A semi-physical simulation method and system based on a prediction model

Abstract

The present application relates to the field of aero-engine simulation technology, more particularly to a semi-physical simulation method and system based on a prediction model.The method comprises: collecting engine test data; identifying a transfer matrix of a state space of the prediction model according to the engine test data, the transfer matrix being used to describe the relationship between an input sequence and an output sequence; calculating matrix coefficients of the prediction model based on the transfer matrix; calculating a predicted value of the output sequence corresponding to the input sequence based on the matrix coefficients and current engine state quantities; and adjusting and updating the operating state of an engine physical system in real time based on the predicted value of the output sequence, so that the engine meets the expected performance.The present application provides high-precision prediction and control for the engine control system by constructing a real-time updated prediction model in combination with actual engine test data.The robustness and adaptability of the control system under fault conditions are enhanced by a fault injection simulation method which simulates various fault types.

Description

Technical Field

[0001] This invention relates to the field of aero-engine simulation technology, and more specifically, to a semi-physical simulation method and system based on a predictive model. Background Technology

[0002] With the continuous advancement of engine technology, the accuracy and real-time performance of engine control systems play a crucial role in their performance and stability. Traditional testing techniques for aero-engine control systems primarily rely on physical models and empirical formulas for modeling and adjustment. However, with the increasing complexity of engine design and operating conditions, traditional testing techniques face numerous challenges, particularly in terms of significant shortcomings in real-time control and accurate prediction.

[0003] Existing technologies typically employ physical model-based full-physics simulation methods or empirical data models to predict engine performance. While these methods can provide some predictive capability, their prediction accuracy is insufficient when faced with complex operating conditions, nonlinear characteristics, and multivariate coupling, and they cannot reflect the engine's dynamic behavior in real time. Furthermore, traditional simulation models are often computationally complex, have poor real-time performance, and are ill-suited for the demands of high-speed dynamic adjustments.

[0004] For real-time adjustment of multivariable aero-engine control systems, existing methods often lack sufficient flexibility and adaptability, resulting in unstable performance of the control system under different operating conditions, which cannot meet the requirements of modern engine control systems for high precision and high real-time performance.

[0005] Semi-physical simulation, also known as physics-mathematical simulation or hardware-in-the-loop simulation, refers to a simulation in which a part of the system being simulated is introduced into the simulation loop as a physical object (or physical model), while the rest of the system is described by a mathematical model and transformed into a simulation computation model. Using physical effect models, real-time mathematical and physical simulations are combined.

[0006] While existing technologies exist for modeling certain components of aero-engines, a comprehensive modeling approach for the entire engine is still lacking. For example, Chinese invention patent CN109470196A discloses a model-based method for evaluating aero-engine blade profile data. This method involves several steps: first, defining the blade profile profile to be evaluated as a line profile profile; then, fitting the line profile profile to a theoretical profile before evaluation; next, thickening the existing blade profile design model and generating a new blade profile using the blade's curve set as the theoretical model for profile profile evaluation. However, the aforementioned patent's technical solution is limited to modeling and analyzing the blade's geometric data, without addressing the comprehensive simulation of the entire engine model.

[0007] Therefore, there is an urgent need for a simulation technology for engine control systems that can solve complex real-time control problems involving multivariable coupling. Summary of the Invention

[0008] The purpose of this invention is to provide a semi-physical simulation method and system based on a predictive model, which solves the problems of poor accuracy and insufficient real-time performance of existing aero-engine control systems in multivariable control.

[0009] To achieve the above objectives, this invention provides a semi-physical simulation method based on a prediction model, comprising the following steps:

[0010] Collect engine test data;

[0011] Based on engine test data, the transfer matrix of the state space of the prediction model is identified, and the transfer matrix is ​​used to describe the relationship between the input sequence and the output sequence.

[0012] Calculate the matrix coefficients of the prediction model based on the transfer matrix;

[0013] Based on the matrix coefficients and the current engine state, the predicted output sequence value corresponding to the input sequence is calculated.

[0014] The operating status of the physical engine system is adjusted and updated in real time based on the predicted values ​​of the output sequence, so that the engine can meet the expected performance.

[0015] In some embodiments, after collecting the engine test data, the process further includes:

[0016] The collected data is preprocessed.

[0017] In some embodiments, the step of identifying the transfer matrix of the predictive model state space based on engine test data further includes:

[0018] Discretize the state-space equations and output the discretized transfer matrix.

[0019] In some embodiments, the discretization method includes a zero-order hold, forward difference, backward difference, and bilinear transform.

[0020] In some embodiments, the matrix coefficients of the prediction model include a first matrix coefficient H and a second matrix coefficient P:

[0021] The calculation method for the coefficients H of the first matrix is ​​as follows:

[0022]

[0023] The calculation method for the coefficient P of the second matrix is ​​as follows:

[0024]

[0025] Among them, A d The first transfer coefficient of the discrete transfer matrix;

[0026] B d The second transfer coefficient of the discrete transfer matrix;

[0027] C d The third transfer coefficient of the discrete transfer matrix;

[0028] Dd is the fourth transfer coefficient of the discretized transfer matrix;

[0029] n y This represents the length of the output sequence.

[0030] In some embodiments, the expression for the predicted output sequence corresponding to the input sequence is:

[0031]

[0032] Where I is the identity matrix, λ is the weight coefficient of the input sequence, and x a Let r be the current engine state variable, r be the expected value of the engine state variable, and k be the discrete index of the current engine state variable.

[0033] In some embodiments, the engine test data includes:

[0034] Sampling time T;

[0035] High-voltage rotor speed N2;

[0036] Low-pressure rotor speed N1;

[0037] Fuel output Wf;

[0038] Turbine exhaust temperature T44.

[0039] In some embodiments, the predicted output sequence is the optimal fuel output.

[0040] The method of adjusting and updating the engine operating status in real time based on the predicted output value to enable the engine to achieve the expected performance further includes:

[0041] Calculate the desired opening of the fuel metering valve based on the optimal fuel output.

[0042] Calculate the current required by the electro-hydraulic servo valve based on the desired opening degree and the current opening degree;

[0043] The calculated current value is output to control the electro-hydraulic servo valve of the fuel metering valve.

[0044] In some embodiments, the method further includes the following steps:

[0045] Select and inject different types of fault signals;

[0046] Identify and adjust the transfer matrix of the state space based on the fault signals;

[0047] The matrix coefficients of the prediction model are updated based on the fault signals;

[0048] Based on the matrix coefficients and the current engine state, the predicted value of the fault response is calculated.

[0049] In some embodiments, the selection and injection of different types of fault signals further includes:

[0050] Different types of fault signals are fused using a fault fusion algorithm;

[0051] The fused fault signals are injected simultaneously.

[0052] To achieve the above objectives, the present invention provides a semi-physical simulation system based on a prediction model, comprising:

[0053] Memory is used to store instructions that can be executed by the processor;

[0054] A processor for executing the instructions to implement the method as described above.

[0055] To achieve the above objectives, the present invention provides a computer-readable medium having computer instructions stored thereon, wherein when the computer instructions are executed by a processor, the method described above is performed.

[0056] This invention provides a semi-physical simulation method and system based on a predictive model. By combining actual engine test data, a real-time updated predictive model is constructed, providing high-precision prediction and control of the engine control system under multivariable coupling and nonlinear dynamic environments. Furthermore, this invention also employs a fault injection simulation method to simulate and fuse multiple fault types, enhancing the robustness and adaptability of the control system under fault conditions, thereby improving the overall stability and performance of the engine system. Attached Figure Description

[0057] The above and other features, properties and advantages of the present invention will become more apparent from the following description taken in conjunction with the accompanying drawings and embodiments, in which the same reference numerals always denote the same features, wherein:

[0058] Figure 1 A flowchart of a semi-physical simulation method based on a prediction model according to an embodiment of the present invention is disclosed;

[0059] Figure 2 A principle block diagram of a semi-physical simulation system based on a prediction model according to an embodiment of the present invention is disclosed. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.

[0061] The semi-physical simulation testing of aero-engine control systems is a crucial support for verifying control algorithms and logic, serving as a loop test for the fuel system-EEC controller. The semi-physical simulation model runs in a real-time simulator, coupling the physical fuel system with the controller to form a closed loop. Model-based semi-physical simulation methods can predict fuel supply and guide vane opening based on the engine's state-space equations and parameter constraints. This model-based prediction approach can effectively validate model-based semi-physical simulation methods within the semi-physical simulation environment.

[0062] Figure 1 A flowchart of a semi-physical simulation method based on a prediction model according to an embodiment of the present invention is disclosed. The semi-physical simulation method based on a prediction model proposed in this invention includes the following steps:

[0063] Step S1: Collect engine test data;

[0064] Step S2: Based on the engine test data, identify the transfer matrix of the state space of the prediction model. The transfer matrix is ​​used to describe the relationship between the input sequence and the output sequence.

[0065] Step S3: Calculate the matrix coefficients of the prediction model based on the transfer matrix;

[0066] Step S4: Based on the matrix coefficients and the current engine state variables, calculate the predicted output sequence value corresponding to the input sequence;

[0067] Step S5: Adjust and update the operating status of the physical engine system in real time based on the predicted values ​​of the output sequence so that the engine meets the expected performance.

[0068] This invention employs a semi-physical simulation method based on a predictive model to predict and simulate an engine model, thereby addressing the problem of real-time model prediction. Within the semi-physical simulation environment, this invention constructs a real-time simulation environment and a real-time predictive model environment. By integrating with a physical fuel system, it achieves real-time multivariable control of the engine model, and the model predictive control algorithm is validated within the semi-physical environment.

[0069] Next, taking the application of the prediction model-based semi-physical simulation method for aero-engine fuel systems as an example, we will illustrate the prediction model-based semi-physical simulation method proposed in this invention.

[0070] Step S1: Collect engine test data.

[0071] In this embodiment, the engine test data originates from information collected and transmitted by the EEC controller when the engine is at idle.

[0072] In this embodiment, the engine test data includes at least: sampling time T, fuel output Wf, high-pressure rotor speed N2, low-pressure rotor speed N1, and turbine exhaust temperature T44.

[0073] The turbine exhaust temperature T44 can also be replaced with other temperature or pressure values.

[0074] Furthermore, preprocessing the collected data can include removing noise, outliers, and standardizing the data to ensure accuracy.

[0075] Step S2: Based on the engine test data, identify the transfer matrix of the state space of the prediction model. The transfer matrix is ​​used to describe the relationship between the input sequence and the output sequence.

[0076] The expression corresponding to the state space of the prediction model is as follows:

[0077]

[0078] y = Cx + D;

[0079] Where x represents the engine state variable, y represents the constraint parameter, u represents the input sequence, A, B, C, and D are the first transfer coefficients of the transfer matrix, B is the second transfer coefficient of the transfer matrix, C is the third transfer coefficient of the transfer matrix, and D is the fourth transfer coefficient of the transfer matrix.

[0080] In this embodiment, the engine state variables include the engine core engine speed, i.e., the high-pressure rotor speed N2, and the low-pressure rotor speed N1, i.e.:

[0081] x = [N1, N2] T ;

[0082] In this embodiment, the limiting parameter is the turbine exhaust temperature T44, which is used as a limiting value in fuel calculation.

[0083] In this embodiment, the input sequence is the fuel output quantity Wf.

[0084] The transmission coefficients A, B, C, and D express the relationship between fuel output and engine speed, as well as the relationship between fuel output and turbine exhaust temperature.

[0085] In this embodiment, identifying the transfer matrix of the prediction model's state space further includes:

[0086] Discretize the state-space equations and output the discretized transfer matrix.

[0087] The expression corresponding to the state space of the prediction model is as follows:

[0088]

[0089] y(k)=C d x(k)+D d u(k)

[0090] Among them, A d The first transfer coefficient of the discrete transfer matrix;

[0091] B d The second transfer coefficient of the discrete transfer matrix;

[0092] C d The third transfer coefficient of the discrete transfer matrix;

[0093] D d It is the fourth transfer coefficient of the discrete transfer matrix.

[0094] The algorithm for the prediction model ultimately runs in an embedded system or a real-time simulator, so it is necessary to convert the continuous state equations into discrete difference equations.

[0095] Alternatively, discretization methods include zero-order hold, forward difference, backward difference, and bilinear transform.

[0096] A zero-order hold keeps the value of each sample point unchanged until the next sample time. Forward differencing uses the difference between the current time and a future time to discretize the system. Backward differencing uses the difference between the current time and a past time to discretize the system. The bilinear transform method converts a continuous system into a discrete system through a specific mapping, preserving stability and frequency response.

[0097] In this embodiment, the forward difference method is used to discretize the engine state variables. The corresponding expression is:

[0098]

[0099] Where T is the sampling time, and k is the discrete index of the current engine state variable x.

[0100] Step S3: Calculate the matrix coefficients of the prediction model based on the transfer matrix;

[0101] The obtained prediction model's matrix coefficients include the first matrix coefficient H and the second matrix coefficient P.

[0102] The calculation method for the coefficients H of the first matrix is ​​as follows:

[0103]

[0104] The calculation method for the coefficient P of the second matrix is ​​as follows:

[0105]

[0106] Among them, A d The first transfer coefficient of the discrete transfer matrix;

[0107] B d The second transfer coefficient of the discrete transfer matrix;

[0108] C d The third transfer coefficient of the discrete transfer matrix;

[0109] D d The fourth transfer coefficient of the discretized transfer matrix;

[0110] n y This represents the length of the output sequence.

[0111] Step S4: Based on the matrix coefficients and the current engine state variables, calculate the predicted output sequence value corresponding to the input sequence.

[0112] The expression for the predicted value of the output sequence corresponding to the input sequence is:

[0113]

[0114] Where I is the identity matrix, λ is the weight coefficient of the input sequence, and x a Let r be the current engine state variable, r be the expected value of the engine state variable, and k be the discrete index of the current engine state variable.

[0115] For example, based on the input sequence such as the current engine speed and fuel quantity, matrix operations are used to calculate the desired output (such as fuel quantity, valve opening, etc.). The output sequence represents the engine's predicted response under given inputs and serves as the basis for the next control calculation.

[0116] The following explains the derivation process of the first matrix coefficient H and the second matrix coefficient P in step S3, as well as the output sequence prediction value in step S4.

[0117] The recurrence equations for the discretized state space are as follows:

[0118] x(k+1)=A dx(k)+B d u(k);

[0119] y(k)=C d x(k)+D d u(k);

[0120] The length of the output sequence used for prediction is n. y Input sequence length n u Assume n y＝ n u The recursive equations for the discretized state space are as follows:

[0121]

[0122]

[0123] Let x^=[x(k+1) x(k+2)…x(k+n) y )] T ,

[0124] y ^ =[y(k+1) y(k+2)…y(k+n) y )] T ,

[0125] By finding the recurrence pattern, we assume...

[0126]

[0127] The calculation method for the first matrix coefficient H and the second matrix coefficient P is as follows:

[0128]

[0129] The calculation of matrices P and H is based on linearization of the theoretical nonlinear model to obtain an initial coefficient matrix. Subsequently, the P and H matrices are corrected using test data. The corrected P and H have higher accuracy, ensuring the accuracy of the prediction model.

[0130] The recursive formula containing ny predicted output sequences is as follows, and this recursive formula will be used to calculate the optimal output sequence:

[0131]

[0132] The recursive equation is transformed into an augmented form, and the output sequence is transformed into an incremental model, which facilitates solving for the optimal output sequence. The corresponding expression is as follows:

[0133]

[0134] To calculate the optimal output sequence, the cost function J is to minimize the predicted output values ​​and fuel output of the ny sequences, and the corresponding expression is:

[0135]

[0136] Where the first term of J represents the sum of prediction errors of each term, and the second term represents the minimum value of J after the input sequence, which yields the optimal fuel input;

[0137] Based on the above formula, we can derive...

[0138]

[0139] Taking the derivative of the cost function J, we get:

[0140]

[0141] The formula is updated online in real time on a real-time simulator, and the input sequence Δu is filtered and limited by fuel parameters.

[0142] By setting the derivative to 0, the expression for the predicted output sequence value corresponding to the input sequence is:

[0143]

[0144] Where I is the identity matrix, λ is the weight coefficient of the input sequence, and x a Let r be the current engine state variable, r be the expected value of the engine state variable, and k be the discrete index of the current engine state variable.

[0145] Step S5: Adjust and update the operating status of the physical engine system in real time based on the predicted values ​​of the output sequence so that the engine meets the expected performance.

[0146] Based on the calculated predicted values, the operating state of the engine system (such as fuel quantity, valve opening, etc.) is adjusted through real-time control algorithms (such as PID control, optimal control, etc.) so that the actual output of the engine is consistent with the predicted values, and the desired state and performance requirements are met.

[0147] Furthermore, step S5 also includes:

[0148] Calculate the desired opening of the fuel metering valve based on the optimal fuel output.

[0149] Calculate the current required by the electro-hydraulic servo valve based on the desired opening degree and the current opening degree;

[0150] The calculated current value is output to control the electro-hydraulic servo valve of the fuel metering valve.

[0151] In this embodiment, the method for calculating the optimal fuel output is as follows:

[0152]

[0153] By looking up the table using the optimal fuel quantity, the relationship between fuel metering flow and valve opening is found, and the corresponding metering valve opening is obtained. The table lookup process is usually based on interpolation or lookup using a pre-calibrated table of flow and opening.

[0154] In this embodiment, the required current for the electro-hydraulic servo valve is calculated using a proportional-integral (PI) control algorithm based on the desired opening degree and the current opening degree. The current opening degree is the actual opening degree converted from the sensitivity data collected by the fuel system.

[0155] In this embodiment, the output current is output to the electro-hydraulic servo valve of the fuel metering valve through a controller or board for flow control.

[0156] To simulate fault injection for various signals in the engine control system, the semi-physical simulation method proposed in this invention further includes the following steps:

[0157] Select and inject different types of fault signals;

[0158] Identify and adjust the transfer matrix of the state space based on the fault signals;

[0159] The matrix coefficients of the prediction model are updated based on the fault signals;

[0160] Based on the matrix coefficients and the current engine state, the predicted value of the fault response is calculated.

[0161] In semi-physical simulation methods, input signals such as temperature, pressure, and speed are typically provided by physical devices or simulation modules. By integrating a fault injection module into the simulation system, fault signals can be dynamically added during simulation. These signals affect the operating state of the engine model during the simulation, thereby testing the response of the engine control system under fault conditions.

[0162] Furthermore, the fault types include position faults, temperature faults, pressure faults, and current faults.

[0163] Based on the characteristics of fault injection, set the fault mode. Common fault modes include: step fault injection, slope fault injection, and pull-off fault injection.

[0164] Based on the fault type and mode, the generated fault signals are injected into the semi-physical simulation system in real time. These signals will affect the system's input or state variables and will also affect the system's state space model (such as A, B, C, and D matrices).

[0165] State space parameters (such as matrices A, B, C, and D) need to be updated and corrected according to the changes in the system state after fault injection.

[0166] Fault signals can cause changes in the dynamic behavior of the system, so it is necessary to update the matrix coefficients (such as the P and H matrices) to reflect the new fault conditions.

[0167] Because fault injection affects the system's state variables, the predictive model needs to perform multivariate control based on the new state-space coefficients and the current system state. The control system dynamically adjusts outputs such as fuel quantity and valve opening to compensate for or correct the effects of the fault.

[0168] The predictive model is updated based on the fault conditions to ensure that it can accurately predict the behavior of the control system under fault conditions. The response of the control system is obtained through real-time simulation detection, and the model and control strategy are optimized based on the feedback.

[0169] Ultimately, through fault injection simulation, the system can adjust the optimal fuel quantity calculation and valve opening control, thereby reflecting the best performance that the system can respond to under fault conditions.

[0170] Furthermore, the selection and injection of different types of fault signals further includes:

[0171] Different types of fault signals are fused using a fault fusion algorithm;

[0172] The fused fault signals are injected simultaneously.

[0173] If multiple signals need to be injected with faults simultaneously, a fault fusion algorithm can be used to fuse different types of faults and inject them synchronously. For example, current faults (such as open circuit, short circuit, overcurrent, etc.) and temperature faults (such as temperature rise, temperature drop, etc.) can be injected simultaneously to simulate complex fault scenarios in which multiple faults occur at the same time. By combining the fault fusion algorithm with the simulation of multi-signal faults, the stability and robustness of the system under various fault conditions can be ensured.

[0174] The fault fusion algorithm selects the corresponding module based on the signal types of more than 20 channels and the fault injection module. It can also fuse different fault types. For example, current signals can be injected with open circuit, short circuit, and overcurrent faults at the same time.

[0175] More specifically, fault fusion algorithms can be implemented using various methods, such as weighting, time series control, and matrix calculation, to fuse multiple fault signals and inject them simultaneously.

[0176] The semi-physical simulation method of the present invention can realize the simulation and verification of fault injection, ensuring that fault injection can affect the simulation model in real time, and test the response and optimization strategy of the control system under fault conditions.

[0177] Although the methods described above are illustrated and depicted as a series of actions for the sake of simplicity, it should be understood and appreciated that these methods are not limited by the order of the actions, as some actions may occur in a different order and / or concurrently with other actions from the illustrations and descriptions herein or not illustrated and described herein but which may be understood by those skilled in the art, according to one or more embodiments.

[0178] The present invention also proposes a semi-physical simulation device based on a prediction model, which can realize the semi-physical simulation method based on the prediction model as described above.

[0179] This invention also proposes a semi-physical simulation device based on a prediction model, comprising, in sequence, a hardware acquisition module, a state space identification module, an expectation calculation module, a matrix coefficient calculation module, an optimal fuel quantity calculation module, an opening degree calculation module, a fuel metering valve displacement calculation module, a fuel metering current calculation module, and an output module.

[0180] The hardware acquisition module collects the frequency of the high-pressure rotor speed, the frequency of the low-pressure rotor speed, and the sensitivity of the fuel metering valve in real time.

[0181] Among them, the high-pressure rotor speed frequency can be calculated from the engine model, the high-pressure rotor speed N2, and then the speed signal is transmitted to the frequency board, and the board outputs the frequency analog sensor acquisition.

[0182] In another embodiment, the high-voltage rotor speed frequency can also be calculated from the engine model to obtain the high-voltage rotor speed N2, which is then transmitted to the electric drive module. The electric drive module controls the rotation of the N2 motor and obtains the motor speed by measuring the actual N2 speed sensor.

[0183] The frequency of the low-pressure rotor speed is similar to that of the high-pressure rotor speed. It can be calculated from the engine model to determine the low-pressure rotor speed N1, and then the speed signal is transmitted to the frequency measurement board, which outputs the frequency analog sensor acquisition. Alternatively, the low-pressure rotor speed N1 can be calculated from the engine model and then transmitted to the electric drive module. The electric drive module controls the rotation of the N1 motor, and the motor speed is obtained by measuring the actual N1 speed sensor.

[0184] In this embodiment, similarly, the frequency of the low-voltage rotor speed is simulated by a board, which represents the low-voltage rotor speed N1.

[0185] The sensitivity of the fuel metering valve is obtained by collecting the Va and Vb voltages of the LVDT (linear displacement sensor) actuator of the fuel metering valve in the fuel metering system circuit, and calculating the sensitivity by the voltage difference and ratio.

[0186] The state space identification module identifies the state space equation of the engine based on the engine test data, and obtains the transfer coefficients (A, B, C, D matrices) of the state space.

[0187] The expectation calculation module calculates the expected values ​​of the state variables based on the state-space model and the current state variables, including the high-pressure rotor speed N2 and the low-pressure rotor speed N1. The expectation calculation module outputs the expectation matrix of the state variables to facilitate further calculation of the optimal fuel quantity and opening degree.

[0188] The matrix coefficient calculation module is used to calculate the first matrix coefficient H and the second matrix coefficient P, which are used to calculate the optimal fuel output.

[0189] The calculation methods for the first matrix coefficient H and the second matrix coefficient P are as follows:

[0190]

[0191] The optimal fuel quantity calculation module calculates the optimal fuel output based on the desired engine speed and the current engine speed.

[0192] More specifically, the inputs of the optimal fuel calculation module are the first matrix coefficient H and the second matrix coefficient P, the current high-pressure rotor speed N2, the low-pressure rotor speed N1, and the output is the fuel output quantity.

[0193] The method for calculating the optimal fuel is as follows:

[0194]

[0195] The opening calculation module takes the optimal fuel level as input and outputs the opening degree of the metering valve.

[0196] The calculation process involves looking up a table using the optimal fuel quantity to determine the relationship between fuel metering flow rate and valve opening, thus deriving the corresponding metering valve opening. This lookup process typically involves interpolation or searching based on a pre-calibrated flow-opening table.

[0197] The fuel metering valve displacement calculation module calculates the actual displacement of the fuel metering valve based on the sensitivity of the fuel metering valve and the calibration relationship between sensitivity and displacement.

[0198] The fuel metering current calculation module is used to calculate the current corresponding to the desired opening degree;

[0199] The inputs to the fuel metering current calculation module include the desired opening degree and the current opening degree. The desired opening degree is calculated by the opening degree calculation module, while the current opening degree is calculated by the fuel metering valve displacement calculation module based on the actual displacement and obtained by conversion using the sensitivity data collected by the fuel system.

[0200] The output module uses a PI control algorithm to perform proportional-integral calculations based on the difference between the desired opening degree and the actual opening degree. The output current signal controls the electro-hydraulic servo valve of the fuel metering valve through a controller or board.

[0201] Furthermore, the semi-physical simulation device based on a predictive model proposed in this invention also includes several fault injection modules. Each fault type (such as temperature fault, location fault, etc.) is encapsulated as an independent module, such as a location fault injection module, a temperature fault injection module, a pressure fault injection module, etc. Each module can handle its specific type of fault injection task, ensuring the system's flexibility and scalability.

[0202] The present invention also proposes a semi-physical simulation system based on a prediction model, which can realize the semi-physical simulation method based on the prediction model described above.

[0203] Figure 2 This is a block diagram of a predictive model-based semi-physical simulation system according to another embodiment of the present invention. The predictive model-based semi-physical simulation system may include an internal communication bus 201, a processor 202, a read-only memory (ROM) 203, a random access memory (RAM) 204, a communication port 205, and a hard disk 207. The internal communication bus 201 enables data communication between components of the predictive model-based semi-physical simulation system. The processor 202 can perform judgments and issue prompts. In some embodiments, the processor 202 may consist of one or more processors.

[0204] Communication port 205 enables data transmission and communication between the predictive model-based semi-physical simulation system and external input / output devices. In some embodiments, the predictive model-based semi-physical simulation system can send and receive information and data from a network via communication port 205. In some embodiments, the predictive model-based semi-physical simulation system can transmit and communicate with external input / output devices via input / output port 206 in a wired manner.

[0205] The semi-physical simulation system based on the predictive model may also include different forms of program storage units and data storage units, such as hard disk 207, read-only memory (ROM) 203, and random access memory (RAM) 204, capable of storing various data files used for computer processing and / or communication, as well as possible program instructions executed by processor 202. Processor 202 executes these instructions to implement the main part of the method. The results processed by processor 202 are transmitted to an external output device via communication port 205 and displayed on the user interface of the output device.

[0206] For example, the implementation process file of the above-mentioned semi-physical simulation method based on prediction model can be a computer program, stored in hard disk 207, and can be loaded into processor 202 for execution to implement the method of this application.

[0207] When the implementation process document of the semi-physical simulation method based on the predictive model is a computer program, it can also be stored as an article of manufacture in a computer-readable storage medium. For example, computer-readable storage media can include, but are not limited to, magnetic storage devices (e.g., hard disks, floppy disks, magnetic stripes), optical discs (e.g., compact discs (CDs), digital multifunction discs (DVDs)), smart cards, and flash memory devices (e.g., electrically erasable programmable read-only memory (EPROM), cards, sticks, key drives). Furthermore, the various storage media described herein can represent one or more devices and / or other machine-readable media used for storing information. The term "machine-readable medium" can include, but is not limited to, wireless channels and various other media (and / or storage media) capable of storing, containing, and / or carrying code and / or instructions and / or data.

[0208] It should be noted that although the semi-physical simulation method, apparatus, and system based on prediction models proposed in this invention are described separately in the embodiments, their contents should be understood as interrelated. For example, if the description of the apparatus includes certain content that is not mentioned in the description of the method, it should be considered that the method implicitly includes the content described in the apparatus.

[0209] This invention provides a semi-physical simulation method and system based on a predictive model. By combining actual engine test data, a real-time updated predictive model is constructed, providing high-precision prediction and control of the engine control system under multivariable coupling and nonlinear dynamic environments. Furthermore, this invention also employs a fault injection simulation method to simulate and fuse multiple fault types, enhancing the robustness and adaptability of the control system under fault conditions, thereby improving the overall stability and performance of the engine system.

[0210] The present invention provides a semi-physical simulation method and system based on a prediction model, which has the following beneficial effects:

[0211] 1) By adopting a semi-physical simulation method and updating the prediction model in real time, the real-time prediction and adjustment of the multivariable state of the engine is realized, which significantly improves the response speed and accuracy of the control.

[0212] 2) Through modular fault injection methods, multiple types of faults can be simulated simultaneously, and fault detection and response can be optimized through fault fusion algorithms to ensure that the engine system can continue to operate stably when a fault occurs;

[0213] 3) By using a state-space identification-based control method combined with real-time feedback, multiple operating parameters of the engine can be precisely adjusted to meet the high-precision control requirements under complex operating conditions.

[0214] As indicated in this application and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" are not specifically singular and may include plural forms. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of explicitly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0215] Those skilled in the art will understand that information, signals, and data can be represented using any of a variety of different techniques and arts. For example, the data, instructions, commands, information, signals, bits, symbols, and chips described throughout the above description can be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, light fields or optical particles, or any combination thereof.

[0216] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in a generalized manner in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the invention.

[0217] The various illustrative logic modules and circuits described in conjunction with the embodiments disclosed herein may be implemented or performed using a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. The general-purpose processor may be a microprocessor, but in alternatives, it may be any conventional processor, controller, microcontroller, or state machine. The processor may also be implemented as a combination of computing devices, such as a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors cooperating with a DSP core, or any other such configuration.

[0218] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of both. The software module may reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to a processor such that the processor can read and write information to / from the storage medium. In an alternative, the storage medium may be integrated into the processor. The processor and storage medium may reside in an ASIC. The ASIC may reside in a user terminal. In an alternative, the processor and storage medium may reside as discrete components in the user terminal.

[0219] The above embodiments are provided for those skilled in the art to implement or use the present invention. Those skilled in the art can make various modifications or changes to the above embodiments without departing from the inventive concept of the present invention. Therefore, the protection scope of the present invention is not limited to the above embodiments, but should be the maximum scope that conforms to the innovative features mentioned in the claims.

Claims

1. A semi-physical simulation method based on a prediction model, characterized in that, Includes the following steps: Collect engine test data; Based on engine test data, the transfer matrix of the state space of the prediction model is identified, and the transfer matrix is ​​used to describe the relationship between the input sequence and the output sequence. Calculate the matrix coefficients of the prediction model based on the transfer matrix; Based on the matrix coefficients and the current engine state, the predicted output sequence value corresponding to the input sequence is calculated. The operating status of the physical engine system is adjusted and updated in real time based on the predicted values ​​of the output sequence, so that the engine can meet the expected performance.

2. The semi-physical simulation method based on a prediction model according to claim 1, characterized in that, After collecting the engine test data, the process further includes: The collected data is preprocessed.

3. The semi-physical simulation method based on a prediction model according to claim 1, characterized in that, The step of identifying the transfer matrix of the prediction model's state space based on engine test data further includes: Discretize the state-space equations and output the discretized transfer matrix.

4. The semi-physical simulation method based on a prediction model according to claim 3, characterized in that, The discretization methods include zero-order hold, forward difference, backward difference, and bilinear transform.

5. The semi-physical simulation method based on a prediction model according to claim 3, characterized in that, The matrix coefficients of the prediction model include a first matrix coefficient H and a second matrix coefficient P: The calculation method for the coefficients H of the first matrix is ​​as follows: The calculation method for the coefficient P of the second matrix is ​​as follows: Among them, A d The first transfer coefficient of the discrete transfer matrix; B d The second transfer coefficient of the discrete transfer matrix; C d The third transfer coefficient of the discrete transfer matrix; D d The fourth transfer coefficient of the discretized transfer matrix; n y This represents the length of the output sequence.

6. The semi-physical simulation method based on a prediction model according to claim 5, characterized in that, The expression for the predicted value of the output sequence corresponding to the input sequence is: Where I is the identity matrix, λ is the weight coefficient of the input sequence, and x a Let r be the current engine state variable, r be the expected value of the engine state variable, and k be the discrete index of the current engine state variable.

7. The semi-physical simulation method based on a prediction model according to claim 1, characterized in that, The engine test data includes at least: Sampling time T; High-voltage rotor speed N2; Low-pressure rotor speed N1; Fuel output Wf.

8. The semi-physical simulation method based on a prediction model according to claim 7, characterized in that, The predicted output sequence is the optimal fuel output. The method of adjusting and updating the engine operating status in real time based on the predicted output value to enable the engine to achieve the expected performance further includes: Calculate the desired opening of the fuel metering valve based on the optimal fuel output. Calculate the current required by the electro-hydraulic servo valve based on the desired opening degree and the current opening degree; The calculated current value is output to control the electro-hydraulic servo valve of the fuel metering valve.

9. The semi-physical simulation method based on a prediction model according to claim 1, characterized in that, It also includes the following steps: Select and inject different types of fault signals; Identify and adjust the transfer matrix of the state space based on the fault signals; The matrix coefficients of the prediction model are updated based on the fault signals; Based on the matrix coefficients and the current engine state, the predicted value of the fault response is calculated.

10. The semi-physical simulation method based on a prediction model according to claim 9, characterized in that, The selection and injection of different types of fault signals further includes: Different types of fault signals are fused using a fault fusion algorithm; The fused fault signals are injected simultaneously.

11. A semi-physical simulation system based on a prediction model, comprising: Memory is used to store instructions that can be executed by the processor; A processor for executing the instructions to implement the method as described in any one of claims 1-10.

12. A computer-readable medium having stored thereon computer instructions, wherein when the computer instructions are executed by a processor, the method as described in any one of claims 1-10 is performed.

Images