In order to obtain high accuracy design result of site reference high temperature gas sensor, starting from the model and solution of temperatures of support housing, outer shield, inner shield and thermocouple wire, the computing method of optimum internal flow velocity of the sensor was studied through predetermination and iteration of key parameters. Difference quotient deviation judgment method was put forward, which solved the problem of insufficient boundary conditions during calculation. The result showed that optimum internal flow velocities of inner shield and annular channel of double shield L type site reference high temperature gas sensor were 146.1 m/s and 544.9 m/s, respectively, under the working condition of 1717.5 K total temperature and 0.5 MPa static pressure. The study results can provide necessary support for optimal design, theoretical analysis of high temperature gas sensors of double shield type, and single shield type, etc.

The effects of swirl numbers on the combustion performance of lean premixed combustor were studied by means of numerical simulation and experiment. The numerical simulation results showed that when the swirl number was between 0.56−0.73, an obvious recirculation zone can be generated at the head of the flame barrel. With the increase of swirl number, the scale of recirculation zone increased, but the fuel mixing uniformity became worse, and the non-uniformity of the main combustion zone decreased from 0.00853 to 0.01047, which was not conducive to the improvement of the uniformity of temperature field and the reduction of NOx emission. The experiment results showed that with the increase of swirl number, the lean blowout fuel-air ratio decreased, the smaller inlet Mach number indicated the more obvious trend, the outlet temperature distribution factor increased from 0.1640 to 0.1915, the uniformity became worse, the emission index of CO decreased from 26.23 g/kg to 18.07 g/kg, the emission index of UHC decreased from 12.55 g/kg to 9.21 g/kg, and the emission index of NOx increased from 0.609 g/kg to 0.850 g/kg.

Experiments were conducted to study the ignition performance on a full annular combustor by simulating the ground start and high-altitude ignition parameters. The test results showed that the fuel-air ratio of both the stable flame establishment and the flame propagation from the ignition dome to the unburned dome decreased with the increasing inlet airflow Mach number; ignition became more and more difficult with the increase of simulating ignition height, especially at low inlet Mach number condition; the fuel-air ratio of stable flame establishment limit and the light around limit decreased with the increase of corrected reference velocity, and there was a power function relationship between them; the functional relationship was obtained by fitting; if compared with the test values, the errors of the estimated fuel-air ratio of stable flame establishment limit and light around limit were mostly less than 10% and 15%, respectively.

To investigate the effect of the cavity combustor on the propagation modes of rotating detonation waves and propulsion performance, experimental study was performed in a laboratory-scale rotating detonation combustor utilizing different combustor configurations including a cavity combustor, two annular combustors with the combustor width of 19 mm and 15 mm, respectively. Ethylene and oxygen-enriched air was used as propellants. The mass flow rate of oxidizer varied from 50 g/s to 200 g/s and the equivalence ratio was fixed at 0.8. The experiments have been conducted in the combustor with and without installing the aerospike nozzles, respectively. For the combustor without installing the nozzle, the dual-wave collision mode and the single detonation mode were obtained and the mass flow rate ranges of different modes were identical for different combustors. For the cavity combustor, the propagation velocity was faster than the values measured from the annular combustors, especially at lower mass flow rate. This phenomenon indicated that more homogeneous mixtures were produced and lower velocity deficit were obtained in the cavity combustor. As the combustors installing with the aerospike nozzles, different propagation modes were observed, including deflagration mode, dual-wave collision, quad-wave collision, single detonation mode and dual wave mode. The stable co-rotating detonations were easily obtained with a wide range of mass flow rate in the cavity combustor, while the detonation counter-rotating detonations were easily acquired at the same operation conditions. Finally, it was found that specific impulse calculated from the cavity combustor decreased by 10% compared with the annular combustor of width 15 mm, and by 7% compared with the annular one of width 19 mm, respectively. In summary, the cavity combustor was beneficial for promoting the stability and reducing the velocity deficit of rotating detonations, but this configuration was adverse for the propulsion performance.

The trajectory and penetration of liquid jet in crossflow with non-uniform velocity distribution were studied experimentally. Using specifically designed perforated plates to realize several non-uniform velocity distributions, water was used as the jet liquid, and the research was carried out by laser sheet combined with phase doppler particle Analyzer. The main concerns involved the average jet momentum flux ratio, the local Weber number and non-uniformity of the transverse velocity distribution. The influence of the crossflow with non-uniform velocity distribution on the liquid jet was mainly reflected in the transformation of the breakup mode along the liquid flow direction. A dimensionless parameter L characterizing the influence range of liquid jet by cross flow was proposed, and an empirical formula suitable for jet trajectory with non-uniformity of −2—2 and jet momentum flux ratio of 10—40 at normal temperature and pressure was constructed. The existence of “primary uplift section” was taken as the criterion to determine whether the subsection fitting was carried out. For non-uniformity less than 1, the jet trajectory was predicted by combining the empirical formula with the linear function diagram of nonuniformity and L. For non-uniformity greater than 1, the location where the liquid jet first developed into composite breakup was taken as the dividing point of “uplift section” and “deflection section”, and the jet trajectory was fitted by segments.

In order to solve the problem that many factors may lead to the increase of calculation cost of optimizing air film cooling efficiency, a coupling model of flow angle, length-diameter ratio and coating thickness at three blowing ratios (0.5, 1.0 and 1.5) was reasonably designed by Box-Behnken design method. Realizable k-ε turbulence model was used for numerical simulation. Response Surface Methodology (RSM) was used to analyze the simulation results to obtain the response equation, finally the optimal parameters were predicted by the regression equation. Researches showed that flow angle and coating thickness are the main factors affecting the cooling efficiency of film holes with thermal barrier coating (TBC), and the length-diameter is the secondary factor affecting the cooling efficiency. The response equation was used to predict the optimization model with best film cooling efficiency. Within the studied blowing ratio range, the results showed that the film cooling efficiency of the optimized model was 55.45%−90.95% higher than that of the reference, and the prediction error range of the response equation was 2.71%−13.42% with higher accuracy.

The effects of hole arrangement of two rows of film holes on the film cooling performance at each side of rotating blade leading edge were investigated. The film cooling performance and flow filed were obtained by using numerical simulation method. The first row of film holes was fixed at ±10° downstream stagnation line and was followed by the second row of holes. The span-wise position and the stream-wise location of the second row of holes were adjusted to achieve different hole arrangement. The results indicated that the effects of hole arrangement of two rows on film cooling were different at two sides of the leading edge. At the leading edge suction side, the relative position between two rows of holes significantly affected the interaction between the adjacent jets from two rows. At the pressure side, closer distance between two adjacent holes in two rows produced better film cooling performance. Meanwhile, the effect of relative position between two rows of holes was mainly located in the near hole region. The film cooling effectiveness at the suction side and pressure side increased by 0.07 and 0.02, respectively.

A full-scale large bypass ratio turbofan engine core nacelle ventilation and heat transfer test system was designed by changing the cooling air flow rate, the heat generation on the core casing surface and the core nacelle shell insulation to study the convective heat transfer characteristics in the cabin. The results showed that with the increase of intake air flow rate, the surface heat transfer coefficient of the core casing of each section increased accordingly. Due to the presence of high-level flanges in the front/rear compartments, the increase in the front compartment intake air flow rate had almost no influence on the convective heat transfer on the surface of the rear compartment casing. When the intake air flow rate reached 0.05 kg/s, the improvement of the cabin space temperature was limited. Spatially, the air flow temperature of the upper part of the front cabin was about 10 K higher than that of the lower part, and the rear cabin was about 20 K higher. At the same intake air flow rate, the heat output of the casing surface increased, and the surface heat transfer coefficient of the casing increased slightly, with a difference of about 10 W/（m²·K）. The existence of the insulation layer caused the heat transfer of the core casing to the environment through radiation heat exchange to decrease, and the heat transfer method mainly relied on convection heat transfer, so the surface heat transfer coefficient was relatively improved, with the difference about 60 W/（m²·K）. The empirical formulae of surface heat transfer for each section of the core casing were obtained by using the least squares method, which can be used as a reference for the design of ventilation and cooling engineering for the core nacelle with large bypass ratio turbofan.

In order to establish an aero-engine inlet total pressure distortion estimation method for the application of engine distortion tolerance control, using the measured steady and dynamic pressure data of inlet distortion flow field simulating test, a study of reconstructing steady total pressure distribution and steady-state circumferential distortion index and dynamic total pressure turbulence based on steady and dynamic wall static pressures by using neural network method was conducted. The result showed that the steady total pressure distribution could be well related to the wall steady static pressure by neural network model, so that the steady total pressure flow field could be reconstructed from wall steady static pressure, the high and low pressure extent and steady-state circumferential distortion index of reconstructed total pressure field were very close to those of measured flow field. Adding center total pressure and more wall steady static pressure probes data into the neural network inputs could improve the reconstructing accuracy. The dynamic total pressure turbulence level can be reconstructed by dynamic static pressure turbulence and airflow Mach number, with the reconstruction error within ±0.25%.

Considering the requirements on integration of aircraft/engines, a design method of over-under turbine based combined cycle engine (TBCC) nozzle oriented thrust optimization under a given geometric constraint was proposed. The optimal allocation of the area expansion ratio of the turbo-engine/ramjet nozzle at under-expansion and over-expansion state was realized by means of theoretical analysis and numerical simulation. Within the range of turbo-engine/ramjet pressure-drop-ratio, the TBCC nozzle designed for thrust optimization can achieve higher thrust performance than the baseline nozzle, and the difference in composite thrust coefficient was particularly obvious under the over-expansion state, where the thrust coefficients at Mach number of 0.2 and 3 can be increased by 4.89% and 4.14%, respectively, through thrust optimization. Furthermore, the lift force and pitching moment of thrust optimization nozzle changing with the Mach number were 33% and 47.3% smaller than those of the baseline nozzle, which can effectively reduce the range of aerodynamic focus of the whole aircraft, helping to reduce the trim resistance of wide-speed-range aircraft and the difficulty of flight control.

For the thermal protection of turbine components of high-speed gas turbine engines, a variable cycle turbofan (VCTF) engine with cooled cooling air (CCA) technology was taken as an example. The library of fuel thermal physical property, and the simulation model of the heat exchanger, the turbine blade cooling, and the combustion chamber with fuel temperature change were established. The iteration simulation model of design point for the high-speed VCTF engine was developed, and the influence on the thermodynamic cycle performance of the VCTF engine with CCA technology was analyzed. Results showed that at the same turbine temperature limits level, the CCA technology can further increase the thrust of the high-speed gas turbine engine. However, the application of high-temperature resistant material into turbine blades was still the key to improving the performance of the high-speed gas turbine engine. For the low-pressure turbine (LPT) without high-temperature resistant material, the relative cooling air bleeds of the LPT stator and rotor increased with the increase of high-pressure turbine (HPT) temperature limits level; after the CCA technology was adopted, the relative cooling air bleed of the LPT stator decreased, and that of the LPT rotor without cooled cooling air further increased. Applying high-temperature resistant material into LPT can reduce this adverse effect.

A nonlinear model predictive control based on a reduced order model and smooth switching control was proposed for the conversion maneuvering of tiltrotor aircraft. The simulation model and the reduced order prediction model of tiltrotor aircraft were established. In the framework of the model prediction algorithm, the reduced order model was used to predict the successor states of the vehicle. Combined with smooth switching control of speed and altitude, along with the nacelle control strategy of segmented nacelle tilting rate, the command tracking objective function suitable for transition maneuvering was derived. The control saturation, pitch angle rate, and pitch angle limits were considered in the constraints. The performance of the conversion control method on attitude and rate command tracking during different conversion phases and dynamic transition maneuvering was verified by flight simulations, respectively. Results showed that the inherent characteristics of aircraft at different nacelle angles had little impact on the tracking performance of the controller. In the conversion maneuvering, the controller and the nacelles control strategy enabled the aircraft to complete the conversion procedures with small altitude variation and smooth pitch attitude variation, and the speed tracking and the attitude holding of lateral and heading were excellent.

Based on the clutch dynamic performance calculation model established by the dynamic analysis of the clutch components, the co-working equations of STOVL propulsion system mode conversion process, namely its dynamic performance simulation model, were set up, according to the coupling relationship of short takeoff and vertical landing (STOVL) propulsion system components in different working states. The control strategies for the conversion between STOVL mode and conventional mode were proposed, and the characteristics of the performance parameters during the conversion process under different clutch engagement speeds were compared. Results showed that the maximum error of the simulation model was 3.76% compared with the foreign research data in the process of mode conversion. In the process of mode conversion, the proposed control strategy can ensure that there was no overlimit of temperature and speed of STOVL propulsion system, and the surge margin change was less than 0.1%. In the process of mode conversion, the shorter clutch engagement time of mode conversion process, the greater instantaneous frictional heat generation, but the lower total heat generation.

In view of the problem that the inertial particle separator has low separation efficiency of sands with small diameter sand, a method was proposed to improve the small diameter sand separation efficiency by introducing local jet to form pneumatic bulge, and the effectiveness of the method was verified by simulation. Results showed that, the formation of pneumatic bulge can significantly improve the small diameter sand separation efficiency, and the separation efficiency of AC sands can be increased by 3.6%, or up to 7%, under the premise that the total pressure recovery at the outlet was no more than 0.5%; the introduction of jet can effectively improve the separation efficiency of sands with diameter of 9 μm and below, and the smaller diameter indicated the lower lifting effect. For sands with diameter of 9 μm, the separation efficiency can be increased to 100%; the injection position of the jet should be set on the wall surface of the centrosome bulge, and within a certain range, the greater angle between the jet and incoming flow and the greater jet pressure indicated the higher improvement effect on the small diameter sand separation efficiency.

In order to reveal the mechanism of strengthening the loading performance of multi-layer gas foil thrust bearing, the finite element method considering elastohydrodynamic lubrication in multi-layer gas foil thrust bearing was established. The variation of stiffness with different loads and the evolution of wedge-shaped film’s characteristic parameters were studied. Results showed that the stiffness of the multi-layer foil presented a nonlinear trend of increment with the increase of load capacity, also the variation amplitude of stiffness of the laminated foil kept in a low level at small load capacity, then it became large as the increase of the load capacity, and the stiffness of the circumferential direction changed alternately. It resulted in variation of the characteristic parameters of wedge-shaped clearance with the increase of the capacity. The wedge height attenuated from 45 μm to 16.8 μm and the pitch ratio first increased and then decreased correspondingly when the bearing capacity varied from 15 N to 75 N. Based on the characteristic change of wedge-shape clearance under different loads, the gas film pressure distribution induced a double peak, which enhanced load compared with the wave foil bearing under the same bearing capacity, and the minimum gas film thickness can be increased about 35% to 50%, which significantly reduced the collision and wear probability between rotor and stator.

The stress state and oil film dynamic boundary value of slipper of valve distribution axial piston pump are different from those of end face distribution pump. In order to study its lubrication characteristics, a coupling model for slipper pair working condition simulation and numerical analysis applied to valve distribution axial piston pump was established to analyze the effects of plunger motion frequency, system load and different graded constant flows on the lubrication characteristics of slipper pair. The results showed that the sliding shoe pair of valve distribution overturned mainly in the direction of friction torque, and eccentric wear was more likely to occur in the transition period from high pressure zone to low pressure zone and in the low pressure zone; the increase of the plunger movement frequency could reduce the risk of overturning and eccentric wear of the slipper, but it may also reduce the stability of the slipper pair; with the increase of the system load, the oil film thickness decreased, and the overturning angle of the slipper in the high pressure area decreased, while the overturning angle of the slipper in the low pressure area increased; under different grading and constant flow rates, when the number of plungers was greater than 3, the slippers were not easy to overturn when the odd number of plungers were combined, while the slippers were easy to overturn and wear when the even number of plungers were combined, and the pressure change range in the high and low pressure zones increased.

Considering the worse effect and generalization ability of rolling bearing fault diagnosis under variable working conditions and cross-type conditions as well as the shortage of serious samples in practice, a fault diagnosis method using modified Fourier mode decomposition (MFMD) and Transformer convolutional neural network （Transformer-CNN） was proposed based on the sequence characteristics of vibration signal. The vibration data preprocessing module was designed, in which MFMD and position encoding were adopted to preprocess the samples and mark the sequence position relationships. The Transformer-CNN sequence modeling unit with the scaled dot-product attention mechanism was then designed, and the cyclic stack structure was optimized by the max-pooling, which reduced the network parameters and improved the sequence modeling capability. The pre-training-fine-tuning transfer learning method was adopted to transfer the trained mode parameters to the target domain and fine-tune, which avoided the over-fitting caused by insufficient data. The results showed that Transformer-CNN can reduce the fault diagnosis error by more than 50% compared with the benchmark algorithms. In the case of cross-working and cross-type conditions with small samples, the algorithm enables to achieve 8.75% diagnosis accuracy improvement and faster convergence.

To investigate the effect of slip boundary and bearing parameters on bearing characteristics, the Reynolds equation considering slip boundary was introduced to the multi-leaf foil bearing with bump-foil support. The Newton-Raphson iterative method and the perturbation method were successively employed to linearize the pressure governing equation, and the curved beam model was used to describe the radial foil deformation. The finite difference method was adopted to help establish the fluid-structure coupling solution model of this type of bearing, and the numerical results agreed well with the experimental results. The influences of bearing number, eccentricity ratio, length-diameter ratio, clearance ratio, the number and thickness of top foil on characteristic parameters were studied. The results indicated that slip boundary caused a general decrease of 3% in the load capacity of eight-leaf bearing when the bearing number and length-diameter ratio were small, meaning that the effect of slip boundary should be taken into consideration. Nevertheless, this effect was not sensitive to the change in the length-diameter ratio and thickness of the top foil. Besides, this type of bearing had a better stability. When the bearing number was small, the slip boundary led to the decline of bearing stability.

Taking a domestic independently-developed civil turboshaft engine in development as the object, the maximum unbalance exploration research based on airworthiness requirements was carried out. The deterioration of unbalance during the first TBO of an in-service turboshaft engine was statistically analyzed, finding this turboshaft engine had the same configuration and the same magnitude with this one researched in this paper. The deterioration of unbalance in normal service of this turboshaft engine was 20−70 g·mm, and in those rare cases it reached above 80 g·mm, close to 100 g·mm. On this basis, the unbalance response of the turboshaft engine was analyzed based on the short bearing π oil film model, and the maximum unbalance exploratory experiment was carried out both on a rotor test device and an engine test device. The calculation and experiment results showed that, the engine could work stably at all speeds in normal service unbalance deterioration (70 g·mm); under special cases, when the unbalance reached and exceeded 100 g·mm, the engine vibrated seriously, especially when the unbalance was of out-of-phase distribution, the engine vibrated violently during the 2st critical speed region. The vibration speed range was wider, and the engine could not work properly.

Test and simulation studies were conducted to investigate the fretting fatigue behaviour and damage characterization of turbine attachment. Seven types of turbine attachment with different pressure angles, tooth angles, and tooth spacing were designed, and fatigue tests were carried out under high and low cycle loads. The second tooth of the tenon of all types suffered fretting fatigue failure. And the effect of the geometric parameters of the tenon on fretting fatigue was obtained: with the increase of pressure angle and tooth pitch, the fretting fatigue life gradually decreased; while with the increase of tooth shape angle, the fretting fatigue life was first kept basically the same, and then increased significantly. Critical evidence was provided to reveal the failure mechanism under combined high and low cycle loads, meanwhile important data were generated for determining the controlling parameter and validating the established model of the fretting fatigue for turbine tenons.

A new nonlinear simulation method and model modification method were proposed for the nonlinear stiffness characteristics of bolted structures with notches in aeroengine casings. A nonlinear thin layer element method was proposed to simulate the nonlinear stiffness characteristics of bolted structures, and the simulation results were evaluated. A new dual weighted response surface correction method was also proposed. The modified method was used to modify the static nonlinear model of bolted structures. The results showed that the simplified nonlinear thin layer element model can predict the nonlinear stiffness characteristics of actual bolted structures considering the roughness of the seam contact surface; a dual weighted response surface correction method was proposed to maintain the stiffness error within 3.5% before and after correction, verifying the feasibility of the modeling and correction method.

In view of the fatigue fracture fault of the main blade girder of a helicopter and the failure of fault warning from the pressure sensor, a fault tree was applied to analyze the reason of the faults. Blade failure and main fault mode of the pressure sensor were analyzed, based on which the mechanism of the accident was discussed. The results of macro and micro fractographic analysis showed that spalling holes appeared from multiple non-metallic tesseras on the surface of the girder and the fracture appeared from the bottom of the fatigue spalling holes on the external surface of the lower main girder, which demonstrated clear characteristics of fracture caused by high-cycle fatigue. Energy spectrum analysis to the source area of the fracture showed the existence of Si and O-rich tesseras on the surface and sub-surface of the main girder. General investigation carried after the accident showed that the pressure sensor had the fault mode of proper manual inspection function but without warning when the pressure level of the base chamber was low. Comprehensive analysis indicated improper shot-peening parameters. There were initial defects on part of the surface of the main girder, which led to the initiation and extension of fatigue crack from the part of defect because of the fatigue load. Pressure level of the main girder dropped out of air leakage when the fatigue crack extended throughout the inner chamber. Meanwhile, the sealing component of the pressure sensor failed, leading to pressure loss of the base chamber. In this way, the warning function and blade monitoring safety mechanism failed and the crack running through the main girder extended along both directions of the girder, leading to the fracture of the blade after 23 ups and downs.

To address the combustion problem of polymer skeleton-reinforced paraffin fuel in hybrid rocket motors, CFD software was used to carry out numerical simulation of the combustion process of helical and hexagonal skeleton-reinforced paraffin fuel in a direct flow/swirling injection solid-gaseous hybrid combustor. The combustion process of the four conditions was compared, and the influences of the skeleton structure and injection method on the combustion were analyzed. The results showed that the regression rates of the skeleton material and the paraffin-based fuel were quite different, and the skeleton structure gradually became prominent as the combustion progressed. The turbulence intensity and fuel mass flow both affected the combustion chamber temperature, and the combustion chamber temperature volatility decreased as the combustion progressed. In the swirling injection condition, the temperature of the combustion chamber was higher than that in the direct flow injection condition, there was a high-temperature area at the head of the combustion chamber, and the axial temperature distribution was more uniform. In the direct flow injection condition, the temperature increased sharply in the middle of the combustion chamber. In addition, compared with the hexagonal skeleton, the spiral skeleton can provide greater turbulence strength under direct flow injection conditions. In the swirling injection conditions, the swirl intensity of the oxidant played a leading role in the increase of the turbulence intensity, and the skeleton structure had little effect.

Taking the inlet pressure of the oxidizer boost pump as the monitoring object, the fuel flow rate into the gas generator can be reduced by the action flow regulator during the cavitation failure of the oxidizer main pump, then the mixture ratio can be stabilized and the combustion temperature can be suppressed, so as to avoid the catastrophic consequences of excessive gas temperature. The control strategies of 98.8%, 97.2%, 89.5% and 74.5% rated flow rate and the control delay of 0, 0.15, 0.23 s and 0.30 s were set up. The effectiveness was studied by simulation. The results showed that the optimum fuel flow rate under cavitation failure was used to keep the mixing ratio rating, and the maximum allowable time delay decreased with the increase of cavitation failure severity. When the tank pressure was reduced to 53%, 43% and 33% of the rated pressure within 0.10 s, the maximum allowable action delay was 0.23, 0.17 s and 0.13 s, respectively. The optimal control time delay reference was 0.09 s, increasing the delay led to overshoot of some component parameters, and reducing the delay led to the possibility of engine flameout.

In view of the contradiction between reducing the pressure peak in low pressure chamber and increasing the exit velocity of self-ejection, the optimization design of self-ejection was carried out. A motor-low pressure coupled internal ballistic solution model of self-ejection was developed, physical experiments of two working conditions with four rockets were carried out. The multi-population differential evolution algorithm using differential strategy was proposed, “eliminate-complement operation” was adopted to deal with the constraints in the optimization process. A two-objective self-ejection constraint optimization model was established with the peak of pressure in low pressure chamber and the exit velocity considering the actual constraints of self-ejection, and this optimization model was calculated by the multi-population differential evolution algorithm. Results showed that, the Pareto front was of two segments with different slopes approximately. With the increase of pressure peak of low pressure chamber, the same increment of pressure led to a smaller increment of exit velocity. Twelve schemes were evenly selected in Pareto front and ranked by using technique for order preference by similarity to an ideal solution; the peak of pressure in low pressure chamber of the final optimized scheme of self-ejection decreased by 16.11%, the exit velocity of the final optimized scheme increased by 54.55%, and the performance of rocket projectile self-ejection had been improved.

To improve the response speed of aero-engine performance seeking control, an aero-engine performance seeking control method based on nonlinear model prediction control was proposed. The full envelope interpolation compact propulsion system dynamic model was used as the on-board model to estimate the engine performance parameters and the future outputs of the limited time domain. With the nonlinear model prediction control method, three typical performance seeking control modes, i.e.: the maximum thrust mode, the minimum fuel consumption mode and the minimum turbine temperature mode were changed into real-time performance seeking, and corresponding real-time control performance indexes were designed to improve engine response speed. The simulation results showed that compared with the traditional method, the proposed control method had better control effect under three performance optimization control modes, and the response speed increased from 0.5 s to 5 s. At the high altitude cruising working point the thrust of the maximum thrust mode increased by 19.8%; the fuel consumption rate of the minimum fuel consumption mode dropped by 3.12%; the temperature of the minimum turbine temperature mode dropped by 17 K, helping to verify the effectiveness of the control method.