A mobile ejector combined cycle engine and its mode conversion control method

Abstract

The present invention discloses a mobile ejector combined cycle engine and a mode conversion control method thereof, which relates to the technical field of combined cycle engines. A mobile ejector engine can be formed by introducing a movable ejector cone and arranging an ejector engine in the ejector cone. The mobile ejector engine can provide a thrust ranging from 1.6 Ma to 3 Ma. Thus, under the coordinated operation of a turbofan engine and the mobile ejector engine, the "thrust gap" problem between the thrust provided by the turbofan engine and the thrust provided by a sub-combustion mode ramjet is solved, and continuous and stable thrust output of the mobile ejector combined cycle engine in a wide speed range from 0 to 7 Ma is achieved.

Description

Technical Field

[0001] The present invention relates to the field of combined cycle engines, and in particular to a mobile ejector combined cycle engine and a mode conversion control method thereof. Background Art

[0002] The propulsion mission of the TBCC engine (Turbine-Based Combined Cycle) from 0 to 1.8 Ma is completed by the turbofan engine, the propulsion mission from 2.8 Ma to 4 Ma is completed by the subsonic mode ramjet engine, and the propulsion mission from 3.8 Ma to 7 Ma is completed by the scramjet engine. The turbofan engine and the subsonic / scramjet dual-mode ramjet engine are relatively independent power units. Among them, there is a "thrust gap" between the thrust provided by the turbofan engine and the thrust provided by the subsonic mode ramjet engine, and continuous thrust output cannot be achieved. The American HyperSpace company proposed a new concept of Hyscram combined cycle engine, whose main structure is as follows: Figure 1 As shown, it can provide full-range thrust for hypersonic vehicles, from a standing start to acceleration to 8 Ma flight. However, this engine envisions the use of numerous new technologies, including superconducting turbine propulsion and power generation, a plasma electromagnetic dual-mode ramjet, a superconducting electric fan, an electric compressor, a plasma combustion chamber, and a magnetic levitation system. Currently, the implementation of these technologies remains challenging and requires extensive technical research. Therefore, this invention improves upon existing TBCC engines and proposes a mobile ejector combined cycle engine and its mode conversion control method. Summary of the Invention

[0003] The purpose of the present invention is to provide a mobile ejector combined cycle engine and a mode conversion control method thereof, which introduces a mobile ejector engine on the basis of the existing TBCC engine to achieve continuous and stable thrust output of the mobile ejector combined cycle engine in a wide speed range from 0 to 7 Ma.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] In a first aspect, the present application provides a mobile ejector combined cycle engine, comprising: a casing, a turbofan engine, a subsonic and scramjet dual-mode ramjet engine, and a mobile ejector engine;

[0006] The housing includes a movable ejector thrust section, a turbofan thrust section, a sub-combustion and scramjet thrust section, and a common tail nozzle;

[0007] The mobile ejector thrust section is arranged in a common air inlet passage; the mobile ejector thrust section, the turbofan thrust section and the common tail nozzle are arranged in sequence along the air inlet direction; the subcombustion and scramjet thrust section is sheathed outside the mobile ejector thrust section and the turbofan thrust section; the subcombustion and scramjet thrust section, the mobile ejector thrust section and the turbofan thrust section are connected; the exhaust passages of the turbofan thrust section and the subcombustion and scramjet thrust sections are connected to the common tail nozzle;

[0008] The mobile ejector engine is arranged in the mobile ejector thrust area; the turbofan engine is arranged in the turbofan thrust area; the sub-combustion and scramjet dual-mode ramjet engine is arranged in the sub-combustion and scramjet thrust area; the sub-combustion and scramjet dual-mode ramjet engine is sleeved on the outside of the turbofan engine;

[0009] The mobile ejector engine includes a movable ejector cone and an ejector engine arranged in the movable ejector cone; the movable ejector cone is movably connected to the shell; a plurality of jet holes are provided at one end of the movable ejector cone close to the turbofan engine; the jet holes are used to eject the jet generated by the ejector engine; the cone tip of the movable ejector cone is located at the air inlet end; the front end cone body of the movable ejector cone is a closed structure, and the rear end cavity of the movable ejector cone is an incompletely closed structure.

[0010] In a second aspect, the present application provides a mode conversion control method for a mobile ejector combined cycle engine, the method comprising a turbojet-to-ejector mode conversion control method; the turbojet-to-ejector mode conversion control method specifically comprising:

[0011] Obtain the total thrust generated by the mobile ejector combined cycle engine, i.e. the first thrust, in real time;

[0012] When the first thrust reaches a turbojet-ejector mode transition point, calculating a first thrust error and a first thrust error integral between the first thrust and a corresponding target thrust;

[0013] The first thrust error, the first thrust error integral, and the first thrust are used as state parameters, and a DDPG algorithm is used to calculate first action parameters under the state parameters to gradually close the turbojet inlet and turbojet exhaust channels of the turbofan engine, gradually open the ejector inlet and ejector exhaust channels of the ejector engine, gradually stop the fuel supply to the turbojet engine, and start to increase the fuel supply to the ejector engine; the first action parameters include the turbojet channel splitter angle and turbojet channel plugging cone stroke of the turbofan engine, the ejector channel splitter angle and ejector channel plugging cone stroke of the ejector engine, the turbojet engine fuel quantity, and the ejector engine fuel quantity.

[0014] In a third aspect, the present application provides a mode conversion control method for a mobile ejector combined cycle engine, characterized in that the method includes an ejector-subcombustion mode conversion control method; the ejector-subcombustion mode conversion control method specifically includes:

[0015] Obtain the total thrust generated by the mobile ejector combined cycle engine, i.e. the second thrust, in real time;

[0016] When the second thrust reaches the ejection-sub-combustion mode transition point, calculating a second thrust error and a second thrust error integral of the second thrust and the corresponding target thrust;

[0017] The second thrust error, the second thrust error integral, and the second thrust are used as state parameters, and the DDPG algorithm is used to calculate second action parameters under the state parameters to gradually stop the fuel supply to the ejector engine and start increasing the fuel supply to the sub-combustion afterburner; the second action parameters include the fuel amount of the ejector engine and the sub-combustion afterburner fuel amount.

[0018] In a fourth aspect, the present application provides a mode conversion control method for a mobile ejector combined cycle engine, the method comprising a subsonic-scram combustible mode conversion control method; the subsonic-scram combustible mode conversion control method specifically comprising:

[0019] Obtain the total thrust generated by the mobile ejector combined cycle engine, i.e. the third thrust, in real time;

[0020] When the third thrust reaches a subsonic-scram mode transition point, calculating a third thrust error and a third thrust error integral between the third thrust and a corresponding target thrust;

[0021] The third thrust error, the third thrust error integral, and the third thrust are used as state parameters, and the DDPG algorithm is used to calculate third action parameters under the state parameters to gradually close the ejector engine's ejector intake channel and ejector exhaust channel, gradually stop the fuel supply to the sub-combustion afterburner, and start increasing the fuel supply to the scramjet combustion chamber; the third action parameters include the ejector engine's ejector channel diverter plate angle and ejector channel blocking cone stroke, ejector channel blocking cone stroke, sub-combustion afterburning fuel quantity, and scramjet fuel quantity.

[0022] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0023] The present invention provides a mobile ejector combined cycle engine and a mode conversion control method thereof. A movable ejector cone is introduced, and an ejector engine is arranged in the ejector cone to form a mobile ejector engine. The mobile ejector engine can provide a thrust ranging from 1.6 Ma to 3 Ma. Therefore, under the coordinated operation of the turbofan engine and the mobile ejector engine, the "thrust gap" problem between the thrust provided by the turbofan engine and the thrust provided by the sub-combustion mode ramjet is solved, and continuous and stable thrust output of the mobile ejector combined cycle engine in a wide speed range from 0 to 7 Ma is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 A schematic structural diagram of a Hyscram combined cycle engine provided by an embodiment of the present invention;

[0026] Figure 2 A schematic diagram of the technical concept of a mobile ejector combined cycle high-speed engine according to an embodiment of the present invention;

[0027] Figure 3 A schematic diagram of the composition and structure of a mobile ejector combined cycle high-speed engine provided by an embodiment of the present invention;

[0028] Figure 4 A schematic structural diagram of a mobile ejector engine provided in an embodiment of the present invention;

[0029] Figure 5 A three-dimensional side view of a mobile ejector combined cycle high-speed engine provided by an embodiment of the present invention;

[0030] Figure 6 A schematic diagram of the mode conversion process of a mobile ejector combined cycle high-speed engine provided by an embodiment of the present invention;

[0031] Figure 7 A schematic diagram of the structure of a switching control system provided by an embodiment of the present invention;

[0032] Figure 8 A schematic diagram of the overall framework of a control system provided by an embodiment of the present invention;

[0033] Figure 9 A schematic diagram of the action network structure provided by an embodiment of the present invention;

[0034] Figure 10 A schematic diagram of the evaluation network structure provided by an embodiment of the present invention;

[0035] Figure 11 A schematic diagram of the overall framework of real-time training for handover control provided by an embodiment of the present invention;

[0036] Figure 12 A flow chart of a real-time control algorithm for a mobile ejector combined cycle high-speed engine according to an embodiment of the present invention;

[0037] Figure 13 Schematic diagram of data interaction between a mobile ejector combined cycle high-speed engine and an RL Agent provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0039] The purpose of the present invention is to provide a mobile ejector combined cycle engine and a mode conversion control method thereof, which introduces a mobile ejector engine on the basis of the existing TBCC engine to achieve continuous and stable thrust output of the mobile ejector combined cycle engine in a wide speed range from 0 to 7 Ma.

[0040] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0041] Example

[0042] The present invention first designs the composition and structure of the mobile ejector combined cycle high-speed engine, obtains the performance parameters of its various components, then designs the shape and position of the ejector cone and models it to obtain its performance parameters. On the above basis, combined with the component-level Simulink model of the existing turbofan engine and the subsonic / scramjet dual-mode ramjet engine, a component-level Simulink model of the mobile ejector combined cycle high-speed engine is developed, which can perform open-loop thrust simulation. Finally, the control strategy and key control technology of the mobile ejector combined cycle high-speed engine are studied to realize the control of each mode conversion. The specific scheme is as follows: Figure 2 The technical approach of this invention is divided into two main parts: the design and modeling of a mobile ejector combined cycle high-speed engine; and the design of control technology for the mobile ejector combined cycle high-speed engine. The key technology of this invention is the multi-task coordinated control technology of the mobile ejector from the turbofan mode to the ejector sub-combustion mode.

[0043] Among them, such as Figure 3As shown, a mobile ejector combined cycle engine (i.e., a mobile ejector combined cycle high-speed engine) in this embodiment includes: a casing, a turbofan engine, a subsonic / scramjet dual-mode ramjet engine, and a mobile ejector engine;

[0044] The shell includes a moving ejector thrust area, a turbofan thrust area, a sub-combustion and scramjet thrust area and a common tail nozzle.

[0045] The mobile ejector thrust section is arranged in a common air inlet passage; the mobile ejector thrust section, the turbofan thrust section and the common tail nozzle are arranged in sequence along the air inlet direction; the subcombustion and scramjet thrust section is sleeved on the outside of the mobile ejector thrust section and the turbofan thrust section; the subcombustion and scramjet thrust section, the mobile ejector thrust section and the turbofan thrust section are connected; the exhaust passages of the turbofan thrust section and the subcombustion and scramjet thrust section are connected to the common tail nozzle.

[0046] The mobile ejector engine is arranged in the mobile ejector thrust area; the turbofan engine is arranged in the turbofan thrust area; the subsonic and scramjet dual-mode ramjet engine is arranged in the subsonic and scramjet thrust area; the subsonic and scramjet dual-mode ramjet engine is sleeved on the outside of the turbofan engine.

[0047] Figure 3 The inlet, separator, scramjet / ramjet combustor, adjustable ramjet tail nozzle, fan blades, high-pressure compressor, combustor, high-pressure turbine, low-pressure turbine, and common tail nozzle shown in the figure adopt the existing TBCC engine structure, and their positional relationships and connections adopt the existing methods. Building on the existing TBCC engine structure, the ejector engine and movable intake cone are newly introduced, further forming the ejector nozzle and ejector section within the space where the scramjet / ramjet engine is located.

[0048] like Figure 4 As shown, the mobile ejector engine includes a movable ejector cone (i.e., an intake cone) and an ejector engine disposed within the movable ejector cone. The movable ejector cone is movably connected to the housing. A plurality of jet holes are disposed at one end of the movable ejector cone near the turbofan engine. The jet holes are used to eject the jet generated by the ejector engine. The tip of the movable ejector cone is located at the intake end. The jet holes correspond one-to-one with the multiple nozzles of the ejector engine.

[0049] The intake cone isn't a completely enclosed structure. The front cone is closed, while the rear is partially enclosed. This allows air to enter the jet engine after passing through the intake duct. The jet holes at the rear of the intake cone are used to eject the jet generated by the jet engine. Specifically, the front cone of the intake cone is a closed structure, while the rear cavity connected to it is a partially enclosed structure. The rear cavity has an air inlet (e.g., located on the front side of the cavity) and jet holes on the rear side of the cavity.

[0050] The air intake is achieved by the air intake duct. The movable ejector cone is the combination of the air intake cone and the ejector engine and is located behind the air intake duct. The air intake cone structure is a cone structure.

[0051] Because the fluid properties of supersonic airflow differ significantly from those of sonic airflow, the engine is essentially ingesting gases in completely different states, creating "turbulence" that can severely disrupt engine operation. The present invention adjusts the airflow through the intake duct, allowing the engine to operate under ideal conditions. The present invention primarily utilizes shock waves to compress the airflow. This involves the airflow passing through a tapered intake cone, forming a conical shock wave. The airflow then passes through the intake duct, undergoing multiple reflections within the duct (in the intake duct, airflow reflects when it strikes the inner wall of the duct; multiple reflections occur when the airflow strikes the wall), ultimately slowing the airflow to subsonic speeds. Furthermore, the slower the speed, the greater the inclination angle of the shock wave. This means that at slow speeds, the intake duct may not be able to enclose the entire conical shock wave, resulting in insufficient air intake for the engine. In this case, the intake cone needs to be moved backward, allowing the shock wave surface to be enclosed by the entire intake duct, ensuring high operating efficiency under all aircraft flight conditions.

[0052] The ejector cone and the shell can be combined in a coaxial nesting manner, and the ejector cone can move along the axial direction.

[0053] like Figure 4As shown, a multi-nozzle ejector engine is integrated with an inlet duct control cone. The ejector engine is built into the inlet cone, forming an integrated, controllable, and movable ejector engine. Multiple engine jet nozzles, for example, six, are arranged on the rear side of the inlet cone. The fuel flow of the embedded ejector engine and the position of the inlet cone are adjusted according to flight speed and thrust requirements. The inlet cone is a conical structure behind the inlet duct. The intake area of ​​the inlet duct opening changes with the movement of the inlet cone, achieving coordinated adjustment of the ejector primary flow and the inlet area. The ejector suction effect is used to introduce secondary air flow, improving the operating range and efficiency of the subsonic ramjet at low Mach numbers, and meeting the inlet ramjet intake volume requirements for the subsonic and scramjet modes. The movable ejector cone is essentially a nested combination of the ejector engine and the inlet cone. The aforementioned effects are achieved by adjusting the fuel flow of the embedded ejector engine, the position of the inlet cone, and the inlet flow according to flight speed and thrust requirements.

[0054] Generally, the position of the intake cone is controlled in a single mode. The purpose of controlling the cone position is to provide the required intake volume to the corresponding engine in different single modes.

[0055] The ejector cone uses a high-speed, high-energy flow (liquid, air, or other material flow) to eject a lower-speed, low-energy flow. This effect is achieved through boundary mixing. After the secondary air is drawn into the intake duct, the ejector flow transfers energy to the ejected flow. The resulting mixing zone gradually expands to fill the entire mixing chamber. After a period of mixing, the flow becomes nearly uniform at the mixing chamber outlet, contributing to efficient engine operation.

[0056] The ejector is actually completed by the ejector. After the airflow passes through the intake cone, the ejector rectifies the gas again before entering the ejector engine. The ejector is a structure that realizes the ejection suction effect to obtain greater thrust. It is a part of the ejector engine. The existing ejector engine has this structure, and its design goal is to achieve ejection suction. The principle of the ejector structure is that the gas enters the suction chamber of the ejector after being accelerated by the nozzle, and the ejected airflow is brought into the mixing chamber of the ejector. The airflow is fully mixed in the mixing chamber and finally discharged to the subsequent engine.

[0057] The air intake cone is nested with the ejector engine, and the air intake cone will move according to the flight speed, causing the ejector engine to move accordingly.

[0058] like Figure 5As shown, the structure of the improved TBCC engine of the present invention comprises: ① a mobile ejector engine composed of an intake cone and a multi-nozzle ejector engine, located at the front end of the engine. ② A turbofan engine is connected to it and located in the middle section of the engine. ③ A subsonic / scramjet dual-mode ramjet engine is nested externally (i.e., outside the space where the mobile ejector engine and turbofan engine reside). Finally, ④ a shared tail nozzle is located at the rear end of the engine. From an innovative perspective, the present invention proposes the innovative concept of a mobile ejector engine, which offers significant advantages such as high ejection efficiency, continuous adjustment of intake volume, avoidance of thermal congestion, improved thrust enhancement ratio, increased space utilization, and reduced weight. This can facilitate the coordinated operation of the turbofan engine and ejector engine, thereby smoothly and effectively bridging the "thrust gap" and successfully completing the propulsion mission from 1.6 Ma to 3 Ma. Existing TBCC engines have a separate intake duct at the front end. Due to the lack of an intake cone, congestion can occur during intake due to the dual effects of ejection and velocity ramjet. The integrated design of the ejector engine and intake cone eliminates a set of servo control devices, improving space utilization and reducing weight.

[0059] The mobile ejector combined cycle high-speed engine of the present invention is based on the traditional turbine-based combined cycle engine (TBCC engine). A turbine-based combined cycle engine is a combined cycle propulsion system constructed by connecting ramjets (including sub-ramjets and scramjets) in parallel or series with a gas turbine engine. The mobile ejector combined cycle high-speed engine of the present invention also incorporates a dual-channel turbofan engine and a sub-ramjets / scramjets dual-mode ramjets (the original structure of the traditional TBCC engine), but adopts a compact, centrosymmetrical, nested structure.

[0060] The present invention further carries out component-level Simulink modeling for an engine with a symmetrical thrust layout including a shared air inlet, a twin-turbojet engine, an ejector rocket, a subsonic combustion chamber, a scramjic combustion chamber, a shared tail nozzle, etc. Specifically, the design of the engine's key components, including the thrust requirements, operating range, and performance parameters of the turbofan engine and scramjet, was based on the existing turbofan engine models and thrust, and scramjet engine models and thrust. The proposed turbofan engine and scramjet engine were then developed, and the thrust requirements, operating range, and performance parameters were iteratively adjusted using GasTurb software. The design then determined the shape and area of ​​the intake cone based on the overall propulsion missions of achieving 0 to 1.8 Ma for the turbofan engine, 1.6 to 3 Ma for the mobile ejector cone and scramjet mode, 2.8 to 4 Ma for the scramjet mode, and 3.8 to 7 Ma for the scramjet mode. The shape and range of the mobile ejector cone were then designed. Using the ejector scramjet CFD calculation program developed by Associate Professor Zhu Chengxiang of the Department of Aerospace at Xiamen University, the ejector engine's outlet pressure, Mach number, equivalence ratio, and other performance parameters were obtained. This resulted in the component models, performance parameters, and overall engine model of the mobile ejector high-speed engine. Then, based on the above-mentioned parameters of each engine component, the mobile ejector engine is modeled by components to obtain the engine Simulink model: According to the subsonic / scramjet dual-mode ramjet parameters obtained by GasTurb software, the thrust and performance parameters of the mobile ejector subsonic mode are calculated respectively. By integrating the data of several calculation points, the thrust requirement, working range and performance parameters of the mobile ejector subsonic mode are established. Then, according to the subsonic mode engine, scramjet mode engine, and intake cone characteristics, and with the help of the ejector subsonic CFD of Associate Professor Zhu Chengxiang of the Department of Aerospace at Xiamen University, the thrust and performance parameters of the mobile ejector subsonic mode are calculated respectively. The calculation program has established an inlet program, established the requirements and performance parameters of the adjustable common air inlet and adjustable common tail nozzle, and finally brought the operating range and performance parameters of the above turbofan engine, sub-combustion / scramjet dual-mode ramjet engine, adjustable common air inlet, and adjustable common tail nozzle into the existing NASA TMATS turbofan engine and sub-combustion / scramjet dual-mode ramjet component-level Simulink model, and focused on developing the Simulink model of the mobile ejector sub-combustion mode to form a complete Simulink model of the mobile ejector combined cycle high-speed engine. At this point, a mobile ejector engine model capable of open-loop thrust simulation is obtained, but the key problem currently existing in this model is the difficulty of multi-modal collaborative control. Therefore, to address this problem, the present invention further designs the control strategy and key control technology of the mobile ejector combined cycle high-speed engine. The technical route is:

[0061] 1. First, design the working mode of the mobile ejector combined cycle high-speed engine and obtain different working modes of the mobile ejector engine

[0062] Under most flight conditions, a mobile ejector combined cycle high-speed engine operates in a single mode, a SISO system. In the mobile ejector combined cycle high-speed engine simulation model, a PID controller is used for closed-loop thrust control, with the input being the engine's fuel flow rate in each single mode. For the single modes, this paper employs a PID tuner for parameter design. Without delving into further detail, the focus will be on the intermediate transitions between modes.

[0063] During the modal conversion process, it is generally necessary to maintain the total output thrust of the mobile ejector combined cycle high-speed engine unchanged, while the altitude and Mach number need to be kept as stable as possible to achieve the transition from providing thrust in the low-speed flow channel to providing thrust in the high-speed flow channel. Unlike when each sub-engine works alone, when working alone, the combined inlet duct splitter and the blocking cone of the common tail nozzle are fixed. During the modal conversion process, the geometric configurations of the combined inlet duct and the common tail nozzle are changed, which brings about changes in the inlet conditions of each flow channel. The change in the inlet conditions will lead to the nonlinear enhancement of the engine system. At the same time, it will face multiple disturbances during the modal conversion process, and strong disturbances will cause the engine to become unstable. Here are the following modal conversion processes:

[0064] (1) Turbojet-jet mode conversion

[0065] The turbojet-ejector mode conversion is the conversion from the turbojet mode to the ejector mode. The present invention selects Mach number 2 as the mode conversion point. The control input and actuation trend of each component in the mode conversion process are shown in Table 1.

[0066] Table 1 Turbojet-ejector mode conversion control input and change law

[0067]

[0068] Among them, the ejector channel diverter plate, the ejector channel blocking cone, the ejector air intake channel and the ejector exhaust channel are all existing structures of the ejector engine itself.

[0069] The safety margins involved in the turbojet-jet mode conversion process include the turbojet engine compressor surge margin SM HPC and the rocket combustion chamber temperature T i The restrictions are as follows:

[0070] SM HPC >10

[0071] T i <3400K (1)

[0072] Here, the subscript i represents the temperature of the rocket combustion chamber involved in the turbojet-ejector mode transition process, which is distinguished from other stages. The subscript of the ejector-subjet mode transition process is s, and the subjet-superjet mode transition process is r.

[0073] (2) Ejection-subcombustion mode conversion

[0074] The ejection-subcombustion mode conversion is the conversion from the ejection mode to the subcombustion mode, which does not involve the switching of the flow channel. Here, Mach number 3 is selected as the mode conversion point. The control input and actuation trend of each component in the mode conversion process are shown in Table 2.

[0075] Table 2 Control input and change rules of ejection-subcombustion mode conversion

[0076]

[0077] The safety boundaries involved in the injection-subcombustion mode conversion process include the inlet duct non-start boundary and the subcombustion chamber temperature T s Limitation, for the characterization of the problem of sub-combustion inlet duct not starting, the inlet duct inlet and outlet back pressure ratio π ram As an indicator, the specific formula is as follows:

[0078]

[0079] T s <3800K (2)

[0080] π ram * It represents the upper limit of the inlet inlet and outlet back pressure ratio during the injection sub-combustion stage. The number without * is the actual inlet inlet and outlet back pressure ratio.

[0081] (3) Subcombustion-supercombustion mode conversion

[0082] The subsonic-scram mode transition is the transition from the subsonic mode to the scramjet mode. Here, Mach number 4.5 is selected as the mode transition point. The control input and actuation trends of each component during the mode transition process are shown in Table 3.

[0083] Table 3 Subcombustion-supercombustion mode conversion control input and change law

[0084]

[0085]

[0086] The safety boundaries involved in the subsonic-scramjet mode transition process include the scramjet engine inlet non-start boundary and the scramjet engine combustion chamber temperature T r Limitation, for the characterization of the problem of the scramjet inlet not starting, the inlet inlet and outlet back pressure ratio π is used scrAs an indicator, the specific formula is as follows:

[0087]

[0088] T r <1200K (3)

[0089] 2. After obtaining the different modes of the engine in the above steps, further design the mode conversion control strategy to achieve a safe and smooth mode conversion process of the engine. The mode conversion process of the mobile ejector combined cycle high-speed engine is as follows: Figure 6 shown.

[0090] In order to solve the multi-mode control problem of the mobile ejector combined cycle high-speed engine, the present invention designs a switching control system. Its main task is to enable the control system to quickly switch to the controller of the corresponding mode as the operating state of the mobile ejector combined cycle high-speed engine changes, and the switching operation cannot affect the stability of the system.

[0091] Due to the multi-objective and multi-task nature of the switching system, the switching controller design must meet the following requirements:

[0092] (1) At the same time, there is only one sub-circuit that remains online;

[0093] (2) When the incoming flow conditions of the mobile ejector combined cycle high-speed engine reach the mode switching point, the switching control system can make the switching operation in time to ensure that the controller in the corresponding mode quickly switches into the system;

[0094] (3) Ensure worry-free system switching during controller switching of each mode.

[0095] To ensure the above requirements, the present invention adopts a two-layer control structure, including a sub-controller that ensures good tracking performance of each thrust closed-loop control sub-circuit and an upper-layer switching controller that ensures worry-free and fast switching. According to the operating characteristics of the mobile ejection combined cycle high-speed engine, a reasonable mode conversion rule is designed, as shown in Table 4. The mode conversion process mainly selects different engines according to the different flight altitudes and Mach numbers in the flight mission. The switching control system diagram is shown in the figure below. Figure 7 As shown, the shaded part in the sub-controller indicates the use of an intelligent controller.

[0096] Table 4 Modal conversion rules

[0097]

[0098] 3. In order to implement the modal conversion control strategy designed in the above steps, the present invention designs a multi-input multi-task intelligent collaborative control algorithm for this strategy.

[0099] For each of the three modal transition processes, three intelligent agents (modal switching controllers) were designed and trained based on the DDPG algorithm. The overall control system framework, neural network structure, reward function, and controller parameter training strategy were designed. The actuator and evaluator update and control processes in the DDPG algorithm follow the existing processes of the traditional DDPG algorithm. This paper applies this DDPG algorithm to the modal switching controller.

[0100] First, the overall framework of the intelligent controller is designed. The intelligent controller of the mobile ejector combined cycle high-speed engine proposed in this invention adopts a unified basic framework. The control-update basic framework is as follows: Figure 8 As shown in FIG, it mainly consists of a mobile ejector combined cycle high-speed engine simulation model and an intelligent controller, wherein the main components of the intelligent controller include an actuator, an evaluator, and an experience replay pool R.

[0101] The basic control-update framework primarily consists of an actuator and an evaluator. The controlled object is a simulation model of a specific mobile-ejector combined-cycle high-speed engine undergoing modal transitions. The actor network in the actuator serves as a controller, receiving inputs such as state, namely, the controlled variable and engine safety parameters. After passing through the controller, an action command is output, which acts on the mobile-ejector combined-cycle high-speed engine and generates the next state and reward value, which are fed back to the controller. The array packet for each round is stored in an experience replay pool. The parameters of the actor network in the actuator are updated using the strategy gradient calculated by optimizer 1. Simultaneously, the parameters of the target actor network are updated using a soft update criterion. The critic network in the evaluator evaluates the action commands generated by the actor network and outputs Q-values, which are used to update the actor network. The critic network parameters are updated by optimizer 2 by minimizing the loss function. Simultaneously, the parameters of the target critic network are updated using a soft update criterion.

[0102] Then the intelligent controller network of mode conversion of mobile ejector combined cycle high-speed engine is designed.

[0103] (1) Network design of intelligent controller for turbojet-jet mode conversion

[0104] Based on the control objectives and the knowledge of traditional controller design, the state s is selected as the error e between the thrust generated by the mobile ejector combined cycle high-speed engine and the target thrust. Thrust , thrust error integral es Thrust And the total thrust Thrust generated (i.e. the thrust generated by the moving ejector combined cycle high-speed engine), the output is the combined inlet turbojet splitter angle Sp Trb1 With Sp Trb2 , injection channel splitter plate angle Sp Ejc , turbojet engine fuel quantity WfTrb1 With Wf Trb2 and the fuel supply of the launch rocket Wf Roc , and the corresponding common tail nozzle plugging cone stroke Cone Trb1 ,Cone Trb2 ,Cone Ejc It should be noted that the state needs to be normalized before being input into the neural network.

[0105] The target thrust is determined based on the flight mission and the target Mach number during flight. The target thrust is generated by CFD software optimization.

[0106] Exploration is crucial for intelligent agents. Since DDPG uses deterministic policy gradients and is inherently lacking in exploration capabilities, Gaussian noise is added to the output of actions in the early stages of training to increase the diversity of the experience pool.

[0107] The update direction of the parameters θ of the online action network is related to Q(s t ,a t ) value function increases in the same direction. Network updates are similar to those of general supervised learning algorithms, and network weights are updated using error backpropagation based on the network loss function. For actor networks, network updates apply the chain rule. The policy gradient obtained by differentiating the expected future reward function J with respect to the network parameters θ can be expressed as follows:

[0108]

[0109] Function J is used to measure the performance of the executor and is defined as follows based on N training samples sampled in batches:

[0110]

[0111] For the turbojet-ejector mode conversion problem, which involves 6-dimensional input and has coupling effects, a simple 3-layer neural network is difficult to complete the search task. Finally, one input layer, three intermediate layers (hidden layers) and one output layer are determined. The input layer is only used to receive state input. Each intermediate layer has 30 neurons. The range of each output parameter is determined based on the prior knowledge of the traditional controller. The tanh activation function is used to map the output layer result to [-1,1] to limit the output range. The action network structure is as follows: Figure 9 shown.

[0112] The target actor network samples actions through the experience pool and provides action input to the target critic network, whose input is the next state s' of the mobile ejection combined cycle high-speed engine. t, the network outputs the action a' at the next moment. Since it is copied from the Actor network, it has the same structure as the Actor network. The target action network adopts the "soft" update mode, as shown in the following formula:

[0113] θ'←ξθ+(1-ξ)θ' (6)

[0114] To ensure the stability of learning, it is necessary to ensure that ξ<1. Although this sacrifices the parameter update speed and reduces the learning speed, it enhances the robustness of the algorithm.

[0115] The input of the online evaluation network is state s and action a. The purpose is to fit the action value function Q(s,a) to provide direction for the parameter update of the action network. Its policy gradient comes from the loss function that minimizes the online evaluation function. Each time, N experiences (s i ,a i ,r i ,s i+1 ), achieve the target value y i The fitting loss function can be expressed as:

[0116]

[0117] in:

[0118] y i =r i +γQ t (s′ i ,π(s′ i |θ t )|φ t ) (8)

[0119] The evaluation network consists of two paths. The state path uses two intermediate layers, each containing 30 neurons for feature extraction. The activation function uses the ReLU function to achieve network sparsity. For the action path, given its high dimensionality and the coupling between various components, the network structure is designed with three intermediate layers and 30 neurons in each layer. The activation function still uses the ReLU function. The output layer outputs the current action value as a scalar with a number of neurons of 1. The network structure is as follows Figure 10 shown.

[0120] The input of the target evaluation network is the state s' at the next moment t And the output a' of the target action network t , the output is the value of the action at the next moment Q'(s' t ,a' t), the purpose is to optimize the update iteration direction of the online action network and improve the convergence speed of the network. It has the same structure as the online evaluation network, and the update method of the parameter φ' also adopts the "soft" update method:

[0121] φ'←ξφ+(1-ξ)φ' (9)

[0122] The mobile ejector combined cycle high-speed engine structure selected in this paper adopts a four-channel structure with symmetrical axis distribution. There are two axisymmetric turbojet channels. Therefore, the control quantity of the turbojet channel is treated in a consistent manner. That is, the final output action set is:

[0123] C={Sp Trb ,Sp Ejc ,Wf Trb ,Wf Roc ,Cone Trb ,Conde Ejc} (10)

[0124] The hyper-parameter design of the intelligent controller for the turbojet-ejector mode conversion process is shown in Table 5.

[0125] Table 5. Hyperparameters of the intelligent controller for turbojet-ejector mode conversion

[0126]

[0127] (2) Design of intelligent controller network for injection-subcombustion mode conversion

[0128] Compared with the turbojet-ejector mode conversion process, the ejector-sub-combustion mode conversion process only involves the switching control of the fuel supply system, the action dimension is relatively low, and the state quantity s remains unchanged. The output becomes the ejector rocket fuel supply Wf Roc The fuel supply volume Wf of the secondary combustion chamber Ram The complexity of the action network and the evaluation network is relatively adjusted, specifically by reducing the number of neurons in the hidden layer of the action network and reducing the depth of the action path of the evaluator network.

[0129] The hyper-parameter design of the intelligent controller for the injection-subcombustion mode transition process is shown in Table 6.

[0130] Table 6 Hyperparameters of the intelligent controller for ejection-subcombustion mode transition

[0131]

[0132]

[0133] (3) Network design of intelligent controller for sub-combustion-supercombustion mode conversion

[0134] The scramjet channel of the mobile ejector combined cycle high-speed engine in this paper is in a normally open state, so the scramjet channel does not include the adjustment of the geometric surface. The output of the Actor network includes the combined inlet ejector channel splitter angle Sp Ejc Oil supply to secondary combustion chamber Wf Ram and the fuel supply to the scramjet combustion chamber Wf Soc , and the corresponding common tail nozzle plugging cone stroke Cone Ejc .

[0135] The hyperparameter design of the intelligent controller for the subsonic-scrammatic mode transition process is shown in Table 7.

[0136] Table 7 Hyperparameters of the intelligent controller for subcombustion-supercombustion mode transition

[0137]

[0138] Finally, the reward function is designed. When using reinforcement learning for controller design, the design of the reward function is crucial. The reward value directly affects the evaluation network's fitting of the value function and indirectly affects the update of the action network parameters. The quality of its design has a decisive impact on the final control effect and even the convergence of the algorithm. For the engine mode conversion process, according to the control objectives mentioned above, the thrust tracking performance and engine safety are directly reflected in the reward function. For this purpose, the present invention designs sub-reward functions r err With r uns The final reward function r is composed of these two parts:

[0139] r=r err +r uns (11)

[0140] For thrust tracking, that is, to maintain the actual thrust F generated by the moving ejector combined cycle high-speed engine during the mode conversion process total and thrust command value F dmd Equal, so the thrust error e is included in the reward function Thrust , and perform segmented processing according to the accuracy parameters in the thrust target:

[0141]

[0142] in:

[0143]

[0144] This paper adopts the error reward function in a unified form as shown in Equation (12) for the three modal conversion processes, and only adjusts the specific proportional coefficients. The specific values ​​are shown in Table 8.

[0145] Table 8 Proportional coefficients of the reward function for each mode conversion error of the mobile ejector combined cycle high-speed engine

[0146]

[0147] For the numerical description of the safety of the mobile ejector combined cycle high-speed engine, this paper introduces a unified safety margin ζ expression as follows:

[0148]

[0149] Among them, x is the physical quantity with safety limit in this paper, x max Indicates the maximum value that the physical quantity can reach before entering a dangerous condition. For safety restrictions, we use the reward function r uns Introducing the control barrier function B y (x) to ensure that each state in the control system defines a safe set:

[0150] C:{x∈R n :ζ(x)≥0 (15)

[0151] Control barrier function B y The significance of (x) is that when x approaches the safety boundary, a corresponding penalty is given to adjust the control strategy. Marvi et al. proved that the control barrier function can ensure the safety of the system. y (x) has the following properties:

[0152] (1)B y (x)≤0,

[0153] (2) When the system state is at the safety boundary, the penalty value is very large, that is, ζ→0, B y (x)→∞;

[0154] (3) When the system state is within the safe set, B y (x) decreases monotonically towards the safety margin.

[0155] The control barrier function used in this paper is as follows:

[0156]

[0157] It meets the above requirements for the control barrier function, and setting different parameters γ according to different safety boundary conditions can make a trade-off between safety and controller performance. And set the corresponding proportional coefficient ω according to the penalty intensity i The final reward function for the security part is

[0158]

[0159] According to the safety boundary conditions of each modal conversion process, the specific reward function of the safety part of each modal conversion process is as follows:

[0160] Turbojet-jet mode conversion:

[0161]

[0162] C1={SM HPC >10,T i <3400K} (18)

[0163] Ejection-subcombustion mode conversion:

[0164]

[0165] C2={π ram <π ram * ,T s <3400K} (19)

[0166] Subcombustion-supercombustion mode conversion:

[0167]

[0168] C3={π scr <π scr * ,T r <1200K} (20)

[0169] Table 9 Parameter settings of the safety reward function for each modal conversion process

[0170]

[0171] 4. The above steps design three intelligent agents based on the DDPG algorithm. These three intelligent agents need to be trained to obtain network parameters for subsequent actual control. Therefore, this step designs the intelligent controller training method.

[0172] Traditional controller design often relies on precise mathematical models. However, for mobile ejector combined cycle high-speed engines that integrate high coupling, strong nonlinearity, and MIMO, traditional mathematical modeling methods struggle to achieve sufficiently high accuracy, which also limits the control performance of traditional algorithms. For DDPG reinforcement learning controller design, the initial weights ω and biases b within the neurons of the actor and critic networks are randomly generated and are not suitable for engine control during modal transitions. Therefore, neural network training for three agents is performed for each of the three modal transitions of a mobile ejector combined cycle high-speed engine, and the trained network parameters are saved. This section constructs a parallel-parallel mobile ejector combined cycle high-speed engine simulation model based on the Matlab / Simulink platform and uses the Matlab platform's ReinforceLearning Toolbox to construct the agent. State data generated by the model is directly input into the agent to improve training data accuracy and avoid the cross-platform data interaction issues that arise when building the agent using third-party platforms such as Python.

[0173] The real-time training environment of the present invention consists of two parts: the model end (Environment) and the intelligent agent end (Agent). The overall framework of real-time training is as follows: Figure 11 shown.

[0174] The model side and the intelligent agent side implement the aforementioned single-step update of the intelligent agent network parameters and training optimization through data interaction. The specific interaction process is as follows: First, the network parameters of the intelligent agent are initialized, and then the initial setting function ResetFcn is called. The modal transition point is randomly selected within the range of each modal transition process, and the engine environmental conditions such as altitude and Mach number are determined to determine the thrust target Thtust. dmd The state input s of the mobile ejector combined cycle high-speed engine is obtained by the obtained instructions and initial network parameters. t , will s t Input Actor network to get output action a t , and acts on the engine to execute a step-length simulation program to obtain the next state s t , and obtain the corresponding reward value r through the set reward function t , the data group s generated in this round t ,a t ,r t Stored in the experience pool, the real-time training process of the agent can be divided into control thread and update thread, such as Figure 12 shown.

[0175] 5. Finally, the effectiveness and safety of the control strategy and cooperative control algorithm are verified and tested on the Simulink model of the mobile ejector combined cycle high-speed engine;

[0176] (1) Implementation method based on Reinforcement Learning Toolbox

[0177] The Reinforcement Learning Toolbox (RL Toolbox) is a simplified configuration tool provided by Matlab for training intelligent agents using reinforcement learning algorithms. It provides apps, functions, and Simulink modules for numerous reinforcement learning algorithms, such as DQN, PPO, SAC, and DDPG. The toolbox supports representing policy-value functions using deep neural networks or table lookups, and implements training through interaction with the modeling environment in Matlab or Simulink. It also supports hyperparameter customization, hardware scheduling, and multithreading.

[0178] In addition to the basic state observation value (observation) and reward value (reward), an additional abort input (isdone) is added for early termination of the round. According to simulation results, in the early stages of agent training, the policy network will generate unreasonable action inputs, causing some solvers in the model to crash. To prevent this phenomenon, we generate a high-level input isdone to terminate the round early when the engine thrust output deviates too much from the command value, and add a severe penalty term to the reward function:

[0179] r=r-isdone*M

[0180] M→+∞ (21)

[0181] The specific structure of the interaction between the model and the agent in the Simulink environment is as follows Figure 13 shown.

[0182] For each modal conversion process, this paper created a corresponding training environment in Matlab. The training environment must include the dimensions of the state quantities in each modal conversion process and the upper and lower limits of each state quantity. The specific values ​​are mapped to the range [-1, 1]. The model training step size is uniformly 0.01s. The training process is as follows: Figure 8 The agent network parameters for each modal transition are consistent with those in Tables 5, 6, and 7. The training conditions for each modal transition are as follows. The training condition parameters for the turbojet-to-jet mode transition are shown in Table 10. The training condition parameters for the jet-to-subsonic mode transition are shown in Table 11. The training condition parameters for the subsonic-to-scramjet mode transition are shown in Table 12.

[0183] Table 10 Turbojet-ejector mode conversion training parameters

[0184]

[0185] Table 11 Training parameters for ejection-subcombustion mode conversion

[0186]

[0187]

[0188] Table 12: Subcombustion-supercombustion mode transition training parameters

[0189]

[0190] At this point, a design scheme for establishing a mobile ejector combined cycle high-speed engine was obtained, an engine open-loop simulation Simulink model was obtained, and a modal conversion control strategy for a mobile ejector combined cycle high-speed engine was designed to ensure a safe and smooth transition from turbofan mode, ejector subcombustion mode, subcombustion mode, and scramjet mode, as well as continuous thrust improvement, to avoid the "thrust gap" phenomenon.

[0191] Therefore, this embodiment also provides a mode conversion control method for a mobile ejector combined cycle engine, which includes a turbojet-ejector mode conversion control method, an ejector-subsonic mode conversion control method, and a subsonic-scram mode conversion control method.

[0192] The modal conversion control process at different stages is specifically as follows: obtaining the total thrust generated by the mobile ejector combined cycle engine in real time; when the total thrust reaches the modal conversion point of the corresponding stage, calculating the thrust error and thrust error integral between the total thrust and the corresponding target thrust; using the thrust error, the thrust error integral, and the total thrust as state parameters, using the DDPG algorithm to calculate the action parameters under the state parameters to achieve modal conversion.

[0193] The turbojet-to-ejector mode switch controls the gradual closing of the turbojet inlet and exhaust ducts of the turbofan engine, the gradual opening of the ejector inlet and exhaust ducts of the ejector engine, the gradual cessation of fuel supply to the turbojet engine, and the start of an increase in fuel supply to the ejector engine. The control parameters for the turbojet-to-ejector mode switch include the turbojet channel splitter angle and the turbojet channel plugging cone stroke of the turbofan engine, the ejector channel splitter angle and the ejector channel plugging cone stroke of the ejector engine, the turbojet engine fuel quantity, and the ejector engine fuel quantity. The ejector-to-post-combustion mode switch controls the gradual cessation of fuel supply to the ejector engine and the start of an increase in fuel supply to the post-combustion afterburner. The control parameters for the ejector-to-post-combustion mode switch include the ejector engine fuel quantity and the post-combustion afterburner fuel quantity. The subcombustion-scramjet mode conversion control is to gradually close the ejector intake channel and the ejector exhaust channel of the ejector engine, gradually stop the oil supply to the subcombustion afterburning chamber and start to increase the oil supply to the scramjet combustion chamber; the action parameters of the subcombustion-scramjet mode conversion control include the ejector channel diverter plate angle and the ejector channel blocking cone stroke, the ejector channel blocking cone stroke, the subcombustion afterburning fuel amount, and the scramjet fuel amount of the ejector engine.

[0194] The improvements of this embodiment compared with the existing technology are as follows: ① The innovative concept of a mobile ejector engine is proposed for the first time, which has significant advantages such as high ejection efficiency, continuous adjustment of the intake volume, avoidance of thermal congestion, improved thrust enhancement ratio, improved space utilization, and reduced weight; ② A compact centrally symmetrical nested turbine-based combined cycle high-speed engine is proposed for the first time, which has a wide speed range propulsion capability from 0 to 7 Ma, as well as the characteristics of low flight resistance and large combat radius; ③ A control strategy for a mobile ejector combined cycle high-speed engine is proposed for the first time, especially a multi-input and multi-task intelligent collaborative control algorithm from turbofan mode to ejection sub-combustion mode, which meets a variety of safety limit control tasks and thrust tracking control targets, and ensures that the thrust output from 0 to 7 Ma is safe, stable and effective.

[0195] Compared with the existing technology, the advantages of this embodiment are as follows: ① The movable ejector cone will significantly improve the operating range and efficiency of the subsonic mode ramjet at low Mach numbers, ensuring a smooth transition from the turbofan mode to the subsonic mode and a thrust increase. The movable ejector cone will effectively adjust the inlet flow rate, ensuring a smooth transition from the subsonic mode to the scramjet mode and a thrust increase, achieving the thrust requirement of the combined cycle high-speed engine from 0 to 7 Ma, and can be applied to the power of hypersonic aircraft, with important application value;

[0196] ② The movable ejector cone can adjust the length of the mixing section to avoid thermal congestion in the mixing chamber, improve the thrust enhancement ratio, and ensure the safety and smooth transition of the ejector subcombustion mode. It can be applied to the transition control from turbofan mode to subcombustion mode, and can also be used to adjust the pressure and Mach number at the inlet of the subcombustion engine combustion chamber, showing great development potential.

[0197] ③ The intelligent control method's interaction, based on Matlab's newly released RL Toolbox, enables direct data interaction between the intelligent agent and the mobile ejector combined cycle high-speed engine model on a single platform, without the need for additional mathematical model construction. This demonstrates the advantages of the present invention, namely its independence from the model and its strong practicality.

[0198] ④ During the training of the intelligent agents in each modal transition process, the DDPG algorithm adds random Gaussian noise to the action network during training to enrich the diversity of the experience pool samples. In addition, the present invention designs a random modal transition initial point in the training strategy, and all training and simulation verification processes include environmental interference such as gusts and random effects of engine health parameters, verifying the adaptability and robustness of the intelligent controller;

[0199] ⑤ Through the design of the error sub-reward function and the safety sub-reward function, the present invention achieves the simultaneous guarantee of thrust tracking and safety performance, solving the problem of "losing sight of one thing while focusing on another" in traditional controllers.

[0200] In addition, in this embodiment, a PID controller may be used to control the fuel flow rate of the high-speed flow channel during the mode conversion process and determine the final action based on the state quantity observation.

[0201] Advantages: low model dependency and easy implementation.

[0202] Disadvantages: poor control performance, difficult to ensure safety and control performance at the same time, and poor anti-interference ability.

[0203] In the mode conversion process, the H-based ∞ Multivariable control of a moving ejector combined cycle high-speed engine with loop shaping is conducted. Based on system identification, the state space equations of each mode conversion process of the moving ejector combined cycle high-speed engine are obtained to solve the robust stabilization problem.

[0204] Advantages: It can achieve a balance between safety performance and control performance to a certain extent and has good robustness.

[0205] Disadvantages: The mathematical model of the mobile ejector combined cycle high-speed engine requires high accuracy, but due to the strong nonlinearity and high coupling of the mobile ejector combined cycle high-speed engine, it is difficult to obtain a high-precision mathematical model, and its practicality is poor.

[0206] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A mobile ejector combined cycle engine, characterized in that: include: Cases, turbofan engines, subsonic and scramjets, and mobile ejector engines; The housing includes a movable ejector thrust section, a turbofan thrust section, a sub-combustion and scramjet thrust section, and a common tail nozzle; The mobile ejector thrust section is arranged in a common air inlet passage; the mobile ejector thrust section, the turbofan thrust section and the common tail nozzle are arranged in sequence along the air inlet direction; the subcombustion and scramjet thrust section is sheathed outside the mobile ejector thrust section and the turbofan thrust section; the subcombustion and scramjet thrust section, the mobile ejector thrust section and the turbofan thrust section are connected; the exhaust passages of the turbofan thrust section and the subcombustion and scramjet thrust sections are connected to the common tail nozzle; The mobile ejector engine is arranged in the mobile ejector thrust area; the turbofan engine is arranged in the turbofan thrust area; the sub-combustion and scramjet dual-mode ramjet engine is arranged in the sub-combustion and scramjet thrust area; the sub-combustion and scramjet dual-mode ramjet engine is sleeved on the outside of the turbofan engine; The mobile ejector engine includes a movable ejector cone and an ejector engine arranged in the movable ejector cone; the movable ejector cone is movably connected to the shell; a plurality of jet holes are provided at one end of the movable ejector cone close to the turbofan engine; the jet holes are used to eject the jet generated by the ejector engine; the cone tip of the movable ejector cone is located at the air inlet end; the front end cone body of the movable ejector cone is a closed structure, and the rear end cavity of the movable ejector cone is an incompletely closed structure.

2. The mobile ejector combined cycle engine according to claim 1, characterized in that: The jet holes correspond one to one with the multiple nozzles of the ejector engine.

3. A mode conversion control method for a mobile ejector combined cycle engine according to any one of claims 1 to 2, characterized in that: The method includes a turbojet-jet mode conversion control method; the turbojet-jet mode conversion control method specifically includes: Obtain the total thrust generated by the mobile ejector combined cycle engine, i.e. the first thrust, in real time; When the first thrust reaches a turbojet-ejector mode transition point, calculating a first thrust error and a first thrust error integral between the first thrust and a corresponding target thrust; The first thrust error, the first thrust error integral, and the first thrust are used as state parameters, and a DDPG algorithm is used to calculate first action parameters under the state parameters to gradually close the turbojet inlet and turbojet exhaust channels of the turbofan engine, gradually open the ejector inlet and ejector exhaust channels of the ejector engine, gradually stop the fuel supply to the turbojet engine, and start to increase the fuel supply to the ejector engine; the first action parameters include the turbojet channel splitter angle and turbojet channel plugging cone stroke of the turbofan engine, the ejector channel splitter angle and ejector channel plugging cone stroke of the ejector engine, the turbojet engine fuel quantity, and the ejector engine fuel quantity.

4. A mode conversion control method for a mobile ejector combined cycle engine according to any one of claims 1 to 2, characterized in that: The method includes an ejection-subcombustion mode conversion control method; the ejection-subcombustion mode conversion control method specifically includes: Obtain the total thrust generated by the mobile ejector combined cycle engine, i.e. the second thrust, in real time; When the second thrust reaches the ejection-sub-combustion mode transition point, calculating a second thrust error and a second thrust error integral of the second thrust and the corresponding target thrust; The second thrust error, the second thrust error integral, and the second thrust are used as state parameters, and the DDPG algorithm is used to calculate second action parameters under the state parameters to gradually stop the fuel supply to the ejector engine and start increasing the fuel supply to the sub-combustion afterburner; the second action parameters include the fuel amount of the ejector engine and the sub-combustion afterburner fuel amount.

5. A mode conversion control method for a mobile ejector combined cycle engine according to any one of claims 1 to 2, characterized in that: The method includes a subcombustion-scram combustion mode conversion control method; the subcombustion-scram combustion mode conversion control method specifically includes: Obtain the total thrust generated by the mobile ejector combined cycle engine, i.e. the third thrust, in real time; When the third thrust reaches a subsonic-scram mode transition point, calculating a third thrust error and a third thrust error integral between the third thrust and a corresponding target thrust; The third thrust error, the third thrust error integral, and the third thrust are used as state parameters, and the DDPG algorithm is used to calculate third action parameters under the state parameters to gradually close the ejector engine's ejector intake channel and ejector exhaust channel, gradually stop the fuel supply to the sub-combustion afterburner, and start increasing the fuel supply to the scramjet combustion chamber; the third action parameters include the ejector engine's ejector channel diverter plate angle and ejector channel blocking cone stroke, ejector channel blocking cone stroke, sub-combustion afterburning fuel quantity, and scramjet fuel quantity.

6. A mode conversion control method for a mobile ejector combined cycle engine according to claim 3 or 5, characterized in that: The structure of the action network used in the DDPG algorithm includes: a first input layer, a first hidden layer, a second hidden layer, a third hidden layer and a first output layer connected in series; The structure of the evaluation network used in the DDPG algorithm includes a first state path module, a first action path module, a first connection layer, and a second output layer: The first state path module includes a first state input layer, a first state hidden layer and a second state hidden layer connected in series; The first action path module includes a first action input layer, a first action hidden layer, a second action hidden layer and a third action hidden layer connected in series; The output end of the second state hidden layer and the output end of the third action hidden layer are connected to the first connection layer; and the output end of the first connection layer is connected to the second output layer.

7. The mode conversion control method of a mobile ejector combined cycle engine according to claim 4, characterized in that: The structure of the action network used in the DDPG algorithm includes: a second input layer, a fourth hidden layer, a fifth hidden layer, a sixth hidden layer, and a third output layer connected in series; The structure of the evaluation network used in the DDPG algorithm includes: a second state path module, a second action path module, a second connection layer and a fourth output layer: The second state path module includes a second state input layer, a third state hidden layer and a fourth state hidden layer connected in series; The second action path module includes a second action input layer, a fourth action hidden layer and a fifth action hidden layer connected in series; The output end of the fourth state hidden layer and the output end of the fifth action hidden layer are connected to the second connection layer; and the output end of the second connection layer is connected to the fourth output layer.

8. The mode conversion control method of a mobile ejector combined cycle engine according to claim 3, characterized in that: The expression of the reward function in the DDPG algorithm is: r1＝r err1 +r uns1 ； in, C1={SM HPC >10,T i <3400K}; Where r1 represents the reward function corresponding to the turbojet-jet mode conversion control; r err1 represents the thrust error sub-reward function corresponding to the turbojet-ejector mode conversion control; r uns1 represents the safety sub-reward function corresponding to the turbojet-ejector mode conversion control; e Thurst1 Indicates the first thrust error; dmd1 Indicates the first thrust; a1, b1, γ 11 , γ 12 Represents the modal conversion coefficient; SM HPC Indicates the surge margin of the turbojet engine compressor; T i represents the temperature of the rocket combustion chamber during the turbojet-ejector mode transition; Indicates the maximum value of the turbojet engine compressor surge margin; The maximum temperature of a rocket's combustion chamber.

9. The mode conversion control method of a mobile ejector combined cycle engine according to claim 4, characterized in that: The expression of the reward function in the DDPG algorithm is: r2＝r err2 +r uns2 ； in, C2={π ram <p ram * ,T s <3400K}; Where r2 represents the reward function corresponding to the injection-subcombustion mode conversion control; r err2 represents the thrust error sub-reward function corresponding to the ejection-subcombustion mode transition control; r uns2 represents the safety sub-reward function corresponding to the injection-subcombustion mode transition control; e Thurst2 Indicates the second thrust error; ThurSt dmd2 Indicates the second thrust; a2, b2, γ 21 , γ 22 represents the modal conversion coefficient; π ram Indicates the inlet inlet and outlet back pressure ratio; T s Indicates the temperature of the secondary combustion chamber; Indicates the maximum value of the back pressure ratio between the inlet and outlet of the sub-combustion inlet; Indicates the maximum temperature of the sub-combustion afterburning chamber.

10. The mode conversion control method of a mobile ejector combined cycle engine according to claim 5, characterized in that: The expression of the reward function in the DDPG algorithm is: r3＝r err3 +r uns3 ； in, C3={π scr <p scr * ,T r <1200K}; Where r3 represents the reward function corresponding to the subcombustion-supercombustion mode transition control; r err3 represents the thrust error sub-reward function corresponding to the subsonic-scramjet mode transition control; r uns3 represents the safety reward function corresponding to the subcombustion-supercombustion mode transition control; e Thurst3 Indicates the third thrust error; dmd3 Indicates the third thrust; a3, b3, γ 31 , γ 32 represents the modal conversion coefficient; π scr It represents the back pressure ratio of the scramjet inlet inlet and outlet; Indicates the maximum value of the back pressure ratio of the scramjet inlet inlet and outlet; T r Indicates the scramjet combustion chamber temperature Indicates the maximum temperature of the scramjet combustion chamber.

Images