Composite loading test system and method applied to solid engine shell

Abstract

The invention discloses a composite loading test system and method applied to a solid engine shell, and belongs to the field of aerospace test devices, and the system comprises an axial loading system, an internal pressure loading system, an external pressure loading system, a temperature rise strain test system and a control system. The system is used for applying axial load, internal pressure, external pressure and high-temperature load to the solid engine shell at the same time. The invention further provides an accurate control method for applying the axial load, the internal pressure and the external pressure. The system has the function of monitoring the stress and strain conditions of the shell in real time, can more comprehensively reflect the mechanical behaviors of the shell under complex working conditions, and provides a more reliable basis for shell structure design and material selection.

Description

Technical Field

[0001] The invention belongs to the field of aerospace test devices, and in particular relates to a composite loading test system and method applied to a solid engine casing. Background Art

[0002] The solid motor is one of the core components of the missile system, responsible for providing propulsion so that the missile can fly along the predetermined trajectory. As the power core of the missile system, its performance and reliability are crucial to the stability and effectiveness of the entire missile system. The solid motor has the advantages of simple structure, stable storage, and fast startup, which make it widely used in the military and aerospace fields. At the same time, the solid motor bears the heavy responsibility of ensuring the success and safety of the mission, so the verification of its reliability and tolerance is particularly important.

[0003] In the testing of solid rocket engines, the use of high temperature or high pressure loading methods alone has obvious limitations. High temperature loading is mainly used to evaluate the thermal stability and heat resistance of materials in high temperature environments, but it ignores the combined effects of high pressure on the material structure, resulting in an inability to fully reflect the mechanical properties during actual work. Although high pressure loading can show the pressure bearing capacity of the shell, it does not take into account the changes in material properties under high temperature conditions and their impact on the overall structural strength. Therefore, testing with a single loading condition is difficult to accurately simulate the complex stress state of solid rocket engines at work, which may lead to an incomplete evaluation of the reliability of materials and structures. Summary of the invention

[0004] In order to solve the problems existing in the prior art, the present invention provides a composite loading test system and method applied to a solid engine casing. The system can monitor the stress and strain conditions of the casing in real time, can more comprehensively reflect the mechanical behavior of the casing under complex working conditions, and provide a more reliable basis for casing structure design and material selection.

[0005] To achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a composite loading test system for a solid motor casing, comprising an axial loading system, an internal pressure loading system, an external pressure loading system, a temperature rise strain testing system and a control system; The axial loading system includes a servo hydraulic station and a loading cylinder; the servo hydraulic station adjusts the flow and pressure of the hydraulic oil in the loading cylinder through a servo valve, and feeds the data back to the control system, thereby controlling the output force of the loading rod and applying the axial load to the housing; The internal pressure loading system and the external pressure loading system include a high-temperature heat transfer oil tank, a high-temperature oil pump, a pressure sensor and a control valve. The two sets of high-temperature oil pumps respectively inject heat transfer oil into the inner cavity of the shell and the annulus formed between the shell and the pressure vessel, and adjust the internal pressure and external pressure of the tested shell through the control valve, and feed back the data to the control system to control the internal pressure and external pressure; The temperature rise strain test system generates a temperature curve by collecting temperature and strain data, and performs synchronous analysis with the strain data.

[0006] As a further improvement of the present invention, the internal pressure loading system injects high-pressure heat transfer oil into the inner cavity of the tested shell through a high-temperature oil pump to simulate the pressurized environment inside the shell when the engine is working; The heat transfer oil used in the internal pressure loading system enters the inner cavity after impurities are removed; a pressure sensor is provided in the inner cavity of the tested shell to monitor the pressure changes in the inner cavity in real time, and dynamically adjust the output flow and pressure of the oil pump through a feedback control loop to control the internal pressure.

[0007] As a further improvement of the present invention, the internal pressure loading system and the external pressure loading system adopt independent dual-circuit control, and independent regulation and coupling control of the internal and external pressures are achieved through a pressure regulating valve.

[0008] As a further improvement of the present invention, the temperature rise strain testing system includes a temperature control cabinet, a heating jacket, a temperature sensor and a strain gauge. The heating jacket is arranged on the outside of the pressure vessel, and the shell is uniformly heated by heat conduction of the heat transfer oil filled in the pressure vessel. The temperature sensor is arranged on the surface of the pressure vessel to monitor the temperature changes on the shell surface in real time. The strain gauge is pasted on the key stress concentration part of the shell to measure the strain of the shell under different load conditions. The data of all sensors are transmitted to the control system through the data acquisition system, and the load and strain are monitored and recorded in real time.

[0009] As a further improvement of the present invention, the control system integrates multi-point pressure sensors and multi-point temperature sensors, which are respectively arranged on the inner and outer surfaces of the shell and in the environment, for real-time monitoring of the internal pressure, external pressure and temperature; Multi-point pressure sensors are arranged on the inner and outer surfaces of the shell under test to monitor the internal and external pressures in real time; multi-point temperature sensors are arranged on the inner and outer surfaces of the shell under test to monitor the temperature in real time; It also includes a fiber Bragg grating sensor: used to monitor the deformation and stress distribution of the tested shell in real time.

[0010] As a further improvement of the present invention, the control system includes a multi-layer neural network optimization controller, which includes a convolutional neural network and a long short-term memory network, performs multi-scale feature extraction and time series prediction on real-time data, and generates optimized control parameters; at the same time, an adaptive fuzzy logic controller is also used, which dynamically adjusts the control strategy based on a multi-layer fuzzy logic algorithm, combining real-time data and historical data; the control system adopts PLC programming logic, has adaptive control capabilities, and dynamically adjusts the loading force and temperature according to sensor feedback.

[0011] As a further improvement of the present invention, the control system also integrates a multi-channel data acquisition and processing system, supports data transmission and storage, supports multi-threading and parallel computing, and supports multiple communication protocols and data encryption; specifically including: Central processing unit, supporting multithreading and parallel computing; Communication module, supporting multiple communication protocols and data encryption functions.

[0012] In a second aspect, the present invention provides a control method based on the composite loading test system applied to a solid motor casing, comprising: S1, accurately place the solid rocket engine shell in the pressure vessel, ensure the sealing and centering between the shell and the container, and confirm that each subsystem is in the best working state; based on the design parameters of the shell and the working environment, set the internal pressure, external pressure, axial load and temperature through the control system; the control system generates and optimizes the synchronous control strategy based on the multi-stress synchronous control algorithm to ensure the coordination and synchronization of the loading process of various physical parameters; calibrate each sensor and actuator; S2, through multi-stress synchronous control technology, accurately coordinates the loading process of temperature rise strain, internal pressure, external pressure and axial load; the high-temperature heating system uses real-time feedback from multi-point high-precision temperature sensors distributed on the shell surface to dynamically adjust the heating power to ensure that the shell surface temperature is strictly distributed according to the preset temperature gradient; the internal and external pressure loading systems use high-precision servo control valves to accurately adjust the flow and pressure of the hydraulic oil to achieve high-precision loading of internal and external pressures, thereby simulating the stress state of the solid rocket engine shell in the actual working environment; the axial load system adjusts the axial load in real time according to the changes in internal and external pressures and the feedback from the strain sensor to ensure the accurate application of the composite load; the control system adopts a closed-loop feedback control mechanism to synchronously adjust the temperature, internal pressure, external pressure and axial load in real time to ensure that all parameters remain coordinated and consistent during dynamic changes; S3, multi-channel data acquisition card synchronously collects axial load, shell inner cavity pressure, shell external pressure, strain and temperature data; the collected data is processed by advanced filtering algorithm, displayed and recorded in real time; a comprehensive test report is generated based on the collected data; abnormal data is identified and prompted.

[0013] As a further improvement of the present invention, the axial load loading specifically includes: S1, select the loading rod stroke, speed and output force as state variables, and establish a reference model based on the axial loading system component parameters:

[0014] in are the loading rod stroke, speed and output force respectively, is the damping coefficient, stiffness and mass of the tested shell after second-order approximation, and is a nonlinear function, To control the input signal of the axial loading servo valve; S2, based on the reference model, the servo valve input signal of the axial loading system is designed. The terminal sliding mode control method is adopted. On the basis of sliding mode control, a nonlinear attractor is introduced. The sliding mode surface is designed as follows:

[0015] in is the error between the output force of the reference model and the expected value, are all hyperparameters, and the reference input signal of the servo valve is obtained for:

[0016] in is the gain coefficient, is the symbolic function, is the expected output force; S3, design an adaptive neural network to provide compensation signal for the servo valve. The compensation signal is expressed as

[0017] in is the force output error between the real loading system and the reference model, is the gain coefficient, is the upper bound of the disturbance, For the neural network Uncertainty estimation, the neural network weights are adaptively updated as the loading process progresses; The final input signal of the servo valve is the reference input signal With compensation signal sum.

[0018] As a further improvement of the present invention, the method for accurately controlling axial load loading, internal pressure loading and external pressure loading specifically includes: S1, precisely placing the solid rocket motor casing in the pressure vessel, and using a distributed sensor network to collect all-round data on key parts of the casing’s inner and outer surfaces;

[0019] in, and They are the multi-point measurement values ​​of the internal pressure and external pressure of the shell to be tested after preloading; S2, preprocess the data

[0020]

[0021] The effect of oil temperature on pressure is introduced, among which, and are the multi-point measurements of the initial temperature; , , , is the preprocessing matrix; S3 uses a combination of convolutional neural networks and long short-term memory networks in a deep learning framework to perform multi-scale feature extraction and time series prediction on real-time collected pressure data to generate optimized control parameters:

[0022]

[0023] Among them, CNN and LSTM are convolutional neural network and long short-term memory network respectively. represents the feature fusion operation; S4, based on fuzzy logic algorithm, dynamically adjusts the control strategy in combination with real-time data; the fuzzy logic controller adopts a hybrid optimization method based on particle swarm optimization algorithm and genetic algorithm to ensure the global optimality of control parameters;

[0024]

[0025] in, and is the control signal output by the adaptive fuzzy logic controller, and is the fuzzy rule membership degree, and is the fuzzy logic parameter; A hybrid optimization algorithm is used to determine the fuzzy logic parameters:

[0026]

[0027] Among them, PSO and GA are particle swarm optimization algorithm and genetic algorithm respectively. represents a mixed optimization operation, is the value range of each optimized parameter; S5, uses multiple redundant actuators with high response speed and performs precise adjustment according to optimized control parameters;

[0028]

[0029] in and To optimize the regulated pressure, and The final control signal generated by the multi-layer neural network optimization controller and the fuzzy logic controller. and is the actuator gain matrix.

[0030] The beneficial effects of the present invention are embodied in: The most significant advantage of the composite loading system proposed in the present invention is that it can simultaneously apply axial load, internal pressure, external pressure and temperature stress to the solid rocket engine casing. This ability to load multiple stresses simultaneously enables the system to highly simulate the multi-dimensional stress state of the casing in the actual working environment, significantly improving the authenticity and reliability of the test. Compared with the traditional single stress loading system, this system has the ability to accurately simulate complex working conditions. It is not only limited to static stress loading, but can also simulate the dynamic stress state of the engine casing under high temperature and high pressure environment. It can accurately reproduce the stress changes of the casing in different stages such as launch, flight and recovery, and comprehensively reflect the mechanical behavior of the casing under complex working conditions, providing a more reliable basis for structural design and material selection.

[0031] In addition, the system can achieve precise control of load application, which not only ensures the synchronization and coordination of multi-stress loading, but also can dynamically adjust the loading force and temperature according to real-time sensor feedback. The combination of multi-layer neural network optimization controller and adaptive fuzzy logic controller improves the prediction accuracy and real-time response capability of the system, ensuring the stability and safety of the test process. Especially in high temperature and high pressure environments, this precise control capability can effectively respond to sudden changes in working conditions and ensure the smooth progress of the test. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a complete structural diagram of the composite loading test system; Figure 2 It is a schematic diagram of the steps for achieving precise control of axial loading; Figure 3 It is a schematic diagram of the steps for achieving precise control of internal and external pressure loading; Figure 4 It is a test flow chart of composite loading system test; Figure 5 It is a schematic diagram of the control system; Figure 6 This is the structure diagram of the neural network used for precise control of axial loading; Figure 7 This is a schematic diagram of the structure of a single LSTM computing unit.

[0033] Among them, 1. Servo hydraulic station; 2. Loading cylinder; 3. PLC; 4. PC; 5. Temperature control cabinet; 6. Heating jacket; 7. Main body bearing frame; 8. Pressure vessel; 9. Loading rod; 10. Test shell; 11. Controller; 12. High-temperature thermal oil tank; 13. High-temperature oil pump. DETAILED DESCRIPTION

[0034] The significance of high temperature and high pressure combined loading is that it can simulate the working environment of solid motors more realistically, thus providing a more comprehensive assessment of material performance and structural strength. This combined testing method can reveal the behavioral characteristics of the shell when it is simultaneously affected by high temperature and high pressure. By reproducing the actual working state of the engine under laboratory conditions, it allows designers to more accurately predict and improve its performance and life under real conditions. In addition, this test can effectively identify potential failure modes, ensure corresponding optimization during the design phase, and further improve the safety and reliability of the missile system. The data obtained through combined loading can provide a basis for material selection, structural optimization and process improvement, promote innovation, and ensure that the design of the new generation of engines is more robust and reliable.

[0035] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0036] The present invention proposes a composite loading test system for solid rocket shells, which specifically includes an axial loading system, an internal pressure loading system, an external pressure loading system, a temperature rise strain test system, and a control system. The complete structural diagram of the test system is shown in the figure. Figure 1 It is used to synchronously apply axial load, internal pressure, external pressure and high temperature load to the solid engine casing and realize precise control of the load, and monitor the stress and strain status of the casing in real time.

[0037] Specifically, the tested shell 10 is placed in a special pressure vessel 8. The design of the pressure vessel 8 fully considers the convenience of entry, exit and connection of the tested shell 10. The two ends of the container are designed to be openable, flanged, and use high-temperature resistant sealing gaskets to ensure that good sealing performance can be maintained after multiple disassembly and assembly. The inner wall of the container is polished to reduce the friction damage that may occur when the shell is entered and exited.

[0038] The axial loading system includes a servo hydraulic station 1, a loading cylinder 2, a loading rod 9 and a main body supporting frame 7, which are used to apply axial pressure to the solid engine casing to evaluate its deformation and strength characteristics under the action of axial load.

[0039] Furthermore, the servo hydraulic station 1 uses a high-response servo valve as a power source, and its closed-loop control system can achieve precise adjustment of the loading force to ensure the force control accuracy and stability during the loading process. The frequency response characteristics of the servo valve have been optimized to meet the needs of different loading rates, thereby meeting a variety of test conditions. The servo hydraulic station 1 adjusts the flow and pressure of the hydraulic oil in the loading cylinder 2 through a high-precision servo valve, and feeds the data back to the control system, thereby accurately controlling the output force of the loading rod 9 and accurately applying the axial load to the shell.

[0040] Among them, the internal pressure loading system and the external pressure loading system are both composed of a high-temperature heat transfer oil tank 12, a high-temperature oil pump 13, a pressure sensor and a control valve. The two groups of high-temperature oil pumps 13 respectively inject heat transfer oil into the inner cavity of the shell and the annulus formed between the shell and the pressure vessel 8. The internal pressure and external pressure of the tested shell 10 are adjusted by the control valve, and the data is fed back to the control system to achieve precise control of the internal pressure and external pressure. The loading cylinder 2 in the axial loading system is connected to the servo hydraulic station 1 through a dedicated high-pressure pipeline. The cylinder has a built-in displacement sensor for real-time monitoring of the piston stroke to ensure the position accuracy during the loading process. The loading rod 9 is made of high-strength alloy material, the surface is chrome-plated, and its wear resistance and corrosion resistance are improved through precision grinding technology. The connection between the loading rod 9 and the pressure vessel 8 adopts high-temperature and high-pressure dynamic sealing technology, specifically using a bidirectional wear-resistant sealing ring to ensure that no leakage occurs during long-term high-pressure testing.

[0041] Specifically, the main bearing frame 7 in the axial loading system is welded from high-strength steel and has undergone stress relief treatment. Its structural design is optimized through finite element analysis and can effectively withstand and disperse high axial loads to avoid additional stress concentration on other components of the system. The fixed support of the frame is designed with an adjustment mechanism that can be flexibly adjusted according to the tested shell 10 of different specifications.

[0042] The internal pressure loading system injects high-pressure heat transfer oil into the inner cavity of the tested housing 10 through the high-temperature oil pump 13 to simulate the pressurized environment inside the housing when the engine is working. The high-temperature oil pump 13 equipped in the system uses a plunger pump with high-temperature resistant and corrosion-resistant materials. The plunger and valve parts inside the pump body are hardened to ensure the service life under high temperature and high pressure conditions.

[0043] Furthermore, the heat transfer oil used in the internal pressure loading system passes through a precision filtration system to remove impurities before entering the inner cavity to prevent the impurities from damaging the inner wall of the shell. A pressure sensor is provided in the inner cavity of the tested shell 10 to monitor the pressure changes in the inner cavity in real time, and dynamically adjust the output flow and pressure of the oil pump through a feedback control loop to achieve precise control of the internal pressure.

[0044] Furthermore, the working principle of the external pressure loading system is similar to that of the internal pressure loading system. The heat transfer oil is injected into the annular outer cavity formed by the tested shell 10 and the pressure vessel 8 through the high-temperature oil pump 13 to simulate the external pressurized environment of the engine shell in actual work.

[0045] As an example, the internal pressure loading system and the external pressure loading system adopt independent dual-circuit control, and realize independent regulation and coupling control of internal and external pressures through high-precision pressure regulating valves. The internal and external pressure loading systems can work simultaneously to perform coupled pressure tests on the shell under complex stress states to evaluate its deformation and strength performance under the combined action of internal and external pressures.

[0046] Specifically, the temperature rise strain test system includes a temperature control cabinet 5, a heating jacket 6, a temperature sensor and a strain gauge. The heating jacket 6 is arranged on the outside of the pressure vessel 8, and the shell is uniformly heated by heat conduction of the heat transfer oil filled in the pressure vessel 8. The temperature sensor is arranged on the surface of the pressure vessel 8 to monitor the temperature change of the shell surface in real time. The strain gauge is pasted on the key stress concentration part of the shell to measure the strain of the shell under different load conditions. The data of all sensors are transmitted to the control system through the data acquisition system to realize real-time monitoring and recording of high-temperature loads and strains.

[0047] As an example, the temperature rise strain test system controls the temperature rise of the tested shell 10 through the heat conduction effect of the heat transfer oil filled in the pressure vessel 8 by means of a heating jacket 6 arranged outside the pressure vessel 8. The heating jacket 6 adopts a specially customized ceramic heater, and its power regulation is realized by a solid-state relay, which can accurately adjust the heating power according to the test requirements, so that the shell surface temperature gradually increases.

[0048] As an example, the temperature rise strain test system is provided with a temperature sensor arranged on the surface of the heating jacket 6. The temperature control cabinet 5 has a built-in PID controller 11, which dynamically adjusts the output power of the heater through the real-time data fed back by the temperature sensor to ensure that the surface temperature of the shell is stable within the set range. The strain gauge is pasted on the key stress concentration part of the shell to measure the strain of the shell under different load conditions. By collecting temperature and strain data, a temperature curve is generated and analyzed synchronously with the strain data to provide data support for evaluating the thermodynamic properties of the shell.

[0049] Furthermore, the control system integrates high-precision multi-point pressure sensors and high-precision multi-point temperature sensors, which are respectively arranged on the inner and outer surfaces of the shell and in the environment for real-time monitoring of internal pressure, external pressure and temperature; fiber Bragg grating (FBG) sensors are used to monitor the deformation and stress distribution of the shell in real time. These sensors are combined with a high-precision data acquisition card to capture transient data in high-temperature and high-pressure environments in real time. The acquisition card has high-speed sampling and high-resolution characteristics to ensure the accuracy and integrity of data under complex working conditions. The collected data is transmitted to the computer through a dedicated interface for real-time monitoring and subsequent analysis to ensure the reliability of the test results.

[0050] As an example, the control system includes a multi-layer neural network optimization controller 11, which includes a convolutional neural network (CNN) and a long short-term memory network (LSTM), and can perform multi-scale feature extraction and time series prediction on real-time data to generate optimized control parameters; at the same time, the system also adopts an adaptive fuzzy logic controller 11, which is based on a multi-layer fuzzy logic algorithm and can dynamically adjust the control strategy to combine real-time data and historical data. The synergy of multi-layer neural networks and adaptive fuzzy logic not only improves the prediction accuracy and real-time response capability of the system, but also enhances its robustness in the face of noise and uncertainty. In addition, the control system adopts optimized PLC 3 programming logic, has adaptive control capabilities, and can dynamically adjust the loading force and temperature according to sensor feedback. This real-time adjustment mechanism ensures the stability and safety of the test process, especially in high temperature and high pressure environments, and can effectively respond to sudden changes in working conditions to ensure the smooth progress of the test.

[0051] In addition, the control system also integrates a multi-channel data acquisition and processing system that supports high-bandwidth, high-precision data transmission and storage; a high-performance central processing unit (CPU) that supports multi-threading and parallel computing; and a high-bandwidth, low-latency communication module that supports multiple communication protocols and data encryption functions. The comprehensive use of these components enables the system to operate efficiently and stably in a complex and changing test environment.

[0052] Among them, the high-precision multi-point pressure sensor is arranged on the inner and outer surfaces of the tested shell 10 to monitor the internal and external pressures in real time; High-precision multi-point temperature sensor: arranged on the inner and outer surfaces of the tested housing 10, used for real-time temperature monitoring; Fiber Bragg Grating (FBG) sensor: used to monitor the deformation and stress distribution of the tested housing 10 in real time; Multi-layer neural network optimization controller 11: including convolutional neural network (CNN) and long short-term memory network (LSTM), which can extract multi-scale features and predict time series of real-time data to generate optimized control parameters; Adaptive fuzzy logic controller 11: Based on multi-layer fuzzy logic algorithm, it can dynamically adjust the control strategy and generate fuzzy logic control signals in combination with real-time data; Multi-channel data acquisition and processing system: supports high-bandwidth, high-precision data transmission and storage; Central Processing Unit (CPU): A high-performance processor that supports multi-threading and parallel computing; Communication module: A high-bandwidth, low-latency communication module that supports multiple communication protocols and data encryption functions.

[0053] As a further improvement of the present invention, a control method for accurately applying the axial load during the combined loading process is designed as follows.

[0054] Since the control equation of the axial loading system process is relatively clear, the model reference adaptive algorithm is used to control the precise loading of the axial load. The overall schematic diagram of the algorithm is shown in Figure 2 The specific steps include: S1, the stroke, speed and output force of the loading rod 9 are selected as state variables, and a reference model is established according to the parameters of the axial loading system components. The reference model is shown in formula (1).

[0055] The nonlinear dynamic equations, the basic laws of fluid mechanics and the dynamic equations considering the compressible fluid characteristics are used to establish a detailed mathematical model for the key components (hydraulic cylinder, servo valve, etc.) in the axial loading system. After obtaining the dynamic equations of these key components, the state space model of the force closed-loop control system is constructed. The state space expression of the axial loading system is:

[0056] in are the stroke, speed and output force of the loading rod 9, are the damping coefficient, stiffness and mass of the tested shell 10 after second-order approximate equivalence, and is a nonlinear function, Input signal to control the axial loading servo valve.

[0057] S2, based on the reference model, the servo valve input signal of the axial loading system is designed, and the terminal sliding mode control method is adopted, that is, a nonlinear attractor is introduced on the basis of traditional sliding mode control to ensure that the system is stable within a limited time. The sliding surface is designed in the form of formula (2).

[0058] Based on the mathematical model established in S1, according to the given desired output force trajectory, the terminal sliding mode control method is used to limit the error convergence time to complete the design of the reference model. The sliding surface expression is designed as:

[0059] in is the error between the reference model output force and the expected value, All are hyperparameters, servo valve reference input signal As shown in formula (3).

[0060]

[0061] in is the gain coefficient, is the symbolic function, is the expected output force.

[0062] S3, design an adaptive neural network to provide compensation signal for the servo valve to eliminate the error between the reference model and the real loading system. The compensation signal expression is shown in formula (4).

[0063] The neural network controller 11 generates a compensation signal which is superimposed on the reference input signal in S1 and acts on the real loading system. The neural network approximation method is used to obtain the uncertainty of the output force error derivative between the reference model and the real system. The neural network output is substituted into the sliding mode control exponential reaching law to obtain the compensation signal calculation formula, which eliminates the tracking error between the real loading system and the reference model, and achieves the purpose of accurately loading the real system according to the predetermined trajectory. The compensation signal expression is shown in formula (4).

[0064]

[0065] in is the force output error between the real loading system and the reference model, is the gain coefficient, is the upper bound of the disturbance, For the neural network Uncertainty estimation. The neural network weights can be adaptively updated as the loading process progresses.

[0066] The final input signal of the servo valve is the reference input signal With compensation signal sum.

[0067] As a further improvement of the present invention, a method for achieving precise control of the internal pressure and the external pressure of the tested housing 10 during the combined loading process is designed as follows.

[0068] Due to the lack of accurate control equations for the shell internal pressure and external pressure loading process, a control system including a high-precision distributed sensor network, a multi-layer neural network optimization controller 11, and an adaptive fuzzy logic algorithm module is designed to complete the accurate and coordinated loading of the internal pressure and external pressure of the tested shell 10.

[0069] Since the loading area inside and outside the shell is large, and the volume of heat transfer oil filled in the inner cavity and the outer annulus is large, there is obvious time lag and nonlinearity in the loading process, and it is difficult to obtain an accurate mathematical expression through traditional methods. Therefore, the present invention adopts a control system including a multi-layer neural network optimization controller 11, an adaptive fuzzy logic controller 11, a high-precision distributed sensor network, a central processing unit and a communication module to achieve real-time monitoring and precise adjustment of the internal and external pressures of the shell, and ensure the stability and reliability of the system under complex working conditions. The overall schematic diagram of the algorithm is shown in the figure. Figure 3 As shown. Specifically including the following steps: S1, the solid rocket engine casing is accurately placed in the pressure vessel 8, and all-round data collection is performed on key parts of the inner and outer surfaces of the casing through a distributed sensor network.

[0070] Shell installation and data acquisition: The solid rocket motor shell is precisely placed in the pressure vessel 8, and comprehensive data collection is performed on the inner and outer surfaces of the shell through a distributed sensor network, including multi-point pressure and temperature measurements to ensure the accuracy of the initial data.

[0071]

[0072] in, and They are the multi-point measurement values ​​of the initial internal pressure and external pressure respectively.

[0073] S2, preprocessing the data: preprocess the collected pressure and temperature data, introduce the effect of oil temperature on pressure, convert the initial data into a multi-dimensional vector that can be used for multi-layer neural network processing, and ensure the consistency and availability of the data.

[0074]

[0075]

[0076] The effect of oil temperature on pressure is introduced, among which, and are the multi-point measurements of the initial temperature; , , , is the preprocessing matrix.

[0077] S3 uses a combination of convolutional neural networks (CNN) and long short-term memory networks (LSTM) in a deep learning framework to perform multi-scale feature extraction and time series prediction on real-time collected pressure data to generate optimized control parameters.

[0078] Multi-layer neural network optimization control: Convolutional neural network (CNN) is used to extract spatial features from pre-processed multi-dimensional vectors, and the time series data of internal and external pressures are processed and predicted through long short-term memory network (LSTM) to generate optimized control parameters. This process combines the spatial feature extraction capability of CNN and the time series prediction capability of LSTM to improve the control accuracy and response speed.

[0079]

[0080]

[0081] Among them, CNN and LSTM are convolutional neural network and long short-term memory network respectively. Represents the feature fusion operation.

[0082] S4, based on fuzzy logic algorithm, dynamically adjusts the control strategy in combination with real-time data to cope with nonlinear pressure changes in complex environments. The fuzzy logic controller 11 adopts a hybrid optimization method based on particle swarm optimization algorithm (PSO) and genetic algorithm (GA) to ensure the global optimality of control parameters.

[0083] Since neural networks are very sensitive to noise in input data, especially in experimental environments, sensor data may have large fluctuations and errors, which will affect the prediction and control effects of the model. Therefore, an adaptive fuzzy logic controller 11 is used as a compensation controller 11 to work together with the multi-layer neural network optimization controller 11 in S3, so that the system can maintain high robustness in the face of uncertainty and noise data.

[0084] The adaptive fuzzy logic controller 11 receives the pre-processed sensor data and dynamically adjusts the control strategy through a multi-layer fuzzy logic algorithm to cope with nonlinear pressure changes in complex environments. The fuzzy logic controller 11 adopts a hybrid optimization method based on a particle swarm optimization algorithm (PSO) and a genetic algorithm (GA) to ensure the global optimality of the control parameters, thereby further improving the stability and reliability of the system.

[0085]

[0086]

[0087] in, and is the control signal output by the adaptive fuzzy logic controller 11, and is the fuzzy rule membership degree, and is the fuzzy logic parameter.

[0088] A hybrid optimization algorithm is used to determine the fuzzy logic parameters:

[0089]

[0090] Among them, PSO and GA are particle swarm optimization algorithm and genetic algorithm respectively. represents a mixed optimization operation, is the value range of each optimized parameter.

[0091] S5, Multiple redundant actuator adjustment: Multiple redundant hydraulic actuators with high response speed are precisely adjusted according to optimized control parameters to ensure instantaneous response and long-term stability of internal and external pressures.

[0092]

[0093]

[0094] in and To optimize the regulated pressure, and The final control signal generated by the multi-layer neural network optimization controller 11 and the fuzzy logic controller 11, and is the actuator gain matrix.

[0095] In addition, the present invention also provides a high temperature and high pressure composite loading test method for a solid rocket engine casing, the method comprising the following main steps: S1, test preparation and parameter setting. First, the solid rocket engine shell is accurately placed in a special pressure vessel 8 to ensure the sealing and centering between the shell and the container, and confirm that each subsystem (including high-temperature heating system, internal and external pressure loading system, and axial loading system) is in the best working state. Based on the design parameters and working environment of the shell, the key test parameters such as internal pressure, external pressure, axial load and temperature are accurately set through the control system. Based on the multi-stress synchronous control algorithm, the control system automatically generates and optimizes the synchronous control strategy to ensure the coordination and synchronization of the loading process of various physical parameters (temperature, pressure, load). At the same time, the system will automatically calibrate each sensor and actuator to ensure that they are in the normal working range to ensure the accuracy and safety of the test.

[0096] S2, synchronous composite loading control. During the loading process, the control system accurately coordinates the loading process of temperature rise strain, internal pressure, external pressure and axial load through multi-stress synchronous control technology. The high-temperature heating system dynamically adjusts the heating power through real-time feedback from multiple high-precision temperature sensors distributed on the shell surface to ensure that the shell surface temperature is strictly distributed according to the preset temperature gradient. The internal and external pressure loading systems accurately adjust the flow and pressure of the hydraulic oil through high-precision servo control valves to achieve high-precision loading of internal and external pressures, thereby simulating the stress state of the solid rocket engine shell in the actual working environment. The axial load system adjusts the axial load in real time according to the changes in internal and external pressures and the feedback from the strain sensor to ensure the accurate application of the composite load. The control system adopts a closed-loop feedback control mechanism to synchronously adjust the temperature, internal pressure, external pressure and axial load in real time to ensure that the parameters are always coordinated and consistent during the dynamic change process, avoiding test errors caused by changes in a single parameter.

[0097] S3, multi-channel data acquisition and real-time monitoring. During the test, the multi-channel high-precision data acquisition card synchronously collects key data such as axial load, shell inner cavity pressure, shell external pressure, strain and temperature. The collected data is processed by advanced filtering algorithms, displayed and recorded in real time to ensure the accuracy and integrity of the data. The control system generates a comprehensive test report based on the collected data. The test personnel can view the changes of various parameters in real time through the human-computer interaction interface and dynamically adjust the test conditions as needed. The system's built-in intelligent algorithm can automatically identify and prompt abnormal data to ensure the safety of the test process and the reliability of the data. In addition, the system also supports real-time storage and backup of data to ensure the long-term preservation and traceability of test data.

[0098] The present invention is described in detail below with reference to specific embodiments.

[0099] Example 1 This embodiment introduces the working process of the composite loading test system. Through the coordinated work of the axial loading system, internal pressure loading system, external pressure loading system, temperature rise strain test system and control system, the precise control and real-time monitoring of the axial load, internal pressure, external pressure and temperature are achieved, and the comprehensive performance test of the solid engine casing under complex stress state is completed. Figure 4 As shown, the specific steps are as follows: 1. Experimental Preparation Before the test begins, first install the solid engine casing to be tested. Open the flanges at both ends of the pressure vessel 8, check the status of the sealing gaskets, and ensure that they are intact. Place the casing stably on the support frame inside the container, and ensure that the casing is aligned with the loading rod 9 to avoid eccentric loading. After tightening the flange of the pressure vessel 8, check the status of each system. Confirm the oil level of the servo hydraulic station 1 to ensure that there is no leakage in the hydraulic pipeline. Check the amount of thermal oil in the high-temperature oil pump 13, and confirm that the thermal oil has been processed by the precision filtration system. Ensure that all sensors (including pressure, temperature, and strain gauges) are connected normally. Check the heating device and temperature control cabinet 5 to ensure that they are in the standby state and the temperature sensor is correctly positioned. After completing these preparations, the system enters the standby state.

[0100] 2. Axial loading operation Start the servo hydraulic station 1, set the initial axial loading force in the control software, usually set to 1.5 times the weight of the shell, and select a suitable loading rate. The servo hydraulic station 1 controls the output force of the loading cylinder 2 through a high-response servo valve and starts to apply axial pressure to the shell. Monitor the data of the built-in displacement sensor of the loading cylinder 2 in real time to ensure that the piston stroke meets the predetermined position and avoid overloading. Gradually increase the axial load according to the test requirements, and the control system adjusts the loading force in real time through closed-loop feedback to ensure force control accuracy.

[0101] 3. Internal pressure loading operation In the internal pressure loading system, the target pressure is set and the pressurization rate is selected. The high temperature oil pump 13 is started to inject the high temperature heat transfer oil that has been precisely filtered into the inner cavity of the shell. The data of the inner cavity pressure sensor is monitored in real time, and the output flow and pressure of the oil pump are dynamically adjusted through the feedback control loop to ensure the stability of the internal pressure.

[0102] 4. External pressure loading operation In the external pressure loading system, set the target pressure and select the pressurization rate. Start another set of high-temperature oil pumps 13 that are different from the internal pressure loading, inject the heat transfer oil into the outer cavity of the annulus between the shell and the pressure vessel 8, monitor the data of the pressure sensor in real time, and control the external pressure through a high-precision pressure regulating valve to ensure pressure stability. The internal and external pressure systems are coupled and controlled through an independent dual-loop control system to ensure synchronous regulation and coordination of internal and external pressures. If the test requires a complex stress state, the control system can realize alternating or synchronous loading of internal and external pressures according to preset parameters to simulate the coupling load under actual working conditions.

[0103] 5. High temperature heating operation Start the temperature rise strain test system, set the target temperature of the shell surface according to the test requirements, and select the heating rate. Start the ceramic heater, adjust the heating power through the solid-state relay, and gradually increase the shell surface temperature. The PID controller 11 in the temperature control cabinet 5 dynamically adjusts the heater output power according to the feedback data of the thermocouple temperature sensor to ensure accurate temperature control. Monitor the temperature changes of the shell surface and the heat transfer oil in real time, and the control software generates a temperature curve and records it synchronously with the strain data. If the temperature exceeds the set range, the system will immediately alarm and automatically reduce the heating power or stop heating to prevent overheating from causing damage to the shell or equipment.

[0104] 6. Data collection and test end operation During the loading and heating process, the strain gauge monitors the tiny deformation of the shell under the action of axial load, internal pressure, external pressure and high temperature in real time. The data is transmitted to the computer through the acquisition card, and the control software generates a stress-strain curve and displays it in real time. The pressure sensors in the inner and outer cavities monitor the pressure changes inside the system in real time. The data is transmitted to the computer through the acquisition card. The control software generates a pressure-time curve and analyzes it synchronously with the temperature data. The control software integrates all the collected data (including strain, pressure, temperature, etc.) and generates a comprehensive test report. Through the control software, the test personnel can view the change trend of each physical quantity in real time and adjust the test parameters as needed, such as increasing or decreasing the axial load, internal pressure, external pressure or temperature. The system automatically records all data to ensure that any changes during the test can be accurately captured for subsequent analysis. The control software will also prompt abnormal data, such as strain values ​​exceeding expectations or large pressure fluctuations, to ensure that the test personnel can take timely countermeasures.

[0105] After the test reaches the predetermined time or loading conditions, the internal and external pressures are gradually reduced, and the load is slowly unloaded through the control system. At the same time, the output power of the heater is gradually reduced to slowly reduce the shell temperature to avoid thermal stress concentration caused by rapid cooling. The temperature control system will automatically adjust through the PID controller 11 according to the preset cooling rate to ensure that the temperature drops evenly. After the internal pressure and temperature of the system are reduced to a safe range, the axial pressure is gradually unloaded through the servo hydraulic station 1, and the control system adjusts the output force of the loading cylinder 2 in real time to ensure a smooth unloading process. When the axial load drops to the initial state, the servo system is turned off and the piston of the loading cylinder 2 returns to its position.

[0106] After the unloading is completed, the data acquisition system is stopped, and all test data (including strain, pressure, temperature, etc.) are saved to the computer hard disk, and the final test report is generated. After checking the status of each system and ensuring that there is no leakage or abnormality, the entire test system is closed. After the system is completely cooled, the flange of the pressure vessel 8 is opened and the tested shell 10 is taken out. The deformation, cracks or other damage of the shell are checked, and compared with the test data to evaluate its performance under complex stress conditions.

[0107] 7. Subsequent analysis After the test is completed, the analysis module of the control software is used to conduct a detailed analysis of the stress-strain curve, temperature-time curve, etc. to evaluate the deformation and strength characteristics of the shell under different load and temperature conditions. By comparing the data under different test conditions, the comprehensive performance of the shell under the combined action of axial pressure, internal pressure, external pressure and high temperature is analyzed. Based on the test data, a detailed test report is generated, including test conditions, loading curves, temperature changes, strain data, etc., to provide data support for subsequent shell design and improvement. If anomalies or problems are found during the test, corresponding improvement suggestions are put forward and provide reference for subsequent tests. The control system structure is as follows: Figure 5 shown.

[0108] Example 2 This embodiment introduces a specific implementation method for controlling the precise loading of the axial load by using a model reference adaptive method, including the following steps: S1, using nonlinear dynamic equations, basic laws of fluid mechanics and dynamic equations considering compressible fluid characteristics, the stroke, speed and output force of the loading rod 9 are selected as state variables , a reference model is established according to the parameters of the axial loading system components, and the reference model is shown in formula (14).

[0109]

[0110] in are the damping coefficient, stiffness and mass after load equivalence. The calculation formula is: When u>0

[0111] When u<0

[0112] The calculation formula is:

[0113] in is the flow area of ​​the servo valve, is the valve core displacement, is the gain coefficient between the input signal and the valve core displacement, is the average density of hydraulic oil before and after the valve, is the cross-sectional area of ​​the high-pressure chamber and the low-pressure chamber, is the elastic modulus of hydraulic oil, is the orifice flow coefficient of the servo valve, is the pump pressure, is the pressure of the chamber directly connected to the pump source, is the corresponding volume, For the volume of another cavity.

[0114] S2, the reference input signal of the servo valve is designed based on the reference model. The design method is to introduce a nonlinear attractor based on the traditional sliding mode control to ensure that the system is stable within a limited time. The sliding surface expression is shown in Equation (18).

[0115]

[0116] in is the error between the output force of the reference model and the expected value, All of them are hyperparameters. The exponential convergence law is used to ensure that s converges to 0 in a finite time, and the system enters the sliding mode:

[0117] Under this design, after the system state point reaches s=0, it converges from the current state e(0)≠0 to e(t s ) = 0 is:

[0118] It is ensured that the error converges to 0 within a limited time, and then the servo valve reference input signal Designed for:

[0119] in is the gain coefficient, is the symbolic function, is the expected output force.

[0120] S3, design an adaptive neural network to provide compensation signals for the servo method to eliminate the error between the reference model and the real loading system. The compensation signal expression is:

[0121] in is the force output error between the real loading system and the reference model, is the gain coefficient, is the upper bound of the disturbance, For the neural network To estimate the uncertainty, the neural network weights can be adaptively updated with the error of the loading process. The specific calculation formula is:

[0122]

[0123]

[0124] in is the neural network input, is the pressure error of the hydraulic cylinder high pressure chamber between the reference model and the real system, is the neural network activation function, and is the adaptive weight, the neural network structure is as follows Figure 6 shown.

[0125] The final input signal of the servo valve of the axial loading system is the reference input signal With compensation signal sum.

[0126] Example 3 Since the goal of internal pressure loading and external pressure loading is to apply uniform pressure to the entire outer and inner walls of the shell, the loading area is large, and the volume of heat transfer oil filled in the inner cavity and the external annulus is large, resulting in obvious time lag and nonlinearity in the process, and it is impossible to obtain a more accurate mathematical expression like axial loading. Therefore, a control system integrating a multi-layer neural network optimization controller 11, an adaptive fuzzy logic controller 11, a high-precision distributed sensor network, a central processing unit (CPU) and a communication module is used to achieve real-time monitoring and precise adjustment of the internal and external pressures of the shell to ensure the stability and reliability of the system under complex working conditions. Specifically, the following steps are included: S1, the solid rocket engine casing is accurately placed in the pressure vessel 8, and all-round and multi-dimensional data collection of the pressure inside and outside the casing is performed through a multi-sensor network.

[0127]

[0128] in, and They are the multi-point measurement values ​​of the initial internal pressure and external pressure respectively.

[0129] S2, preprocess the data.

[0130]

[0131]

[0132] The effect of oil temperature on pressure is introduced, among which, and are the multi-point measurements of the initial temperature; , , , is the preprocessing matrix.

[0133] S3, using convolutional neural network is mainly used to extract spatial features from raw sensor data. The input of convolutional neural network is a preprocessed multidimensional vector:

[0134] The convolutional neural network consists of multiple convolutional layers and pooling layers, and its output can be expressed as

[0135] in represents the convolution operation, and are the weights and biases of the convolutional layer, respectively.

[0136] S4, uses long short-term memory network to process and predict the time-varying sequence data of internal and external pressure. By combining the spatial features extracted by CNN, LSTM can more accurately predict the future system state.

[0137] set up is the output of the LSTM layer, and its calculation process can be expressed as:

[0138] The output of the convolutional neural network is used as the input of the LSTM computing unit. The specific structure is as follows Figure 7 As shown, the output calculation formulas of each part of the LSTM calculation unit are as follows:

[0139]

[0140]

[0141]

[0142]

[0143]

[0144] Optimizing the objective function It can be expressed as:

[0145] in and is the set target pressure.

[0146] The optimization algorithm updates the weights via gradient descent:

[0147] in, is the learning rate, is the gradient of the objective function.

[0148] The output layer generates optimized control parameters and :

[0149] S5, using fuzzy logic controller 11 to provide compensation signal. Since the multi-layer neural network controller 11 in S4 is very sensitive to the noise of input data, especially in the test environment, the sensor data may have large fluctuations and errors, which will affect the prediction and control effect of the model. Therefore, an adaptive fuzzy logic controller 11 is used as a compensation controller 11 to work together with the multi-layer neural network optimization controller 11 in S3, so that the system can maintain high robustness in the face of uncertainty and noise data.

[0150] The adaptive fuzzy logic controller 11 receives the pre-processed sensor data. For each sensor, the input provided to the fuzzy logic controller 11 is:

[0151] Convert the input data into fuzzy sets. and Respectively represent the internal and external pressure on The function value of the membership function of the fuzzy rules is:

[0152]

[0153] in , , and For the The parameters to be optimized of the membership function of the fuzzy rules are then optimized through the hybrid optimization module of particle swarm optimization algorithm (PSO) and genetic algorithm (GA).

[0154]

[0155]

[0156] in Indicates the value range of each parameter to be optimized.

[0157] The optimization objective function can be expressed as:

[0158] According to the fuzzified input data, fuzzy rules are applied to generate fuzzy output. and Corresponding to the The control parameters of internal and external pressures obtained from the sensor data:

[0159]

[0160] By averaging the control parameters obtained from all sensors, the parameters of the fuzzy logic controller 11 for the final internal and external pressure can be written as:

[0161]

[0162] The combination of CNN and LSTM multi-layer neural network optimization control and adaptive fuzzy logic control not only improves the system's prediction accuracy and real-time response capability, but also enhances its robustness in the face of noise and uncertainty, and performs well in handling complex working conditions and multi-objective optimization, achieving more comprehensive, precise and reliable control effects.

Claims

1. A composite loading test system for solid motor casing, characterized in that: Including axial loading system, internal pressure loading system, external pressure loading system, temperature rise strain testing system and control system; The axial loading system includes a servo hydraulic station and a loading cylinder; the servo hydraulic station adjusts the flow and pressure of the hydraulic oil in the loading cylinder through a servo valve, and feeds the data back to the control system, thereby controlling the output force of the loading rod and applying the axial load to the housing; The internal pressure loading system and the external pressure loading system include a high-temperature heat transfer oil tank, a high-temperature oil pump, a pressure sensor and a control valve. The two sets of high-temperature oil pumps respectively inject heat transfer oil into the inner cavity of the shell and the annulus formed between the shell and the pressure vessel, and adjust the internal pressure and external pressure of the tested shell through the control valve, and feed back the data to the control system to control the internal pressure and external pressure; The temperature rise strain test system generates a temperature curve by collecting temperature and strain data, and performs synchronous analysis with the strain data.

2. A composite loading test system for solid motor casing according to claim 1, characterized in that: The internal pressure loading system injects high-pressure heat transfer oil into the inner cavity of the tested shell through a high-temperature oil pump to simulate the pressurized environment inside the shell when the engine is working; The heat transfer oil used in the internal pressure loading system enters the inner cavity after impurities are removed; a pressure sensor is provided in the inner cavity of the tested shell to monitor the pressure changes in the inner cavity in real time, and dynamically adjust the output flow and pressure of the oil pump through a feedback control loop to control the internal pressure.

3. A composite loading test system for solid motor casing according to claim 1, characterized in that: The internal pressure loading system and the external pressure loading system adopt independent dual-circuit control, and independent regulation and coupling control of internal and external pressures are achieved through a pressure regulating valve.

4. A composite loading test system for solid motor casing according to claim 1, characterized in that: The temperature rise strain test system includes a temperature control cabinet, a heating jacket, a temperature sensor and a strain gauge. The heating jacket is arranged on the outside of the pressure vessel, and the shell is uniformly heated by heat conduction of the heat transfer oil filled in the pressure vessel. The temperature sensor is arranged on the surface of the pressure vessel to monitor the temperature change of the shell surface in real time. The strain gauge is pasted on the key stress concentration part of the shell to measure the strain of the shell under different load conditions. The data of all sensors are transmitted to the control system through the data acquisition system, and the load and strain are monitored and recorded in real time.

5. The composite loading test system for solid motor casing according to claim 1, characterized in that: The control system integrates multiple pressure sensors and multiple temperature sensors, which are arranged on the inner and outer surfaces of the shell and in the environment, respectively, for real-time monitoring of internal pressure, external pressure and temperature; Multi-point pressure sensors are arranged on the inner and outer surfaces of the tested shell to monitor the internal and external pressures in real time; Multi-point temperature sensors are arranged on the inner and outer surfaces of the tested shell to monitor the temperature in real time; It also includes a fiber Bragg grating sensor: used to monitor the deformation and stress distribution of the tested shell in real time.

6. The composite loading test system for solid motor casing according to claim 1, characterized in that: The control system includes a multi-layer neural network optimization controller, which contains a convolutional neural network and a long short-term memory network. The controller performs multi-scale feature extraction and time series prediction on real-time data to generate optimized control parameters. At the same time, an adaptive fuzzy logic controller is also used to dynamically adjust the control strategy based on a multi-layer fuzzy logic algorithm, combining real-time data and historical data. The control system adopts PLC programming logic and has adaptive control capabilities, dynamically adjusting the loading force and temperature according to sensor feedback.

7. The composite loading test system for solid motor casing according to claim 1, characterized in that: The control system also integrates a multi-channel data acquisition and processing system, supports data transmission and storage, supports multi-threading and parallel computing, and supports multiple communication protocols and data encryption; specifically including: Central processing unit, supporting multithreading and parallel computing; Communication module, supporting multiple communication protocols and data encryption functions.

8. A control method, based on the composite loading test system for solid motor casing according to any one of claims 1 to 7, characterized in that: include: S1, accurately place the solid rocket engine shell in the pressure vessel, ensure the sealing and centering between the shell and the container, and confirm that each subsystem is in the best working state; based on the design parameters of the shell and the working environment, set the internal pressure, external pressure, axial load and temperature through the control system; the control system generates and optimizes the synchronous control strategy based on the multi-stress synchronous control algorithm to ensure the coordination and synchronization of the loading process of various physical parameters; calibrate each sensor and actuator; S2, through multi-stress synchronous control technology, accurately coordinates the loading process of temperature rise strain, internal pressure, external pressure and axial load; The high-temperature heating system uses real-time feedback from high-precision temperature sensors distributed on the shell surface to dynamically adjust the heating power to ensure that the shell surface temperature is strictly distributed according to the preset temperature gradient. The internal and external pressure loading systems use high-precision servo control valves to accurately adjust the flow and pressure of the hydraulic oil to achieve high-precision loading of internal and external pressures, thereby simulating the stress state of the solid rocket engine shell in the actual working environment. The axial load system adjusts the axial load in real time according to the changes in internal and external pressures and the feedback from the strain sensor to ensure the accurate application of the composite load. The control system adopts a closed-loop feedback control mechanism to synchronously adjust the temperature, internal pressure, external pressure and axial load in real time. S3, multi-channel data acquisition card synchronously collects axial load, shell inner cavity pressure, shell external pressure, strain and temperature data; the collected data is processed by advanced filtering algorithm, displayed and recorded in real time; a comprehensive test report is generated based on the collected data; abnormal data is identified and prompted.

9. A control method according to claim 8, characterized in that: Axial load loading, including: S1, select the loading rod stroke, speed and output force as state variables, and establish a reference model based on the axial loading system component parameters: in are the loading rod stroke, speed and output force respectively, is the damping coefficient, stiffness and mass of the tested shell after second-order approximation, and is a nonlinear function, To control the input signal of the axial loading servo valve; S2, based on the reference model, the servo valve input signal of the axial loading system is designed. The terminal sliding mode control method is adopted. On the basis of sliding mode control, a nonlinear attractor is introduced. The sliding mode surface is designed as follows: in is the error between the reference model output force and the expected value, are all hyperparameters, and the reference input signal of the servo valve is obtained for: in is the gain coefficient, is the symbolic function, is the expected output force; S3, design an adaptive neural network to provide compensation signal for the servo valve. The compensation signal is expressed as in is the force output error between the real loading system and the reference model, is the gain coefficient, is the upper bound of the disturbance, For the neural network Uncertainty estimation, the neural network weights are adaptively updated as the loading process progresses; The final input signal of the servo valve is the reference input signal With compensation signal sum.

10. A control method according to claim 8, characterized in that: Axial load loading, internal pressure loading, and external pressure loading precise control methods include: S1, precisely placing the solid rocket motor casing in the pressure vessel, and using a distributed sensor network to collect all-round data on key parts of the casing’s inner and outer surfaces; in, and They are the multi-point measurement values ​​of the internal pressure and external pressure of the shell to be tested after preloading; S2, preprocess the data The effect of oil temperature on pressure is introduced, among which, and are the multi-point measurements of the initial temperature; , , , is the preprocessing matrix; S3 uses a combination of convolutional neural networks and long short-term memory networks in a deep learning framework to perform multi-scale feature extraction and time series prediction on real-time collected pressure data to generate optimized control parameters: Among them, CNN and LSTM are convolutional neural network and long short-term memory network respectively. represents the feature fusion operation; S4, based on fuzzy logic algorithm, dynamically adjusts the control strategy in combination with real-time data; the fuzzy logic controller adopts a hybrid optimization method based on particle swarm optimization algorithm and genetic algorithm to ensure the global optimality of control parameters; in, and is the control signal output by the adaptive fuzzy logic controller, and is the fuzzy rule membership degree, and is the fuzzy logic parameter; A hybrid optimization algorithm is used to determine the fuzzy logic parameters: Among them, PSO and GA are particle swarm optimization algorithm and genetic algorithm respectively. represents a mixed optimization operation, is the value range of each optimized parameter; S5, uses multiple redundant actuators with high response speed and performs precise adjustment according to optimized control parameters; in and To optimize the regulated pressure, and The final control signal generated by the multi-layer neural network optimization controller and the fuzzy logic controller. and is the actuator gain matrix.

Images