less than 1 calculates f (xn+2), if f (xn+2) < f (x0), it is transferred to 2.5), otherwise as f (xn+2) < f (xn-1) when turn
Enter 2.6), as f (xn+2)≥f(xn-1) it is transferred to (7)；
2.5) x is calculatedn+3=xn+1+γ(xn+2-xn+1), check xn+3Whether in feasible zone, if not in feasible zone, it will
Spreading coefficient γ decreases up to xn+3In feasible zone, minishing method is the extraction of root or is greater than 0 coefficient less than 1 multiplied by one.
If f (xn+3) < f (x0), enable xn=xn+3, it is transferred to 2.2), is otherwise transferred to 2.6)；
2.6) x is enabledn=xn+2, it is transferred to 2.2)；
2.7) enable xn={ xi|f(xi)=min (f (xn),f(xn+2)), calculate xn+4=xn+1+β(xn-xn+1), check xn+4
Whether in feasible zone, if constriction coefficient β is decreased up to x not in feasible zonen+4In feasible zone, minishing method is evolution
Method is greater than 0 coefficient less than 1 multiplied by one.If f (xn+4) < f (xn), enable xn=xn+4, it is transferred to 2.2), is otherwise transferred to
2.8)；
2.8) x is enabledj=x0+θ(xj-x0), j=0,2.2) 1 ..., n are transferred to.
(3) method solving optimization problem is improved and optimizated using based on MBDFO algorithm.
Optimized model shown in formula (1) is expressed as min f (x), x ∈ Rn, s.t.g (x) >=0, wherein x=[z1；z2],
Indicate the z coordinate of two focus points, objective function f (x) indicates receiver outlet temperature steady-state value Tfss, constraint function g (x) packet
Include Tv-Tfss、F0- F andThree constraints, improved MBDFO algorithm steps are as follows:
3.1) an initial interpolated sample collection Y={ y is selected1,y2,…,yq, wherein yi∈Rn, i=1,2 ..., q, selection
One point x0, it is allowed to for any one yi∈ Y meets f (x0)≤f(yi), select an initial Trust Region Radius Δ0, one often
It measures η ∈ (0,1), wherein for the Optimized model shown in formula (1), yi=[z1 i；z2 i], q can be calculated according to formula (4)
3.2) when meeting the condition of convergence, i.e. when the continuous n times variation of the temperature value of iteration point position is less than convergence precision, i.e.,
Flag=N, N are setup parameter, stop iteration；Otherwise, integrate Y for objective function interpolation as quadratic function m using interpolationk(x), have
Body form is mk([z1；z2])=A+Bz1+Cz2+Dz1 2+Ez1 2+Fz1z2, wherein A, B, C, D, E, F are the ginseng that interpolation method acquires
Number；Constraint function is fitted to quadratic function gk(x)；It is solved in Trust Region Radius as shown in formula (19) with trust region method
Optimization problem obtains a step-length p, obtains model optimal value xp=xk+p；
3.3) judge xpWhether in feasible zone, i.e. g (xp) whether it is greater than 0, if g (xp) >=0, then according to formula (6) calculating ratio
Rate ρ；If g (xp) < 0, then x is adjusted according to given scaling criterionpUntil g (xp) >=0, according to formula (6) calculating ratio ρ；
Wherein scaling criterion described in step 3.3) is as follows:
1) for current initial value x0=[z1；z2], if F0- F > 1000, then z1=z1+ 0.1, z2=z2-0.1；
If 2) 100 < F0<=1000-F, then z1=z1+ 0.01, z2=z2-0.01；
If 3) 10 < F0<=100-F, then z1=z1+ 0.001, z2=z2-0.001；
If 4) 0 < F0<=10-F, then z1=z1+ 0.0001, z2=z2-0.0001；
If 5) F0<=0-F, then do not scale.
If 3.4) meet ρ >=η, by a data y in interpolation collection Y-Replace with xp, the data y that is replaced-MeetWherein L (xp, y) and indicate Lagrange interpolation polynomial, shown in calculation method such as formula (15), more
New Trust Region Radius is Δk+1, meet Δk+1≥Δk, update next iteration point xk+1For xp, it is transferred to next iteration, i.e. step
3.2)；If being unsatisfactory for ρ >=η, judge whether interpolation collection Y needs to update, if | | xk-yi||≤Δk,yi∈ Y illustrates interpolation collection Y
It does not need to update, is transferred to step 3.5)；If being unsatisfactory for | | xk-yi||≤Δk,yi∈ Y illustrates that interpolation collection Y needs to update, is transferred to
Step 3.6)；
3.5) Trust Region Radius Δ is updatedk+1, meet Δk+1< Δk, next iteration point xk+1=xk, it is transferred to and changes next time
Generation, i.e. step 3.2)；
3.6) update interpolation collection Y: found out with Trust Region Algorithm meet formula (18) be inserted into valueThen it obtains to be deleted
PointTrusted zones Δk+1=Δk；Selection makes the smallest element of objective function in YAs next
Iteration pointIt is transferred to step 3.2)；
Present example is applied to the radial pattern Jing Chang comprising 1909 face heliostats (known to coordinate)；Emulate place warp
Latitude is (34, -116), and the time is morning 10:00 on October 1st, 1997；Receiver center is highly 76.4m from the ground, is received
Device diameter is 5.1m, is highly 6.2m；Heliostat long 7.5m, wide 5m, from the ground height 6m；It is spread at random a little on every face heliostat
10000,100 light of trace in the light cone of each random point reflection, when using improving complex method solving optimization problem, just
Beginning complex takes [0.5,0.6,1.2；- 0.5, -2.8, -3], reflection coefficient α=1.2, spreading coefficient γ=8, constriction coefficient β=
0.3 and precision ε=1e-3, energy variance threshold values setting value F02.0000e+04 is taken, optimum results focus point iterative method such as Fig. 5
Shown, it is as shown in Figure 6 that receiver exports temperature of molten salt iterative process, wherein local best points are [z1；Z2]=[1.1347；-
0.2932], the local optimum of objective function is 5.1687 DEG C, i.e. local optimum temperature value TfssIt is 594.8313 DEG C；Energy mark
Quasi- difference is 1.9999e+04；The optimization time is 5.2h, emulates to obtain receiver surface using the local best points that optimization obtains
Heat distribution is as Figure 7-8.When using improved MBDFO algorithm solving optimization problem, take initial interpolation collection be [0.5,1.5,
2.2,1.4,1.5,2；- 0.5, -2.5, -2, -1.3, -0.5, -1.5], objective function reduces degree threshold value η=0.01, initial to believe
Rely domain radius Δ0=0.1 precision ε=1e-3 takes coefficient 1.01 when expanding Trust Region Radius, take coefficient when reducing Trust Region Radius
0.9, energy variance threshold values setting value F02.0000e+04 is taken, optimum results focus point iterative method as shown in figure 9, receiver goes out
Mouth temperature of molten salt iterative process is as shown in Figure 10, wherein local best points are [z1；Z2]=[1.1461；- 0.2831], target
The local optimum of function is 5.1508, i.e. local optimum temperature value TfssIt is 594.8492 DEG C；Energy scale difference is 1.9999e+
04；Optimizing the time is 44 minutes, emulates to obtain heat distribution and the figure on receiver surface using the local best points that optimization obtains
7-8 is similar.As can be seen that the local best points that two kinds of optimization methods obtain are almost the same, but when the solution of improved MBDFO algorithm
Between greatly reduce.The computer operating system used is Windows7, and emulation platform MATLAB has invoked gPROMS and CUDA
Interface.
Claims (4)
1. a kind of optimization method of tower type solar heat and power system light and heat collection subsystem, it is characterised in that include the following steps:
1) by heliostat field piecemeal, using receiver surface focus position as decision variable, existed with receiver outlet temperature of molten salt
Target a height of as far as possible in nonvolatile situation, being less than threshold value with receiver surface energy distribution standard deviation is constraint, constitution optimization
Problem；
The step 1) specifically: Jing Chang is divided into two regions, using double focus strategy, i.e., a part of heliostat irradiation
To focus point 1, another part heliostat is irradiated to focus point 2, and two focus point coordinates are respectively (0,0, z1), (0,0, z2), with
Receiver exports temperature of molten salt target a height of as far as possible in nonvolatile situation, is less than with receiver surface energy distribution standard deviation
Threshold value is constraint, shown in constitution optimization problem such as formula (1):
Wherein, TvIt indicates the volatilization temperature value of fused salt, is constant, TfssTemperature of molten salt steady-state value is exported for receiver, it is expressed as
z1, z2Function fitness (z1,z2), z1And z2Indicate two focus point z-axis coordinates, it is vertical that value must not exceed receiver
Height ht, F is the standard deviation of receiver surface energy values, is expressed as independent variable z1, z2Function g (z1,z2), F0Most for setting
Big standard deviation；
2) method solving optimization problem is improved and optimizated using improvement complex method or based on MBDFO algorithm；According to solving result
Realize the optimization of tower type solar heat and power system light and heat collection subsystem in the position for adjusting focus point；
When optimization method is the process of the solving optimization problem based on MBDFO algorithm when improving and optimizating method specifically:
Optimized model shown in formula (1) is expressed as minf (x), x ∈ Rn, s.t.g (x) >=0, wherein x=[z1；z2], it indicates
The z coordinate of two focus points, objective function f (x) indicate receiver outlet temperature steady-state value Tfss, constraint function g (x) includes Tv-
Tfss、F0- F andThree constraints, improved MBDFO algorithm steps are as follows:
5.1) an initial interpolated sample collection Y={ y is selected1,y2,…,yq, wherein yi∈Rn, i=1,2 ..., q select one
Point x0, it is allowed to for any one yi∈ Y meets f (x0)≤f(yi), select an initial Trust Region Radius Δ0, a constant η
∈ (0,1), wherein for the Optimized model shown in formula (1), yi=[z1 i；z2 i], q can be calculated according to formula (4)
5.2) when meeting the condition of convergence, i.e. when the continuous n times variation of the temperature value of iteration point position is less than convergence precision, i.e. flag
=N, N are setup parameter, stop iteration；Otherwise, integrate Y for objective function interpolation as quadratic function m using interpolationk(x), specific shape
Formula is mk([z1；z2])=A+Bz1+Cz2+Dz1 2+Ez1 2+Fz1z2, wherein A, B, C, D, E, F are the parameter that interpolation method acquires；It will
Constraint function is fitted to quadratic function gk(x)；
The optimization problem as shown in formula (19) is solved in Trust Region Radius with trust region method and obtains a step-length p, obtains mould
Type optimal value xp=xk+p；
5.3) judge xpWhether in feasible zone, i.e. g (xp) whether it is greater than 0, if g (xp) >=0, then according to formula (6) calculating ratio ρ；
If g (xp) < 0, then x is adjusted according to scaling criterionpUntil g (xp) >=0, according to formula (6) calculating ratio ρ；
If 5.4) meet ρ >=η, by a data y in interpolation collection Y-Replace with xp, the data y that is replaced-MeetWherein L (xp, y) and indicate Lagrange interpolation polynomial, shown in calculation method such as formula (15), more
New Trust Region Radius is Δk+1, meet Δk+1≥Δk, update next iteration point xk+1For xp, it is transferred to next iteration, i.e. step
5.2)；If being unsatisfactory for ρ >=η, judge whether interpolation collection Y needs to update, if | | xk-yi||≤Δk,yi∈ Y illustrates interpolation collection Y
It does not need to update, is transferred to step 5.5)；If being unsatisfactory for | | xk-yi||≤Δk,yi∈ Y illustrates that interpolation collection Y needs to update, is transferred to
Step 5.6)；
5.5) Trust Region Radius Δ is updatedk+1, meet Δk+1< Δk, next iteration point xk+1=xk, it is transferred to next iteration, i.e.,
Step 5.2)；
5.6) update interpolation collection Y: found out with Trust Region Algorithm meet formula (18) be inserted into valueThen point to be deleted is obtainedTrusted zones Δk+1=Δk；Selection makes the smallest element of objective function in YIt changes as next
Dai DianIt is transferred to step 5.2)；
2. a kind of optimization method of tower type solar heat and power system light and heat collection subsystem as described in claim 1, feature
It is when optimization method is improves complex method, the process of the solving optimization problem specifically:
Optimized model shown in formula (1) is expressed as minf (x), x ∈ Rn, s.t.g (x) >=0, wherein x=[z1；z2], it indicates
The z coordinate of two focus points, objective function f (x) indicate receiver outlet temperature steady-state value Tfss, constraint function g (x) includes Tv-
Tfss、F0- F andThree constraints, the algorithm steps of complex method are as follows:
3.1) intial compound form { x is chosen0,x1,…,xn, reflection coefficient α > 1, spreading coefficient γ > 1, constriction coefficient β ∈ (0,
And precision ε > 0 1)；
3.2) intial compound form is judged whether in feasible zone, if adjusting initial value position according to scaling criterion not in feasible zone,
Until intial compound form is in feasible zone；If in feasible zone, by the size weight of n+1 vertex of complex according to target functional value
New number, makes the number on vertex meet f (x0)≤f(x1)≤…≤f(xn-1)≤f(xn)；
3.3) it enablesIfStop iteration, exports x0, otherwise it is transferred to step
It is rapid 3.4)；
3.4) x is calculatedn+2=xn+1+α(xn+1-xn), check xn+2Whether in feasible zone, i.e., whether meet g (xn+2) >=0, if not
In feasible zone, reflection coefficient α is decreased up into xn+2In feasible zone, f (x is calculatedn+2), if f (xn+2) < f (x0), it is transferred to step
It is rapid 3.5), otherwise as f (xn+2) < f (xn-1) when be transferred to step 3.6), as f (xn+2)≥f(xn-1) it is transferred to step 3.7)；It is described
The minishing method of reflection coefficient α is the extraction of root or is greater than 0 coefficient less than 1 multiplied by one；
3.5) xn is calculated+3=xn+1+γ(xn+2-xn+1), check xn+3Whether in feasible zone, if not in feasible zone,
Spreading coefficient γ is decreased up into xn+3In feasible zone, if f (xn+3) < f (x0), enable xn=xn+3, it is transferred to step 3.2),
Otherwise it is transferred to step 3.6)；The spreading coefficient γ minishing method is the extraction of root or is greater than multiplied by one and 0 is less than 1
Number；
3.6) x is enabledn=xn+2, it is transferred to step 3.2)；
3.7) x is enabledn={ xi|f(xi)=min (f (xn),f(xn+2)), calculate xn+4=xn+1+β(xn-xn+1), check xn+4Whether
In feasible zone, if constriction coefficient β is decreased up to x not in feasible zonen+4In feasible zone, if f (xn+4) < f (xn), it enables
xn=xn+4, it is transferred to step 3.2), is otherwise transferred to step 3.8)；The minishing method of the constriction coefficient β be the extraction of root or multiplied by
One is greater than 0 coefficient less than 1；
3.8) x is enabledj=x0+θ(xj-x0), j=0,1 ..., n are transferred to step 3.2).
3. a kind of optimization method of tower type solar heat and power system light and heat collection subsystem as claimed in claim 2, feature
It is the scaling criterion of the step 3.2) are as follows:
1) for current initial value x0=[z1；z2], if F0- F > 1000, then z1=z1+ 0.1, z2=z2-0.1；
If 2) 100 < F0<=1000-F, then z1=z1+ 0.01, z2=z2-0.01；
If 3) 10 < F0<=100-F, then z1=z1+ 0.001, z2=z2-0.001；
If 4) 0 < F0<=10-F, then z1=z1+ 0.0001, z2=z2-0.0001；
If 5) F0<=0-F, then do not scale.
4. a kind of optimization method of tower type solar heat and power system light and heat collection subsystem as described in claim 1, feature
It is that scaling criterion described in the step 5.3) is as follows:
1) for current initial value x0=[z1；z2], if F0- F > 1000, then z1=z1+ 0.1, z2=z2-0.1；
If 2) 100 < F0<=1000-F, then z1=z1+ 0.01, z2=z2-0.01；
If 3) 10 < F0<=100-F, then z1=z1+ 0.001, z2=z2-0.001；
If 4) 0 < F0<=10-F, then z1=z1+ 0.0001, z2=z2-0.0001；
If 5) F0<=0-F, then do not scale.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610565843.5A CN106227927B (en) | 2016-07-15 | 2016-07-15 | The optimization method of tower type solar heat and power system light and heat collection subsystem |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610565843.5A CN106227927B (en) | 2016-07-15 | 2016-07-15 | The optimization method of tower type solar heat and power system light and heat collection subsystem |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106227927A CN106227927A (en) | 2016-12-14 |
| CN106227927B true CN106227927B (en) | 2019-08-16 |
Family
ID=57530881
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610565843.5A Active CN106227927B (en) | 2016-07-15 | 2016-07-15 | The optimization method of tower type solar heat and power system light and heat collection subsystem |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106227927B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107036303B (en) * | 2017-04-21 | 2023-09-22 | 华电电力科学研究院 | Tower type solar energy receiver with protection system |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104408527A (en) * | 2014-11-14 | 2015-03-11 | 浙江大学 | Focusing strategy optimizing method for mirror fields of tower type solar thermoelectric power system |
-
2016
- 2016-07-15 CN CN201610565843.5A patent/CN106227927B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104408527A (en) * | 2014-11-14 | 2015-03-11 | 浙江大学 | Focusing strategy optimizing method for mirror fields of tower type solar thermoelectric power system |
Non-Patent Citations (3)
| Title |
|---|
| Strategies in tower solar power plant optimization;A.Ramos等;《Solar Energy 86 (2012) 》;20120711;第2536-2548页 |
| 塔式太阳能电站定日镜场的聚焦策略研究;黎韦偲等;《可再生能源》;20160430;第34卷(第4期);第475-480页 |
| 塔式太阳能电站接收器的建模及动态仿真;盛玲霞等;《化工学报》;20160331;第67卷(第3期);第736-742页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106227927A (en) | 2016-12-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Guo et al. | The multi-objective capacity optimization of wind-photovoltaic-thermal energy storage hybrid power system with electric heater | |
| He et al. | Multi-objective planning-operation co-optimization of renewable energy system with hybrid energy storages | |
| Li et al. | Optimization of a heliostat field layout using hybrid PSO-GA algorithm | |
| CN106877338B (en) | The alternating current-direct current micro-capacitance sensor uncertain optimization operation method of the intermittent energy source containing high density | |
| He et al. | The many-objective optimal design of renewable energy cogeneration system | |
| CN106383937B (en) | Water cools down photovoltaic-solar-thermal generating system output power and calculates method and system | |
| CN104408527B (en) | Tower type solar heat and power system Jing Chang focusing strategy optimization method | |
| CN105160435B (en) | Tower type solar thermo-power station mirror field focusing strategy optimization method | |
| Chang et al. | Coordinated frequency regulation using solar forecasting based virtual inertia control for islanded microgrids | |
| CN106300423B (en) | Based on the method and device of three-dimensional trapezoidal fuzzy determining photovoltaic generation daily generation | |
| CN110991092A (en) | Complementary energy integrated utilization calculation method and system | |
| CN106227927B (en) | The optimization method of tower type solar heat and power system light and heat collection subsystem | |
| Pacaud et al. | Stochastic optimal control of a domestic microgrid equipped with solar panel and battery | |
| CN109830987A (en) | The active distribution network Probabilistic Stability method of meter and distributed photovoltaic randomness | |
| CN109494816A (en) | The methods of risk assessment and device of the multipotency streaming system of electric-thermal coupling | |
| Wei et al. | Model-free approach to fair solar pv curtailment using reinforcement learning | |
| Li et al. | Impact on traditional hydropower under a multi-energy complementary operation scheme: An illustrative case of a ‘wind–photovoltaic–cascaded hydropower plants’ system | |
| CN107103189B (en) | A method of carrying out the search of reactor critical buckling | |
| CN108667030A (en) | A kind of polynary duty control method based on prediction model | |
| CN110224391B (en) | Method for solving mixed probability-interval optimal power flow | |
| CN107346474A (en) | The three-dimensional trapezoidal blur method that water cooling photovoltaic and photothermal integral system generated energy calculates | |
| CN110460085A (en) | A method of consider wind-powered electricity generation and part throttle characteristics to effect on power system | |
| Nayak et al. | Frequency regulation of hybrid distributed power systems integrated with renewable sources by optimized type-2 fuzzy PID controller | |
| Verma et al. | Big Bang Big Crunch Optimization for Determination of Worst Case Loading Margin | |
| CN113094871A (en) | Wind power area boundary accurate modeling method based on diamond convex hull set theory |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |