AERO-ENGINE SURGE ACTIVE CONTROL SYSTEM BASED ON FUZZY CONTROLLER SWITCHING

Abstract

An aero-engine surge active control system based on fuzzy controller switching is provided. The present invention selects a basic controller with the most appropriate current state for switching control according to the operating state of a compressor based on the principle of fuzzy switching, and can realize large-range, adaptive and performance-optimized surge active control. Controllers designed by the present invention realize large-range surge active control through fuzzy switching, so that the effective operating ranges of the controllers are expanded and the reliability of the controllers is improved. The designed controllers can be applied to the active control of surge caused by various causes, so that the adaptability of the controllers is improved and is closer to the actual operating condition of the engine. Some optimization indexes are added in the design process of the controllers, which can realize optimal control under corresponding optimization objectives.

Claims

1. An aero-engine surge active control system based on fuzzy controller switching, the control system mainly comprising three parts: a basic controller design module, a fuzzy switching module and a control signal fusion module, wherein the design process of each part comprises the following steps: S1 designing N.sub.c basic controllers according to stability requirements in surge active control, wherein the basic controllers are used for generating a basic control signal u.sub.base of a fuzzy switching controller designed by the present invention, and an implementation process is as follows: S1.1 designing N.sub.c basic controllers by a Lyapunov stability theory based modal control method, with N.sub.c not less than 2; in each basic controller, using a compressor average flow coefficient Φ and a disturbance first-order mode A as feedback quantities respectively to determine a relationship between the feedback quantities and a control quantity u.sub.c required by a compressor, i.e.:
u.sub.base,1=k.sub.1(Φ−Φ.sub.0)
u.sub.base,2=k.sub.2A
u.sub.base,N.sub.c=k.sub.N.sub.c(−Φ+Φ.sub.0+A) in the formulas, k.sub.1, k.sub.2, . . . , k.sub.N.sub.c are controller parameters to be determined; u.sub.base,j represents a basic control signal outputted by an jth basic controller, and Φ.sub.0 is the average flow coefficient of the compressor during steady-state operation; S1.2 determining the operating ranges of the N.sub.c basic controllers: when the operating state of the compressor is within the operating ranges of the basic controllers, the basic controllers can ensure the stable operation of the compressor through tip jet, and the operating ranges of the basic controllers are expressed by the size of the disturbance to the compressor; in the design process of the basic controllers, selecting different compressor parameter variables as feedback variables of the basic controllers, and designing the basic controllers operated in the range of small disturbance and large disturbance respectively to obtain the basic controllers with different operating ranges, wherein the compressor parameter variables comprise average flow Φ and first-order mode A; S1.3 sequencing the basic controllers according to the size of the disturbance range that can be used for operation, based on the operating ranges of the N.sub.c basic controllers, that is, with the increase of the disturbance to the compressor, the controllers in the operating ranges are converted in this order, wherein the rank of the ith basic controller is recorded as rank.sub.i, and rank.sub.i is an integer from 1 to N.sub.c; S2 designing the fuzzy switching module: the fuzzy switching module obtains a selection trend c.sub.tre of the basic controllers by fuzzy reasoning according to a state variable x of the compressor, for representing the weight of a basic controller; the design of the fuzzy switching module needs to determine a state variable x of module input, fuzzy division of module input and output, fuzzy rules used in fuzzy reasoning, and a defuzzification method; a specific design process is as follows: S2.1 determining the state variable x which can represent the operating state of the compressor, as the input of the fuzzy switching module; S2.2 conducting fuzzy division for the state variable x of the input of the fuzzy switching module, and obtaining N.sub.a fuzzy sets in each variable division; conducting fuzzy division for the selection trend c.sub.tre of the outputted basic controller to obtain N.sub.b fuzzy sets; determining that the input variable belongs to a membership function ƒ.sub.in,(i,C)(x.sub.i) of each fuzzy set and the output variable belongs to a membership function ƒ.sub.out,B(c.sub.tre) of each fuzzy set; the membership μ.sub.in,(i,C) of the input variable belonging to different fuzzy sets can be calculated in the following mode:
ρ.sub.in,(i,C)=ƒ.sub.in,(i,C)(x.sub.i) wherein x.sub.i is the value of the ith input variable; C is a division fuzzy set; ƒ.sub.in,(i,C) is a membership function that the ith input variable belongs to the fuzzy set C; μ.sub.in,(i,C) is a membership of the calculated ith input variable in the fuzzy set C; the membership μ.sub.out,B of the output variable belonging to different fuzzy sets can be calculated in the following mode:
μ.sub.out,B=ƒ.sub.out,B(c.sub.tre) wherein c.sub.tre is the selection trend of the output; B is a division fuzzy set; ƒ.sub.out,B(c.sub.tre) represents a membership function of output c.sub.tre in an output fuzzy set B; μ.sub.out,B is a membership of the calculated selection trend c.sub.tre in the fuzzy set B; S2.3 establishing a fuzzy rule table of the fuzzy switching module and designing N.sub.rules fuzzy rules; the fuzzy rules can be explained in if-then format, namely:
If x.sub.1∈C.sub.1,x.sub.2∈C.sub.2, . . . ,x.sub.n∈C.sub.n then c.sub.tre∈B wherein C.sub.1 represents a fuzzy set to which the first input variable belongs in the fuzzy rule, C represents a fuzzy set to which the second variable belongs, and so on; B represents a fuzzy set to which output c.sub.tre belongs in the fuzzy rule; in the fuzzy rules, (x.sub.1∈C.sub.1, x.sub.2 ∈C.sub.2, . . . , x.sub.n∈C.sub.n) is a prior condition of the fuzzy rules, and then a prior membership μ.sub.rule,i of the fuzzy rules can be calculated as: ${\mu }_{\mathrm{rule},i}=\underset{k=1}{\overset{N_{\mathrm{ins}}}{.Math.}}{\mu }_{in,\left(k,C_k\right)}$ wherein μ.sub.in,(k,C.sub.k.sub.) is a membership of the input variables calculated in S2.2 in the fuzzy set; N.sub.ins is the number of input variables in the fuzzy rule; μ.sub.rule,i is a prior membership of the calculated ith fuzzy rule; S2.4 conducting defuzzification for the output of the fuzzy switching module so that a result after defuzzification is the calculated controller selection trend c.sub.tre of the fuzzy switching module; conducting defuzzification for output results by a centroid method, with a calculation process as follows: (1) calculating the prior membership μ.sub.rule,i of each fuzzy rule, namely: ${\mu }_{\mathrm{rule},i}=\underset{k=1}{\overset{N_{\mathrm{ins}}}{.Math.}}{\mu }_{in,\left(k,C_k\right)}$ (2) calculating the controller selection trend c.sub.tre by the centroid method: $c_{\mathrm{tre}}=\frac{{\int }_0^{1}c.Math.\left({\Sigma }_{j=1}^{25}{\mu }_{\mathrm{rule},j}.Math.f_{\mathrm{out},B_j}\left(c\right)\right)dc}{{\int }_0^{1}c.Math.\left({\Sigma }_{i=1}^{25}{\mu }_{\mathrm{rule},i}.Math.f_{\mathrm{out},B_i}\left(c\right)\right)dc}$ wherein μ.sub.rule,i and μ.sub.rule,j are the calculated prior memberships of ith and jth fuzzy rules respectively; ƒ.sub.out,B.sub.i and ƒ.sub.out,B.sub.j are membership functions of the output fuzzy set B in ith and jth fuzzy rules respectively; c.sub.tre is the calculated controller selection trend; S3 designing a control signal fusion module; the input of the control signal fusion module comprises the controller selection trend c.sub.tre and a basic control signal u.sub.base generated by the basic controller, and the output is a control signal u.sub.out after fusion; the control signal fusion module calculates the weight w.sub.i of each basic controller according to the input controller selection trend c.sub.tre, and then conducts weighted fusion for the basic control signal u.sub.base according to the calculated weight, to finally obtain an actual control signal u.sub.out of the controller; the design of the control signal fusion module needs to determine the controller fusion weight and the weighted fusion method, comprising the following specific steps: S3.1 designing the controller fusion weight; according to the number N.sub.c of the basic controllers, conducting fuzzy division for the selection trend c.sub.tre to obtain a fuzzy set center, with the following formula for calculation: $c_i=\frac{1}{N_c-1}\left({\mathrm{rank}}_i-1\right),{\mathrm{rank}}_i=1,2,.Math.,N_c$ in the formula, c.sub.i is the ith fuzzy set center; N.sub.c is the number of the basic controllers; rank.sub.i is a range determined by the ith controller in step S1.3, and the value is an integer from 1 to N.sub.c; for example, when N.sub.c=3, fuzzy set centers are c.sub.i=0, c.sub.2=0.5 and c.sub.3=1; the weight w.sub.i corresponding to the ith controller is calculated according to the following formula:
w.sub.i=ƒ.sub.w,i(c.sub.tre) wherein c.sub.tre is the input of the fusion module, i.e., the controller selection trend; ƒ.sub.w,i is a membership function corresponding to the ith controller, and the membership functions can select the form mentioned in step S2.2; w.sub.i is the calculated weight corresponding to the ith controller; S3.2 conducting weighted fusion for the basic control signal to obtain an actual control signal u.sub.out of the basic controller; the actual control signal u.sub.out of the basic controller can be obtained by weighted fusion through the following formula: $u_{out}=\frac{{.Math.}_{j=1}^{N_c}{w}_j.Math.u_{base,j}}{{.Math.}_{i=1}^{N_c}{w}_i}$ wherein u.sub.base,j represents the basic control signal outputted by the jth basic controller; w.sub.i and w.sub.i represent the weights corresponding to the ith and jth basic controllers respectively; N.sub.c is the number of the basic controllers; u.sub.out is the actual output control quantity of the fused controller, i.e., the output control quantity of the fuzzy switching controller designed by the present invention.

2. The aero-engine surge active control system based on fuzzy controller switching according to claim 1, wherein in step S1.1, a determining method of the controller parameter is as follows: conducting linearization based on a traditional compressor Moore-Greitzer model, to obtain the controller parameter in combination with the Lyapunov stability theory.

3. The aero-engine surge active control system based on fuzzy controller switching according to claim 1, wherein in step S2, the compressor state variable x comprises average flow and average pressure rise of the compressor.

4. The aero-engine surge active control system based on fuzzy controller switching according to claim 1, wherein in step S2, the selection trend c.sub.tre of the basic controller is within a range of 0-1.

Description

DESCRIPTION OF DRAWINGS

[0041] FIG. 1 is a design flow chart of an aero-engine surge active control system based on fuzzy controller switching;

[0042] FIG. 2 is a structural schematic diagram of an aero-engine surge active control system based on fuzzy controller switching;

[0043] FIG. 3 is a structural diagram of an aero-engine surge active control system based on fuzzy controller switching in an embodiment of the present invention;

[0044] FIG. 4 is a schematic diagram of fuzzy sets and fuzzy rules of input and output of a fuzzy switching module; wherein Fig.(a) is fuzzy set division of an average flow coefficient Φ, Fig.(b) is fuzzy set division of a first-order modal amplitude A, Fig.(c) is fuzzy set division of output selection trend c.sub.tre of a fuzzy switching module, and Fig. (d) shows fuzzy switching rules used by the fuzzy switching module;

[0045] FIG. 5 is a schematic diagram of fuzzy set division of a control signal fusion module;

[0046] FIG. 6 shows a surge active control process under small disturbance, wherein Fig.(a) shows a change process of a local compressor flow coefficient when a fuzzy switching controller and basic controllers proposed in the present invention implement control in the same operating range; Fig. (b) shows jet flow coefficients generated by the fuzzy switching controller and the basic controllers; Fig.(c) shows a switching signal generated by the fuzzy switching module at this moment;

[0047] FIG. 7 shows a surge active control process under moderate disturbance, wherein the meanings of Fig.(a), Fig.(b) and Fig.(c) are the same as those described in FIG. 6; and

[0048] FIG. 8 shows a surge active control process under large disturbance, wherein the meanings of Fig.(a), Fig.(b) and Fig.(c) are the same as those described in the previous figure.

DETAILED DESCRIPTION

[0049] The present invention is further described below in combination with drawings and embodiments of the present invention.

[0050] An aero-engine surge active control system based on fuzzy controller switching is provided. The control system mainly comprises three parts: a basic controller design module, a fuzzy switching module and a control signal fusion module. The design flow chart of the aero-engine surge active control system based on fuzzy controller switching is shown in FIG. 1.

[0051] FIG. 2 is a structural schematic diagram of the aero-engine surge active control system based on fuzzy controller switching. It can be seen from the figure that the controller mainly comprises three parts: the fuzzy switching module, the control signal fusion module and basic controllers, wherein the fuzzy switching module further comprises three parts: state variable fuzzification, fuzzy rule reasoning and defuzzification. The fuzzy switching controller generates a controller selection trend c.sub.tre through the fuzzy switching module, and the control signal fusion module obtains an actual controller control signal u.sub.out through weighted fusion according to the selection trend c.sub.tre and a basic control signal u.sub.base.

[0052] FIG. 3 is a structural diagram of the aero-engine surge active control system based on fuzzy controller switching in the present embodiment.

[0053] A specific implementation process comprises the following specific steps:

[0054] S1 Designing basic controllers: designing 3 basic controllers in combination with traditional Lyapunov stability theory according to stability requirements in surge active control, specifically as follows:

[0055] S1.1 Designing 3 basic controllers by a Lyapunov stability theory based modal control method, and using a compressor average flow coefficient Φ and a disturbance first-order mode A as feedback quantities respectively to determine a relationship between the feedback quantities and a control quantity u.sub.base,j required by a compressor: designing N.sub.c=3 basic controllers for the basic controllers in FIG. 3 according to design step S1.1; selecting pressure rise ψ.sub.j generated by a compressor jet valve as the control quantity to obtain the control laws of the basic controllers:

u.sub.base,1=k.sub.1(Φ−Φ.sub.0)  Controller 1:

u.sub.base,2=k.sub.2A  Controller 2:

u.sub.base,3=k.sub.3(−Φ+Φ.sub.0+A)  Controller 3:

[0056] wherein Φ is the compressor average flow coefficient; A is a first-order modal amplitude; Φ.sub.0 is the average flow coefficient at a balance point of the compressor; k.sub.1, k.sub.2 and k.sub.3 are controller parameters to be determined. According to the Lyapunov stability theory, it can be determined that in the present embodiment, the values of k.sub.1, k.sub.2 and k.sub.3 are as follows: k.sub.1=−0.1, k.sub.2=0.1 and k.sub.3=0.1.

[0057] A determining method of the controller parameter is as follows: conducting linearization based on a traditional compressor Moore-Greitzer model, to obtain the controller parameter in combination with the Lyapunov stability theory. The embodiment of the present invention listed herein takes the Moore-Greitzer model of the compressor as a controlled object. The Moore-Greitzer model of the compressor is shown below:

[00009]$\frac{dA}{d\xi }=\frac{3\alpha H}{\left(1+m\alpha \right)W}A\left(1-{\left(\frac{\Phi }{W}-1\right)}^2-\frac{1}{4W^2}A\right)$$\frac{d\Phi }{d\xi }=\frac{H}{l_c}\left(-\frac{\Psi -{\Psi }_{C0}}{H}+\frac{3}{2}\left(\frac{\Phi }{W}-1\right)\left(1-\frac{1}{2W^2}A\right)-\frac{1}{2}{\left(\frac{\Phi }{W}-1\right)}^3+1\right)$$\frac{d\Psi }{d\xi }=\frac{1}{4B^2{l}_c}\left(\Phi \left(\xi \right)-{\Phi }_T\left(\xi \right)\right)$

[0058] In the equations, A(ξ) is a first-order modal amplitude, Φ(ξ) is the compressor average flow coefficient, ψ(ξ) is the average pressure rise coefficient of the compressor and Φ.sub.T(ξ) is the average flow coefficient of a downstream valve of the compressor; in the equations, other parameters are inherent parameters of the compressor, and select the following values here:

ψ.sub.C0=0.30,H=0.14,W=0.25,l.sub.C=8.0,α=1/3.5 and m=1.75.

[0059] S1.2 Determining the operating ranges of the 3 basic controllers: when the operating state of the compressor is within the operating ranges of the basic controllers, the basic controllers can ensure the stable operation of the compressor through tip jet, and the operating ranges of the basic controllers can be expressed by the size of the disturbance to the compressor. The performance characteristics of the above three basic controllers are shown in Table 2.

TABLE-US-00001 TABLE 2 Performance Characteristics of Basic Controllers Controllers Performance Characteristics Controller 1 (average Applicable to small disturbance conditions flow coefficient Φ) with small flow coefficient change and slow development, with less jet quantity Controller 2 (first-order Applicable to large disturbance state, with modal amplitude A) larger jet quantity Controller 3 Applicable to moderate disturbance state, (comprehensive with less and smooth fluctuation of controlled feedback (Φ, A)) quantity and moderate jet quantity

[0060] S1.3 Sequencing the basic controllers according to the size of the disturbance range that can be used for operation, based on the operating ranges of the 3 basic controllers, that is, with the increase of the disturbance to the compressor, the controllers in the operating ranges are converted in this order, wherein the rank of the ith basic controller is recorded as rank, and rank.sub.i is an integer from 1 to 3. It can be seen from Table 2 that the 3 basic controllers have respective operating ranges. With the increase of the disturbance to the compressor, the controllers in the operating ranges are converted from the controller 1 to the controller 3, and then converted to the controller 2. Therefore, the range of the basic controllers can be recorded as follows:

rank.sub.1=1  Controller 1:

rank.sub.2=3  Controller 2:

rank.sub.3=2  Controller 3:

S2 Designing the fuzzy switching module.

[0061] FIG. 4 is a schematic diagram of fuzzy sets and fuzzy rules of input and output of the fuzzy switching module, which reflects the design process of the fuzzy switching module, wherein Fig.(a) is fuzzy set division of an average flow coefficient Φ, Fig.(b) is fuzzy set division of a first-order modal amplitude A, Fig.(c) is fuzzy set division of output controller selection trend c.sub.tre of the fuzzy switching module, and Fig.(d) shows fuzzy switching rules used by the fuzzy switching module.

[0062] The fuzzy switching module obtains a selection trend x of the basic controllers by traditional fuzzy reasoning according to a state variable c.sub.tre of the compressor. The state variable x of the compressor comprises but is not limited to the compressor average flow c and average pressure rise ψ. The selection trend c.sub.tre of the basic controllers is a parameter within a range of 0-1, and is used for representing a weight of a basic controller.

[0063] S2.1 Determining the average flow Φ and the first-order mode A which can represent the operating state of the compressor, as the input of the fuzzy switching module.

[0064] S2.2 Conducting fuzzy division for the state variable x inputted by the fuzzy switching module, as shown in FIG. 4(a):

[0065] Fuzzy division is conducted for the average flow coefficient Φ and the first-order modal amplitude A to obtain N.sub.a=5 fuzzy sets. A reminder membership function and a triangular membership function are used in membership functions. The parameters of the selected membership functions are shown in Table 3 and Table 4 respectively (in order to unify the format, the triangular membership function is regarded as a trapezoidal membership function with two endpoints on upper bottom overlapping). Fuzzy set division of the average flow coefficient Φ and the first-order modal amplitude A is shown in Fig. (a) and Fig. (b) respectively.

TABLE-US-00002 TABLE 3 Membership Function Parameters of Fuzzy Sets of Average Flow Coefficients Fuzzy sets a b c d Z 0 0 0 0.15 PZ 0.1 0.25 0.25 0.4 PS 0.25 0.4 0.4 0.43 PM 0.43 0.6 0.6 0.63 PB 0.63 0.8 1 1

TABLE-US-00003 TABLE 4 Membership Function Parameters of Fuzzy Sets of First-Order Modal Amplitude Fuzzy sets a b c d Z 0 0 0 0.15 PZ 0.05 0.18 0.18 0.34 PS 0.2 0.35 0.35 0.55 PM 0.4 0.55 0.55 0.75 PB 0.63 0.75 1 1

[0066] Fuzzy division is conducted for the selection trend c.sub.tre of the output of the fuzzy switching module to obtain N.sub.b=5 fuzzy sets. The triangular membership function is used in the membership function. The parameters of the selected membership function are shown in Table 5. Fuzzy set division of the selection trend c.sub.tre is shown in Fig.(c).

TABLE-US-00004 TABLE 5 Membership Function Parameters of Fuzzy Sets of Selection trend Fuzzy sets a b c S 0 0 0.25 MS 0 0.25 0.5 M 0.25 0.5 0.75 MB 0.5 0.75 1 B 0.75 1 1

[0067] S2.3 Establishing a fuzzy rule table, as shown in FIG. 4(d):

[0068] According to the performance characteristics of the basic controllers, it can be seen that as the disturbance to the compressor is continuously increased, the controllers are gradually transitioned from average flow coefficient feedback to comprehensive feedback, and then gradually transitioned to first-order modal amplitude feedback according to the operating ranges of the basic controllers. Meanwhile, with the gradual increase of the selection trend c.sub.tre, the used basic controllers are gradually transitioned from the average flow coefficient feedback to the comprehensive feedback, and then gradually transitioned to the first-order modal amplitude feedback. Thus, the design of the fuzzy rules can follow the principle that the greater the compressor disturbance is, the greater the selection trend c.sub.tre is, that is, with the continuous increase of the average flow coefficient Φ and the first-order modal amplitude A, the selection trend c.sub.tre is gradually transitioned from the fuzzy set S to the fuzzy set B. The fuzzy rule table corresponding to the above principle is shown in FIG. 4(d).

[0069] S2.4 Conducting defuzzification for the output of the fuzzy switching module to obtain the selection trend c.sub.tre. The process of defuzzification is illustrated by a specific example here.

[0070] The average flow coefficient Φ=0.275 and the first-order mode A=0.387 are taken as an example:

[0071] (1) Calculating the prior membership μ.sub.rule,i of each fuzzy rule

[0072] The following fuzzy rule is taken as an example, i.e.

if Φ∈PZ,A∈PS then c.sub.tre∈MS

[0073] The rule of the prior membership can be calculated as

[00010]$\begin{array}{c}{\mu }_{\mathrm{rule},1}=\underset{k=1}{\overset{2}{.Math.}}{\mu }_{in,\left(k,C_k\right)}={\mu }_{in,\left(1PZ\right)}\times {\mu }_{in,\left(2PS\right)}\\ =f_{in,\left(1PZ\right)}\left(\Phi \right)\times f_{in,\left(2PS\right)}\left(A\right)=0.833\times 0.815\\ =0.679\end{array}$

[0074] (2) Calculating the controller selection trend c.sub.tre by the centroid method

[0075] After the prior membership ρ.sub.rule,i of each fuzzy rule is determined, the controller selection trend c.sub.tre can be calculated according to the defuzzification method by the centroid method in S2.4.

[00011]$c_{\mathrm{tre}}=\frac{{\int }_0^{1}c.Math.\left({\Sigma }_{j=1}^{25}{\mu }_{\mathrm{rule},j}.Math.f_{\mathrm{out},B_j}\left(c\right)\right)dc}{{\int }_0^{1}c.Math.\left({\Sigma }_{i=1}^{25}{\mu }_{\mathrm{rule},i}.Math.f_{\mathrm{out},B_i}\left(c\right)\right)dc}=0.286$

[0076] S3 Designing a control signal fusion module. FIG. 5 is a schematic diagram of fuzzy set division of the control signal fusion module.

[0077] S3.1 Designing a controller fusion weight. In FIG. 5, the fuzzy set P represents the weight of the average flow coefficient feedback, the fuzzy set PA represents the weight of the comprehensive feedback, and the fuzzy set A represents the weight of the first-order modal amplitude feedback, all of which use the triangular membership functions. In the present embodiment, three basic controllers are used, i.e., N.sub.c=3. According to step S3.1, fuzzy set centers can be determined as c.sub.1=0, c.sub.2=0.5 and c.sub.3=1 respectively. Membership function parameters of the fuzzy sets of the control signal fusion module are shown in Table 6.

TABLE-US-00005 TABLE 6 Membership Function Parameters of Fuzzy Sets of Control Signal Fusion Module Fuzzy sets a b c P 0 0 0.3 PA 0.2 0.5 0.8 A 0.7 1 1

[0078] S3.2 Conducting weighted fusion for the basic control signal, and calculating weights corresponding to each basic controller according to the fuzzy membership of the control signal fusion module labeled with 6 in S3.1 to obtain the actual output control quantity of the controller after fusion, i.e., the output control quantity u.sub.out of the fuzzy switching controller designed by the present invention. Calculation results are shown in FIG. 6, FIG. 7 and FIG. 8.

[0079] The process of weighted fusion of the control signal is further illustrated by an example here:

[0080] The selection trend c.sub.tre=0.286 of the controller is calculated in step S2.4 of the detailed description. According to the method in S3.2, the weights corresponding to the basic controllers can be obtained as follows:

w.sub.1=ƒ.sub.w,1(c.sub.tre)=0.0457

w.sub.2=ƒ.sub.w,2(c.sub.tre)=0

w.sub.3=ƒ.sub.w,3(c.sub.tre)=0.288

[0081] Then, the outputs of the basic controllers are respectively

u.sub.base,1=k.sub.1(Φ−Φ.sub.0)=0.737

u.sub.base,2=k.sub.2A=1.285

u.sub.base,3=k.sub.3(−Φ+Φ.sub.0+A)=1.369

[0082] The controller output after fusion is

[00012]$u_{out}=\frac{{.Math.}_{j=1}^3{w}_j.Math.u_{base,j}}{{.Math.}_{i=1}^3{w}_i}=\frac{0.0457\times 0.737+0\times 1.285+0.288\times 1.369}{0.0457+0+0.288}=1.283$

[0083] The simulation calculation results of this implementation case are shown in FIG. 6, FIG. 7 and FIG. 8: FIG. 6 shows a surge active control process under small disturbance. This disturbance is generated by smaller initial disturbance, wherein Fig. (a) shows a change process of a compressor flow coefficient when a fuzzy switching controller and basic controllers proposed in the present invention implement control; Fig. (b) shows jet flow coefficients generated by the fuzzy switching controller and the basic controllers; Fig.(c) shows a selection trend c.sub.tre generated by the fuzzy switching module at this moment. It can be seen from the figure that under the action of the fuzzy switching controller, the control effect is consistent or even slightly better than that of the basic controllers, and the jet quantity of the jet valve can also be basically consistent. Meanwhile, the selection trend c.sub.tre generated by the fuzzy switching module is basically kept at a low level, that is, the controller which uses the average flow coefficient Φ as the feedback quantity. This is consistent with the performance characteristics and the operating ranges of the basic controllers in the design process.

[0084] FIG. 7 shows a surge active control process under moderate disturbance. This disturbance is generated by larger initial disturbance, wherein the meanings of Fig.(a), Fig.(b) and Fig.(c) are the same as those described in FIG. 6. The fuzzy switching controller can achieve the surge active control effect consistent with the basic controllers under moderate disturbance. However, the jet quantity generated by the jet valve used by the fuzzy switching controller is significantly less than that of the basic controllers, which can reflect that the fuzzy switching controller can achieve a more optimized or even optimal control strategy under the required performance indexes. The selection trend c.sub.tre of the controller indicates that at this moment, the controller is gradually transitioned from comprehensive feedback to average flow coefficient feedback, which also conforms to the performance characteristics and the operating ranges of the basic controllers in the design process.

[0085] FIG. 8 shows a surge active control process under large disturbance. This disturbance is generated by placing distortion plates in front of the compressor, wherein the meanings of Fig.(a), Fig.(b) and Fig.(c) are the same as those described in the previous figure. The distortion plates can significantly change the inlet flow field of the compressor, which has great influence on the stable operation of the compressor. Under this condition, the fuzzy switching controller can achieve the same or even better performance than the basic controllers in the process of surge active control. Meanwhile, the jet quantity generated by the jet valve is also significantly reduced after comparison, so that the control process has less influence on the operating performance of the engine. In this state, the change process of the controller selection trend c.sub.tre among the basic controllers can be clearly observed. The selection trend is decreased gradually with the size of the disturbance to the compressor, which indicates that the controller is gradually transitioned from the first-order modal amplitude feedback to the comprehensive feedback and then to the average flow coefficient feedback at this moment. It can also be seen from the above process that the surge active control method based on fuzzy switching can control the stable operation of the compressor under the conditions of large disturbance and multiple surge causes, and the reliability and the adaptability of the controllers are significantly improved.

[0086] The above embodiments only express the implementation of the present invention, and shall not be interpreted as a limitation to the scope of the patent for the present invention. It should be noted that, for those skilled in the art, several variations and improvements can also be made without departing from the concept of the present invention, all of which belong to the protection scope of the present invention.