CN112861426A - Aeroengine acceleration process optimal control method based on improved particle swarm optimization - Google Patents
Aeroengine acceleration process optimal control method based on improved particle swarm optimization Download PDFInfo
- Publication number
- CN112861426A CN112861426A CN202110046345.0A CN202110046345A CN112861426A CN 112861426 A CN112861426 A CN 112861426A CN 202110046345 A CN202110046345 A CN 202110046345A CN 112861426 A CN112861426 A CN 112861426A
- Authority
- CN
- China
- Prior art keywords
- particle swarm
- optimization
- engine
- optimal control
- algorithm
- 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.)
- Pending
Links
- 239000002245 particle Substances 0.000 title claims abstract description 138
- 238000000034 method Methods 0.000 title claims abstract description 88
- 238000005457 optimization Methods 0.000 title claims abstract description 87
- 230000001133 acceleration Effects 0.000 title claims abstract description 52
- 230000008569 process Effects 0.000 title claims abstract description 49
- 230000006872 improvement Effects 0.000 claims abstract description 11
- 230000006870 function Effects 0.000 claims description 49
- 230000000739 chaotic effect Effects 0.000 claims description 27
- 239000013598 vector Substances 0.000 claims description 10
- 239000000446 fuel Substances 0.000 claims description 9
- 230000003044 adaptive effect Effects 0.000 claims description 7
- 238000002485 combustion reaction Methods 0.000 claims description 7
- 230000000694 effects Effects 0.000 claims description 7
- 238000013178 mathematical model Methods 0.000 claims description 7
- 230000007547 defect Effects 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000001276 controlling effect Effects 0.000 claims description 3
- 230000007423 decrease Effects 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000001105 regulatory effect Effects 0.000 claims description 3
- 230000033228 biological regulation Effects 0.000 claims 1
- XOFYZVNMUHMLCC-ZPOLXVRWSA-N prednisone Chemical compound O=C1C=C[C@]2(C)[C@H]3C(=O)C[C@](C)([C@@](CC4)(O)C(=O)CO)[C@@H]4[C@@H]3CCC2=C1 XOFYZVNMUHMLCC-ZPOLXVRWSA-N 0.000 description 16
- 230000009286 beneficial effect Effects 0.000 description 5
- 238000011160 research Methods 0.000 description 5
- 230000002028 premature Effects 0.000 description 4
- 230000009194 climbing Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000004886 process control Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
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
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/04—Constraint-based CAD
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Feedback Control In General (AREA)
Abstract
The invention provides an aircraft engine acceleration process optimal control method based on an improved particle swarm algorithm. The improvement is mainly carried out from two aspects: improving important parameters in a particle swarm algorithm; the method is combined with a GuoA algorithm, and the method takes advantages and makes up for the shortages and improves in a targeted manner. The improved particle swarm optimization is used for optimizing the acceleration process, and the optimal control variable is output to the aero-engine. The invention can realize the optimal control of the acceleration process of the engine, shorten the acceleration time of the engine, effectively improve the acceleration performance of the engine and improve the maneuverability of the airplane on the premise of ensuring the safe work of the engine.
Description
Technical Field
The invention relates to the technical field of control of aero-engines, in particular to an optimum control method for an accelerating process of an aero-engine based on an improved particle swarm algorithm.
Background
The aircraft engine is the heart of an aircraft and is one of important indexes for measuring the development level of a national aviation industry, so that the research on the reinforced power system has important significance for improving the integral level of the national aviation technology. Because the working process of the aero-engine is complex and changeable, and the aero-engine has the structural characteristics of strong nonlinearity, multiple control variables, time variation and complexity, the research on the engine control problem is more difficult than that of a common control system.
Modern fighters have very high requirements on the maneuverability of the aircraft, and good maneuverability requires good acceleration performance of the engine. The acceleration process control is one of transition state control of the aircraft engine, and compared with engine starting, switch-on/switch-off boosting and deceleration control, the acceleration process control has more obvious influence on the performance of the engine and the aircraft. The acceleration process of the engine directly influences important flight indexes (such as acceleration, climbing, emergent landing and fly-back and the like) of the fighter, so that the research on the optimal control of the acceleration process of the engine and the improvement of the acceleration performance of the engine have important significance.
Although certain results are achieved in the research of optimal control of the acceleration process of the engine at home and abroad, a plurality of unsolved technical problems or points to be improved exist. For example, the particle swarm algorithm often has the defects of premature convergence, poor global convergence performance and the like, and cannot be directly used in optimization control of the acceleration process of the aircraft engine.
Disclosure of Invention
In order to solve the problems in the prior art, the invention provides an optimal control method for an aircraft engine acceleration process based on an improved particle swarm algorithm, which is used for improving the particle swarm algorithm and applying the improved particle swarm algorithm to the optimization control of the engine acceleration process, so that the optimal control of the engine acceleration process is realized, the performance of the engine in the acceleration process is improved, and the maneuverability of an aircraft is improved.
The technical scheme of the invention is as follows:
firstly, a nonlinear mathematical model of the aero-engine is established, and then the optimization of the acceleration process of the aero-turbofan engine is carried out by improving a particle swarm algorithm, so that the optimization of the acceleration process of the aero turbofan engine of a certain type is realized.
The optimal control method for the accelerating process of the aircraft engine based on the improved particle swarm optimization is characterized in that: the improved particle swarm optimization is improved on the basis of the basic particle swarm optimization, and the particle swarm optimization is improved aiming at the defects that the original particle swarm optimization is easy to fall into local optimization in the later iteration stage, the convergence precision is low, the divergence is easy and the like. The improvement is mainly carried out from the following two aspects:
(1) the improvement of important parameters in the particle swarm optimization mainly comprises the following steps: adjustment of the inertial weight ω, learning factor c1And c2Improvement of (1).
(2) The method is combined with a GuoA algorithm, and the method takes advantages and makes up for the shortages and improves in a targeted manner.
The nonlinear mathematical model of the aircraft engine is
y＝f(x)
Wherein
For controlling input vector, including regulating main fuel flow WfArea A of the tail nozzle9Fan guide vane angle dvgl and compressor guide vane angle dvgh,
to output a vector, comprising the specific fuel consumption sfc and the engine thrust F, F (-) is a non-linear vector function that produces the system output.
The acceleration process takes into account the following constraints: the temperature in front of the turbine is not over-heated, the high-pressure compressor is not surged, the high-pressure rotor is not over-rotated, the fan is not over-rotated, the combustion chamber is not rich in oil and is flameout, the oil supply of the main combustion chamber is not more than the maximum oil supply, and the like. The mathematical description of the optimization problem is as follows:
wherein the control variable x ═ Wf,A9,dvgl,dvgh]TThe above variables are all initial values within the corresponding variation range.
And converting the multi-objective function into a single objective function by adopting a linear weighting method to determine the optimizing objective function. Namely, it is
Discretizing and normalizing the formula. The purpose of this processing is to eliminate the influence of the difference of the dimension and the magnitude variation range of each parameter in the objective function on the optimization result. The final optimization objective function can be written as follows:
in the above formula, ωaAnd ωbSatisfy omega for the weight coefficient of the corresponding objective functiona≥0,ωbAnd the size of the optimization target function is more than or equal to 0, and the importance degree of the corresponding optimization target function in the multi-objective optimization problem is reflected.
And discretizing and normalizing the constraint conditions of the aircraft engine according to the form of the objective function:
above gi(x) (i ═ 1, 2.., 11) form a constraint function matrix g (x), and considering the constraint conditions, the objective function can be:
wherein ω is [ ω ]1,ω2,ω3,ω4,ω5,ω6,ω7,ω8,ω9,ω10,ω11]Adjusting the coefficient matrix for the weight of the constraint function, where ω1,ω2,ω3,ω4,ω5,ω6,ω7,ω8,ω9,ω10,ω11The weighting factor can be adjusted for the corresponding constraint conditions, and ω · g (x) is designed to satisfy the constraint conditions of the engine.
The algorithm flow of the improved particle swarm optimization is
(1) The subspace is initialized. Chaotic initialization is carried out to find N particles, fitness function values corresponding to the particles are calculated, the N particles are sequenced according to the function values, the first M particles are taken as initial particles of each population, M initial speeds are randomly generated, and the first S particles are taken to form a subspace:
wherein-0.5 is not less than ai≤1.5，
Let P ═ X1,X2,…,Xs}；i＝1。
(2) Determining an optimal point XbestSum worst point XworstSatisfy the following requirements
Judging a termination condition:
||f(Xbest)-f(Xworst) | | ≦ epsilon (epsilon is convergence accuracy)
Or the number of iterations T < Tmax
If the above convergence condition is satisfied, outputting Xbest,f(Xbest) As an optimum point X*,f(X*) Finishing the calculation, otherwise, turning to 3);
(3) the particle velocity is updated. For each particle, dimension D (1. ltoreq. D. ltoreq.D) updates the speed as follows and is limited to VmaxAnd (4) the following steps.
Vid,t+1＝ωVid,t+c1r1(pid,t-Xid,t)+c2r2(pgd,t-Xid,t)
(4) The particle position is updated. The speed of the update is as follows,
Xid,t+1＝Xid,t+Vid,t+1
(5) a variation operation error threshold Δ E is calculated. The variation operation error threshold Δ E is calculated as follows:
ΔE＝f(Xi,t+1)-f(Xi,t)
if Δ E < 0, go to 6); if Δ E ≧ 0, turn 7)
(6) According to the Logistic chaotic signal generator, generating
u1＝(u11,u12,u13,...,u1D)，
u1j＝4u0j(1-u0j)(j＝1,2,...,D)
And (3) chaotic disturbance:
the chaotic variation generates a sub-population, wherein h represents a step length and gradually decreases along with an evolution algebra t. Beta represents the chaotic disturbance quantity, and is set according to practical problems, wherein the beta is generally 10 e-7-10 e-5, and the effect is better; k represents the size of the subgroup, and k is generally 2 to 6, and the values are compared
Selecting the optimal Xi,best，Xi,t+1＝Xi,best，count[i]＝0；
(7) If Δ E is greater than or equal to 0, count [ i ]]＝count[i]+1, if Δ E ≦ E, E is the last time that the particle was allowed to deteriorate, accepting the new value Xi,t+1＝Xi,t(ii) a Otherwise, update X is not acceptedi,t＝Xi,t+1。
(8) If count [ i ]]G (G is an algebra allowing the particles not to be updated), a new entity X is generated from the subspace VsonAnd making out-of-range judgment
i ═ i +1, single particle completes the update, go 3).
(9) t is t +1, if the adaptive value of the particle is better than the original individual extreme value, the current adaptive value is set as the individual extreme value PiIf P isiIs superior to XworstIs true, then Xworst＝PiTurning to 2).
Further, the control variable is the adjustment of the main fuel flow WfArea A of the tail nozzle9Fan guide vane angle dvgl and compressor guide vane angle dvgh.
Advantageous effects
Compared with the prior art, the optimized control method for the accelerating process of the aircraft engine based on the improved particle swarm optimization improves the particle swarm optimization, can improve the convergence rate of the optimization, reduce the iteration times, ensure the understanding quality and avoid falling into the local optimal solution to a certain extent. The improved particle swarm algorithm is applied to optimizing control in the accelerating process of the engine, the optimal control in the accelerating process of the engine is realized, the accelerating time of the engine is shortened on the premise of ensuring the safe working of the engine, the accelerating performance of the engine is effectively improved, and the maneuverability of the airplane is improved.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a flow chart of an engine acceleration process optimization control of the present invention;
FIG. 2 is a flow chart of the GuoA algorithm of the present invention;
FIG. 3 is a flow chart of the improved particle swarm algorithm of the present invention.
Detailed Description
The invention solves the problem of optimizing control of the acceleration process of the aircraft engine. The optimization problem of the engine is to select an optimal control method to search a group of optimal control quantity (main fuel flow W) in order to optimize the acceleration process of the enginefArea A of the tail nozzle9Fan guide vane angle dvgl, compressor guide vane angle dvgh).
The nonlinear mathematical model of a certain type of aviation turbofan engine is taken as a research object, a corresponding objective function in the acceleration process is established, the optimization algorithm is utilized to carry out optimization calculation on the engine, and the optimal control variable meeting the optimal performance index in the acceleration process can be obtained, so that the acceleration time of the engine is shortened and the acceleration performance of the engine is effectively improved on the premise of ensuring the safe operation of the engine.
On the basis of summarizing previous achievements, the particle swarm optimization is improved according to the characteristics of the aero-engine and is applied to engine optimization control.
1. Because the optimization control of the acceleration process of the aero-engine needs to make a control decision according to the current working state parameters of the engine, when the optimization control method of the acceleration process is researched, the true engine is usually replaced by an aero-engine mathematical model. As the modeling technology of the aeroengine is mature, the detailed description is omitted, and the established nonlinear model of the engine is directly provided
y＝f(x)
Wherein
For controlling input vector, including regulating main fuel flow WfArea A of the tail nozzle9Fan guide vane angle dvgl and compressor guide vane angle dvgh,
for output vectors, including fuel consumption sfc and engine thrust F, F (-) for the production system outputAnd (4) outputting a nonlinear vector function.
2. Design of improved particle swarm algorithm
The shortest response time control mode in the dynamic performance optimization control of the aero-engine means that the acceleration time of the engine is shortened on the premise of ensuring the safe work of the engine. The shortest response time control mode is usually used in the engine acceleration process, and effectively improves the acceleration performance of the engine. The optimizing control flow of the engine acceleration process is shown in figure 1, and the basic idea is as follows: firstly, on the basis of an established turbofan engine nonlinear mathematical model, on the premise of ensuring the safe operation of the engine, the optimization aim of shortening the acceleration time of the engine is taken, then an optimal control plan is sought, and the performance potential of the engine is fully excavated so as to achieve the optimization aim. Because the aeroengine has the characteristics of strong nonlinearity, high complexity and the like, the optimization precision and speed are difficult to be simultaneously improved by using the traditional optimization method, so that the problem must be solved by adopting a more effective optimization algorithm.
The particle swarm algorithm often has the defects of premature convergence, poor global convergence performance and the like, and cannot be directly used in optimization control of the acceleration process of the aircraft engine. Therefore, aiming at the defects that the original particle swarm algorithm is easy to fall into local optimization in the later iteration stage, low in convergence precision, easy to disperse and the like, the particle swarm algorithm is improved, and the improved particle swarm algorithm is used for optimizing control in the accelerating process of the aircraft engine. The improvement is mainly carried out from the following two aspects:
(1) the improvement of important parameters in the particle swarm optimization mainly comprises the following steps: adjustment of the inertial weight ω, learning factor c1And c2Improvement of (1).
(2) The method is combined with a GuoA algorithm, and the method takes advantages and makes up for the shortages and improves in a targeted manner.
Mainly for the inertia weight omega and the learning factor c1And c2The improvement is carried out, the inertia weight omega is the most important parameter in the particle swarm optimization, the proper selection of the inertia weight omega is beneficial to the balance of global and local search performance of the optimization, when the particles are searched towards the optimal solution of the target according to the experience of the whole particles and the self, the selection of the inertia weight omega is very criticalThis may result in a lack of accurate search capability for the particles if selected improperly. Learning factor c1And c2Determines the influence of the self experience and the group experience of the particles on the motion trail of the particles and reflects the strength of information exchange among the particles, thereby reasonably setting c1And c2The method is beneficial to finding the optimal solution as fast as possible for the population.
Improvement of inertial weight:
the standard particle swarm algorithm coordinates the global and local optimizing capability of the particle swarm algorithm through an inertia weight omega. The specific method is that the speed equation of the basic particle swarm is modified as shown in the following formula, and the position equation is kept unchanged:
Vid,t+1＝ωVid,t+c1r1(pid,t-xid,t)+c2r2(pgd,t-xid,t)
in the formula, omega is an inertia weight, the size of the inertia weight determines the inheritance of the particles to the current speed, and a proper omega is selected to help the particle swarm to balance the exploration capacity and the development capacity of the particle swarm. The inertia weight ω mainly balances the effects of local optimization and global optimization, and a larger inertia weight is beneficial to the global optimization, while a smaller inertia weight is beneficial to the local optimization. According to the invention, a linear decreasing strategy from 0.9 to 0.4 is adopted, so that the particle swarm algorithm has good global search performance at the beginning, can be quickly positioned to an area close to a global optimal point, has good local search performance at the later stage, and can accurately obtain a global optimal solution. In the weight calculation formula, TmaxIs the maximum number of iterations, t is the current number of iterations, ωstart，ωendInitial inertial weight and termination weight, respectively.
Learning factor c1And c2The improvement is as follows:
in the formula of the particle swarm optimization, a factor c is learned1And c2Not only reflects the strength of information exchange among particles, but also determines the influence of the self experience and the group experience of the particles on the motion trail of the particles, thereby reasonably setting c1And c2Will be beneficial to the population to be fastAnd finding out the optimal solution. Albeit by weighting the inertia weight omega and learning the factor c1And c2Improvements can be made to some extent to achieve the optimization effect, but some problems still remain to be solved. When the particles are closer to the optimal value, the searching speed is smaller, and the particles lose searching diversity, so that the particles are easy to fall into local optimization, the convergence precision is not high, and the later convergence speed is low. The invention provides an improved algorithm fused with a GuoA algorithm aiming at the defects of the particle swarm algorithm.
Aiming at the problems of premature convergence, easy falling into local optimal solution and the like in the basic particle swarm algorithm, the invention introduces the idea of the GuoA algorithm on the basis of the basic particle swarm algorithm and introduces the chaotic initialization and chaotic variation operation of the swarm into the improved algorithm, and the main idea is as follows: when the particle swarm algorithm cannot find the optimal solution for multiple continuous generations, the particles are searched by adopting a subspace in the GuoA algorithm, when the subspace finds the optimal particles, the original particles are replaced, the updated particles perform chaotic variation again to generate subgroups, and the optimal particles are selected to enter the next generation. Due to the globality of subspace search, the improved algorithm is theoretically a global convergence algorithm.
The chaotic sequence is a pseudo-random sequence determined by a mapping table, a generation rule and an initial condition, and has the characteristics of rich sources, simple generation method and the like. The chaotic initialization and chaotic disturbance of the population are applied to the improved particle swarm algorithm, so that the convergence speed of the algorithm can be improved, the iteration times are reduced, the understanding quality is ensured, and the situation that the population falls into the local optimal solution is avoided to a certain extent. The present invention uses Logistic chaotic mapping to generate the position and velocity of the initial population.
The GuoA algorithm (Guotao algorithm) is an evolution algorithm proposed by Guotao based on combination of multi-father recombination and a group mountain climbing method, is mostly used for solving some numerical optimization problems, and obtains good optimization effect. The core idea of the GuoA algorithm is to randomly generate new individuals by using the subspace spanned by a few individuals, and to make the subspace searched by the algorithm to cover the convex combination space of multiple parents. The GuoA algorithm is integrated into the particle swarm algorithm, the new algorithm is not easy to get early, and the global search capability is realized.
The flow of the GuoA algorithm is shown in fig. 2. The particle swarm optimization is very different from the GuoA optimization.
Particle swarm optimization:
the optimizing process is linear, social and guided;
information is shared among particles, so that the algorithm convergence speed is high;
and thirdly, the solution is easy to fall into a local optimal solution, and premature convergence is easy to cause.
GuoA algorithm:
firstly, the optimization process is random search in a scattered point shape and is not constrained by global optimization;
non-convexity of random search in the subspace ensures that no dead angle exists in the solution space;
and when the optimal solution is not unique, the algorithm may find multiple optimal solutions at one time.
Due to the difference of the optimization modes of the two algorithms, the two algorithms need to solve several problems in the fusion process:
firstly, no subspace is involved in the particle swarm algorithm, and the GuoA algorithm needs to generate the subspace;
generating a searching mode of the particles in the subspace;
the updating mode of the subspace;
and fourthly, switching modes among the searching modes of the algorithm.
In order to solve the above problems, the particle swarm algorithm is improved as follows.
Initializing a subspace:
randomly generating N chaotic sequences, and randomly generating a D-dimensional vector z for each chaotic sequence1＝(z11,z12,z13,...,z1D)，z1jE (0,1), j is 1,2,3, D, and logic is taken as a chaotic signal generator, zi+1j＝μ(1-zij) (j ═ 1,2,3,. times, D; i 1, 2.., N-1) to yield N z1,z2,z3,...,zN. Will ziMaps to the value range of the optimization variable,
xij＝aj+(bj-aj)zij(j＝1,2,...,D；i＝1,2,...,N)
Xi＝(xi1,xi2,xi3,...,xiD)Tn, finding N particles through chaotic initialization, sequencing the N particles according to the sizes of adaptive values, taking the first M initial particles as a population, and randomly generating M initial speeds; taking the first S particles to form a subspace:
wherein-0.5 is not less than ai≤1.5，
Let P ═ X1,X2,…,Xs}。
And (3) subspace searching:
the subspace is searched by using a mutation operation error threshold value delta E and an algebra G allowing the particles not to be updated, and the method specifically comprises the following steps:
(1) according to Δ E ═ f (X)i,t+1)-f(Xi,t) Calculating a variation operation error threshold;
(2) if delta E is less than 0, chaotic variation generates sub-population, and optimal X is selectedi,best，Xi,t+1＝Xi,best，count[i]＝0；
(3) If Δ E is greater than or equal to 0, count [ i ]]＝count[i]+1, if Δ E ≦ E, E is the upper bound that allows the particle to go bad, accept the new value Xi,t+1＝Xi,t(ii) a Otherwise, update X is not acceptedi,t＝Xi,t+1。
(4) If count [ i ]]G, (G is an algebra allowing non-updating of the particles) generates a new individual X in the subspace VsonComparison of f (X)son) And f (X)i,t+1) Selecting the better individual to replace Xi,t+1。
Updating the subspace:
the update of the subspace V is essentially the subspace P ═ X1,X2,…,XSUpdating, in the present algorithm, the historical optimal position P of the individual particle generation is determinediAnd X in PworstMaking a comparison if PiIs superior to XworstThen Xworst＝Pi。
Switching algorithm searching modes:
the switching design of the algorithm is that in the process of updating a single particle, when the particle cannot find better solution updating for continuous generations, the count [ i ] is more than or equal to G, the algorithm searching mode is switched to a subspace to generate the particle, and random searching is carried out. If the particle generated in the subspace is better than the particle, the algorithm switches back to the update mode of the particle group.
The improved particle swarm algorithm flow is shown in FIG. 3:
(1) the subspace is initialized. Chaotic initialization is carried out to find N particles, fitness function values corresponding to the particles are calculated, the N particles are sequenced according to the function values, the first M particles are taken as initial particles of each population, M initial speeds are randomly generated, and the first S particles are taken to form a subspace:
wherein-0.5 is not less than ai≤1.5，
Let P ═ X1,X2,…,Xs}；i＝1。
(2) Determining an optimal point XbestSum worst point XworstSatisfy the following requirements
Judging a termination condition:
||f(Xbest)-f(Xworst) | | ≦ epsilon (epsilon is convergence accuracy)
Or the number of iterations T < Tmax
If the above convergence condition is satisfied, outputting Xbest,f(Xbest) As an optimum point X*,f(X*) Finishing the calculation, otherwise, turning to 3);
(3) the particle velocity is updated. For each particle, dimension D (1. ltoreq. D. ltoreq.D) updates the speed as follows and is limited to VmaxAnd (4) the following steps.
Vid,t+1＝ωVid,t+c1r1(pid,t-Xid,t)+c2r2(pgd,t-Xid,t)
(4) The particle position is updated. The speed of the update is as follows,
Xid,t+1＝Xid,t+Vid,t+1
(5) a variation operation error threshold Δ E is calculated. The variation operation error threshold Δ E is calculated as follows:
ΔE＝f(Xi,t+1)-f(Xi,t)
if Δ E < 0, go to 6); if Δ E ≧ 0, turn 7)
(6) According to the Logistic chaotic signal generator, generating
u1＝(u11,u12,u13,...,u1D)，
u1j＝4u0j(1-u0j)(j＝1,2,...,D)
And (3) chaotic disturbance:
the chaotic variation generates a sub-population, wherein h represents a step length and gradually decreases along with an evolution algebra t. Beta represents the chaotic disturbance quantity, and is set according to practical problems, wherein the beta is generally 10 e-7-10 e-5, and the effect is better; k represents the size of the subgroup, and k is generally 2 to 6, and the values are compared
Selecting the optimal Xi,best，Xi,t+1＝Xi,best，count[i]＝0；
(7) If Δ E is greater than or equal to 0, count [ i ]]＝count[i]+1, if Delta E is less than or equal toe, e is the last time that the particle is allowed to go bad, accepting the new value Xi,t+1＝Xi,t(ii) a Otherwise, update X is not acceptedi,t＝Xi,t+1。
(8) If count [ i ]]G (G is an algebra allowing the particles not to be updated), a new entity X is generated from the subspace VsonAnd making out-of-range judgment
i ═ i +1, single particle completes the update, go 3).
(9) t is t +1, if the adaptive value of the particle is better than the original individual extreme value, the current adaptive value is set as the individual extreme value PiIf P isiIs superior to XworstIs true, then Xworst＝PiTurning to 2).
3. Acceleration process optimization control based on improved particle swarm optimization
On the premise of ensuring the safe operation of the engine, the improved particle swarm algorithm is adopted to carry out optimization control on the acceleration process of a certain turbofan engine, and on the premise of ensuring the safe operation of the engine, the improved particle swarm algorithm can effectively shorten the acceleration time and achieve the purpose of optimization.
The acceleration time of the engine is defined as
In the formula: i is the moment of inertia of the rotor; n ismaxThe rotating speed at the end of the acceleration process; n isidleThe rotation speed when the vehicle is slow; delta NacTo accelerate the remaining power of the turbine.
From the above formula, it can be seen that: the factor for determining the acceleration time is mainly the turbine residual power Δ N during accelerationac. The residual power of the turbine is mainly determined by the high-pressure rotor speed nHAnd a high pressure turbine front total temperature Tt4. To shorten the acceleration time, the residual power of the turbine must be increased, andit is necessary to increase the high-pressure rotor speed of the engine and raise the temperature after the combustion chamber. Therefore, the invention selects the high-pressure rotor speed nHAnd a high pressure turbine front total temperature Tt4As an objective function for optimizing control of the acceleration process. The mathematical expression of the objective function is as follows:
in the above formula, nHdIs the target speed of the high-pressure rotor, nHIs the actual rotational speed of the high pressure rotor. T ist4dTarget total temperature before high pressure turbine, Tt4Is the actual total temperature before the high-pressure turbine.
In order to ensure the stable work of the engine in the acceleration process, the invention considers the following constraint conditions: the temperature in front of the turbine is not over-heated, the high-pressure compressor is not surged, the high-pressure rotor is not over-rotated, the fan is not over-rotated, the combustion chamber is not rich in oil and is flameout, the oil supply of the main combustion chamber is not more than the maximum oil supply, and the like.
Considering the influence of the objective function, constraint conditions and control variables, a suitable set of W needs to be foundf，A9Dvgl, dvgh, minimize the engine acceleration time, i.e. the following nonlinear constraint problem needs to be solved:
wherein the control variable x ═ Wf,A9,dvgl,dvgh]TThe above variables are all initial values within the corresponding variation range.
The acceleration process of the engine is a dynamic process, the optimization result required to be obtained is a track curve of a control variable changing along with time, but the improved particle swarm algorithm is only suitable for a static problem, and an objective function, the control variable and a constraint condition need to be properly processed to solve the dynamic problem. According to the formula, the method adopts a multi-objective optimal control method, and a linear weighting method is adopted to convert a multi-objective function into a single objective function so as to determine the optimal objective function. Namely, it is
Discretizing and normalizing the formula. The purpose of this processing is to eliminate the influence of the difference of the dimension and the magnitude variation range of each parameter in the objective function on the optimization result. The final optimization objective function can be written as follows:
in the above formula, ωaAnd ωbSatisfy omega for the weight coefficient of the corresponding objective functiona≥0,ωbAnd the size of the optimization target function is more than or equal to 0, and the importance degree of the corresponding optimization target function in the multi-objective optimization problem is reflected.
And discretizing and normalizing the constraint conditions of the aircraft engine according to the form of the objective function:
above gi(x) (i ═ 1, 2.., 11) form a constraint function matrix g (x), and considering the constraint conditions, the objective function can be:
wherein ω is [ ω ]1,ω2,ω3,ω4,ω5,ω6,ω7,ω8,ω9,ω10,ω11]Adjusting the coefficient matrix for the weight of the constraint function, where ω1,ω2,ω3,ω4,ω5,ω6,ω7,ω8,ω9,ω10,ω11The weighting factor can be adjusted for the corresponding constraint conditions, and ω · g (x) is designed to satisfy the constraint conditions of the engine.
Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made in the above embodiments by those of ordinary skill in the art without departing from the principle and spirit of the present invention.
Claims (5)
1. An aircraft engine acceleration process optimal control method based on an improved particle swarm algorithm is characterized in that: firstly, establishing a nonlinear mathematical model of an aeroengine;
secondly, determining a corresponding target function and a corresponding constraint function according to the acceleration process of the engine;
thirdly, optimizing and calculating by using an improved particle swarm algorithm;
and fourthly, outputting the optimal control variable to the aircraft engine.
The optimal control method for the accelerating process of the aircraft engine based on the improved particle swarm optimization is characterized in that: the improved particle swarm optimization is improved on the basis of the basic particle swarm optimization, and the particle swarm optimization is improved aiming at the defects that the original particle swarm optimization is easy to fall into local optimization in the later iteration stage, the convergence precision is low, the divergence is easy and the like. The improvement is mainly carried out from the following two aspects:
(1) the improvement of important parameters in the particle swarm optimization,the method mainly comprises the following steps: adjustment of the inertial weight ω, learning factor c1And c2Improvement of (1).
(2) The method is combined with a GuoA algorithm, and the method takes advantages and makes up for the shortages and improves in a targeted manner.
2. The optimal control method for the accelerating process of the aeroengine based on the improved particle swarm optimization algorithm according to claim 1, wherein the optimal control method comprises the following steps: the nonlinear mathematical model of the aircraft engine is
y＝f(x)
Wherein
For controlling input vector, including regulating main fuel flow WfArea A of the tail nozzle9Fan guide vane angle dvgl and compressor guide vane angle dvgh,
to output a vector, comprising the specific fuel consumption sfc and the engine thrust F, F (-) is a non-linear vector function that produces the system output.
3. The optimal control method for the accelerating process of the aeroengine based on the improved particle swarm optimization algorithm according to claim 1, wherein the optimal control method comprises the following steps: the acceleration process takes into account the following constraints: the temperature in front of the turbine is not over-heated, the high-pressure compressor is not surged, the high-pressure rotor is not over-rotated, the fan is not over-rotated, the combustion chamber is not rich in oil and is flameout, the oil supply of the main combustion chamber is not more than the maximum oil supply, and the like. The mathematical description of the optimization problem is as follows:
wherein the control variable x ═ Wf,A9,dvgl,dvgh]TThe above variables are all initial values within the corresponding variation range.
And converting the multi-objective function into a single objective function by adopting a linear weighting method to determine the optimizing objective function. Namely, it is
Discretizing and normalizing the formula. The purpose of this processing is to eliminate the influence of the difference of the dimension and the magnitude variation range of each parameter in the objective function on the optimization result. The final optimization objective function can be written as follows:
in the above formula, ωaAnd ωbSatisfy omega for the weight coefficient of the corresponding objective functiona≥0,ωbAnd the size of the optimization target function is more than or equal to 0, and the importance degree of the corresponding optimization target function in the multi-objective optimization problem is reflected.
And discretizing and normalizing the constraint conditions of the aircraft engine according to the form of the objective function:
above gi(x) (i ═ 1, 2.., 11) form a constraint function matrix g (x), taking into account the constraint stripsAfter completion, the objective function can be:
wherein ω is [ ω ]1,ω2,ω3,ω4,ω5,ω6,ω7,ω8,ω9,ω10,ω11]Adjusting the coefficient matrix for the weight of the constraint function, where ω1,ω2,ω3,ω4,ω5,ω6,ω7,ω8,ω9,ω10,ω11The weighting factor can be adjusted for the corresponding constraint conditions, and ω · g (x) is designed to satisfy the constraint conditions of the engine.
4. The optimal control method for the accelerating process of the aeroengine based on the improved particle swarm optimization algorithm according to claim 1, wherein the optimal control method comprises the following steps: the algorithm flow of the improved particle swarm optimization is
(1) The subspace is initialized. Chaotic initialization is carried out to find N particles, fitness function values corresponding to the particles are calculated, the N particles are sequenced according to the function values, the first M particles are taken as initial particles of each population, M initial speeds are randomly generated, and the first S particles are taken to form a subspace:
wherein-0.5 is not less than ai≤1.5，
Let P ═ X1,X2,…,Xs}；i＝1。
(2) Determining an optimal point XbestSum worst point XworstSatisfy the following requirements
Judging a termination condition:
||f(Xbest)-f(Xworst) | | ≦ epsilon (epsilon is convergence accuracy)
Or the number of iterations T < Tmax
If the above convergence condition is satisfied, outputting Xbest,f(Xbest) As an optimum point X*,f(X*) Finishing the calculation, otherwise, turning to 3);
(3) the particle velocity is updated. For each particle, dimension D (1. ltoreq. D. ltoreq.D) updates the speed as follows and is limited to VmaxAnd (4) the following steps.
Vid,t+1＝ωVid,t+c1r1(pid,t-Xid,t)+c2r2(pgd,t-Xid,t)
(4) The particle position is updated. The speed of the update is as follows,
Xid,t+1＝Xid,t+Vid,t+1
(5) a variation operation error threshold Δ E is calculated. The variation operation error threshold Δ E is calculated as follows:
ΔE＝f(Xi,t+1)-f(Xi,t)
if Δ E < 0, go to 6); if Δ E ≧ 0, turn 7)
(6) According to the Logistic chaotic signal generator, generating
u1＝(u11,u12,u13,...,u1D)，
u1j＝4u0j(1-u0j) (j＝1,2,...,D)
And (3) chaotic disturbance:
the chaotic variation generates a sub-population, wherein h represents a step length and gradually decreases along with an evolution algebra t. Beta represents the chaotic disturbance quantity, and is set according to practical problems, wherein the beta is generally 10 e-7-10 e-5, and the effect is better; k represents the size of the subgroup, and k is generally 2 to 6, and the values are compared
Selecting the optimal Xi,best，Xi,t+1＝Xi,best，count[i]＝0；
(7) If Δ E is greater than or equal to 0, count [ i ]]＝count[i]+1, if Δ E ≦ E, E is the last time that the particle was allowed to deteriorate, accepting the new value Xi,t+1＝Xi,t(ii) a Otherwise, update X is not acceptedi,t＝Xi,t+1。
(8) If count [ i ]]G (G is an algebra allowing the particles not to be updated), a new entity X is generated from the subspace VsonAnd making out-of-range judgment
i ═ i +1, single particle completes the update, go 3).
(9) t is t +1, if the adaptive value of the particle is better than the original individual extreme value, the current adaptive value is set as the individual extreme value PiIf P isiIs superior to XworstIs true, then Xworst＝PiTurning to 2).
5. The optimal control method for the accelerating process of the aeroengine based on the improved particle swarm optimization algorithm according to claim 1, wherein the optimal control method comprises the following steps: the control variable being the regulation of the main fuel flow WfArea A of the tail nozzle9Fan guide vane angle dvgl and compressor guide vane angle dvgh.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110046345.0A CN112861426A (en) | 2021-01-14 | 2021-01-14 | Aeroengine acceleration process optimal control method based on improved particle swarm optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110046345.0A CN112861426A (en) | 2021-01-14 | 2021-01-14 | Aeroengine acceleration process optimal control method based on improved particle swarm optimization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112861426A true CN112861426A (en) | 2021-05-28 |
Family
ID=76003689
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110046345.0A Pending CN112861426A (en) | 2021-01-14 | 2021-01-14 | Aeroengine acceleration process optimal control method based on improved particle swarm optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112861426A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113361777A (en) * | 2021-06-08 | 2021-09-07 | 淮阴工学院 | Runoff prediction method and system based on VMD decomposition and IHHO optimization LSTM |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108133258A (en) * | 2016-12-01 | 2018-06-08 | 南京理工大学 | A kind of mixing global optimization method |
| WO2020000248A1 (en) * | 2018-06-27 | 2020-01-02 | 大连理工大学 | Space reconstruction based method for predicting key performance parameters of transition state acceleration process of aircraft engine |
-
2021
- 2021-01-14 CN CN202110046345.0A patent/CN112861426A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108133258A (en) * | 2016-12-01 | 2018-06-08 | 南京理工大学 | A kind of mixing global optimization method |
| WO2020000248A1 (en) * | 2018-06-27 | 2020-01-02 | 大连理工大学 | Space reconstruction based method for predicting key performance parameters of transition state acceleration process of aircraft engine |
Non-Patent Citations (3)
| Title |
|---|
| LINFENG GOU: "Predictive Control of Turbofan Engine Model Based on Improved Elman Neural Network", 《IEEE》 * |
| 梁爱侠等: "基于SA-EFA的航空发动机结构化控制器参数优化", 《中国航天第三专业信息网第三十九届技术交流会暨第三届空天动力联合会议论文集》 * |
| 滕志军等: "基于动态加速因子的粒子群优化算法研究", 《微电子学与计算机》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113361777A (en) * | 2021-06-08 | 2021-09-07 | 淮阴工学院 | Runoff prediction method and system based on VMD decomposition and IHHO optimization LSTM |
| CN113361777B (en) * | 2021-06-08 | 2023-11-03 | 淮阴工学院 | Runoff prediction method and system based on VMD decomposition and IHHO optimization LSTM |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Vicini et al. | Airfoil and wing design through hybrid optimization strategies | |
| CN109345019B (en) | Improved particle swarm algorithm-based micro-grid economic dispatching optimization strategy | |
| WO2022061871A1 (en) | Hybrid-adaptive differential evolution-based iterative algorithm for aeroengine model | |
| CN111594322B (en) | Variable-cycle aero-engine thrust control method based on Q-Learning | |
| CN108167802B (en) | Multi-model intelligent optimizing and predicting control method for boiler load under low load | |
| CN110363343A (en) | A kind of the GROUP OF HYDROPOWER STATIONS intelligent optimization dispatching method and system of mixed self-adapting | |
| CN111079290A (en) | Engine model correction method based on improved multivariate cosmic algorithm | |
| CN112131670B (en) | Aero-engine model iterative algorithm based on hybrid adaptive differential evolution | |
| Lei et al. | Online optimization of fuzzy controller for coke-oven combustion process based on dynamic just-in-time learning | |
| CN112861426A (en) | Aeroengine acceleration process optimal control method based on improved particle swarm optimization | |
| Fang et al. | Deep reinforcement learning method for turbofan engine acceleration optimization problem within full flight envelope | |
| CN110399697A (en) | Control distribution method based on the aircraft for improving genetic learning particle swarm algorithm | |
| CN112836427A (en) | Aeroengine minimum oil consumption control optimization method based on improved genetic algorithm | |
| CN112949159A (en) | IPSO-based optimal control method for acceleration process of variable cycle engine | |
| CN115217635A (en) | Turbofan engine full-envelope self-adaptive acceleration control method | |
| CN115358167A (en) | Flying and launching integrated pneumatic accompanying optimization design method considering engine parameters | |
| CN114815616A (en) | Intelligent regulation and control method and system for mode conversion of turbine stamping combined type engine | |
| CN113189871B (en) | Thermal power generating unit flexible operation-oriented coordination control strategy | |
| CN112949160A (en) | IPSO-based engine acceleration process optimal control method under gas path component failure | |
| CN112943453A (en) | IGA-based engine maximum thrust control optimization method under gas circuit component failure | |
| CN112943479A (en) | Aero-engine acceleration process optimal control method based on improved simplex method | |
| Gong et al. | A path planning method based on improved particle swarm optimization algorithm | |
| CN113027614A (en) | Aero-engine maximum thrust control optimization method based on improved simulated annealing algorithm | |
| CN115544779B (en) | Aeroengine maximum thrust control method based on dual-mode linear spiral attraction method | |
| Zhang et al. | Optimization of adaptive cycle engine performance based on improved particle swarm optimization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20210528 |
|
| WD01 | Invention patent application deemed withdrawn after publication |