An engine intake simulation intake control method, system and storage medium

Abstract

The application belongs to the technical field of wind tunnel test, and particularly relates to an air intake control method and system for engine air intake simulation and a storage medium. The air intake control method sets a target flow rate, acquires a measured flow rate, establishes an input / output local time domain linear relationship function based on a pseudo partial derivative, establishes a data updating model based on a Kalman filter, acquires an estimated flow rate at a current time through the data updating model, takes the estimated flow rate at the current time as a starting point, establishes a predicted output trajectory, and acquires a fan speed difference value at the current time based on the predicted output trajectory, so as to adjust the fan speed and make the estimated flow rate at a next time approach the target flow rate. The control method and system can reduce output data fluctuation of the engine air intake simulation system, improve response speed and flow rate control precision.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wind tunnel test, and particularly relates to an air intake control method and system for engine air intake simulation and a storage medium. BACKGROUND

[0002] When an aircraft engine works in icing weather conditions, the air intake components are prone to icing, which causes the engine performance to deteriorate, and even causes the engine to be damaged and shut down, thus seriously endangering flight safety. The Aviation Engine Airworthiness Regulation issued by China has made clear requirements for the icing conditions, anti-icing methods and icing and de-icing tests of the air intake components of the aircraft engine, to ensure that the aircraft can enter the expected natural icing environment and the engine can still operate normally. Under the background of airworthiness certification, carrying out airworthiness verification tests of engine air intake component icing and de-icing is a necessary prerequisite for flight safety.

[0003] At present, the research on the icing problem of the engine air intake components and the verification of the de-icing system are mainly realized through icing wind tunnel tests. The icing wind tunnel test needs to use an engine air intake simulation system to complete air intake simulation. The engine air intake simulation system has a pipeline, a flowmeter and a centrifugal fan arranged in sequence on the pipeline, and the pipeline inlet is connected to the wind tunnel test area. The centrifugal fan sucks the gas to flow through the pipeline, and the flowmeter measures the air intake flow, so that the desired flow acts on the fan to realize engine air intake simulation.

[0004] When the icing wind tunnel test is carried out, the engine air intake simulation system is located in a low-temperature, cloud and mist environment. When the gas is sucked, the spray icing constituting the cloud and mist will cause the inside surface of the pipeline to continuously ice, and its shape will continuously change, changing the air intake flow resistance in the pipeline, so that the control model of the whole system presents dynamic changes, and the change trend is difficult to capture. In addition, the spray will change the gas entering the engine air intake simulation system, affecting the accuracy of the flowmeter in measuring the flow.

[0005] In the traditional scheme, the engine air intake simulation system adopts a PID closed-loop control mode. The dynamic flow data value output by the flowmeter in a certain time is extracted and processed by averaging to obtain the average flow as the PID control feedback. At this time, the traditional PID control scheme has the characteristics of simple principle, programmatic easy implementation and strong universality. However, PID only uses the flow as the feedback in the control process, does not pay attention to the internal state change of the system, and lacks the ability to constrain the control state of the system. In the PID control process, the output data of the system flow control will have a large fluctuation, the response speed will not be timely, and the flow control precision will be poor. SUMMARY

[0006] In order to solve the technical problems of large output data fluctuation, slow response speed and poor flow control precision of the engine intake simulation system in the prior art, the application provides an engine intake simulation air intake control method, system and storage medium, which are used for reducing the output data fluctuation of the engine intake simulation system, improving the response speed and flow control precision. Specifically as follows:

[0007] In the first aspect, the application provides an engine intake simulation air intake control method, which comprises the following steps:

[0008] Setting a target flow ;

[0009] Obtaining a measured flow ;

[0010] Based on the pseudo partial derivative , an input / output local time domain linear relationship function containing the pseudo partial derivative is established, wherein the input term is the fan speed difference , and the output term is the estimated flow ;

[0011] Based on the input / output local time domain linear relationship function and the Kalman filter, a data update model containing the filtering gain is obtained, and based on the fan speed difference at the previous moment and the measured flow at the next moment, the data update model is used to update the estimated flow , the filtering gain and the pseudo partial derivative at the previous moment to the estimated flow , the filtering gain and the pseudo partial derivative at the next moment;

[0012] The estimated flow at the current moment is obtained through the data update model;

[0013] Based on the input / output local time domain linear relationship function, the estimated flow at the current moment is taken as the starting point to establish a prediction output trajectory, and the fan speed difference at the current moment is obtained based on the prediction output trajectory;

[0014] Based on the fan speed difference at the current moment, the fan speed is adjusted so that the estimated flow at the next moment approaches the target flow .

[0015] It should be noted that the fan speed difference at the current moment This refers to the difference between the current fan speed and the previous fan speed. Based on this, the fan speed differences at other times can be calculated. This can be extrapolated from there.

[0016] Optional, based on pseudopartial derivatives Establish a system that includes pseudo-partial derivatives. The input / output local time-domain linear relationship function includes the following steps:

[0017] Establish the input / output nonlinear relationship for a general discrete-time nonlinear system:

[0018] ,

[0019] In the formula, k is the current time, and f(...) is an unknown nonlinear function. It is the estimated flow at time k. It is the fan speed at time k. and For two unknown orders;

[0020] If the input / output nonlinear relationship satisfies the Lipschitz condition, then it is determined that it has a bounded parameter, which is used as a pseudo-partial derivative. And form a local time-domain linear relationship function between input and output:

[0021] ,

[0022] In the formula, k represents the current time, and k+1 represents the time after k. The estimated flow rate at time k to time k+1, For the estimated flow at time k, The difference in fan speed at time k, It is the pseudo-partial derivative at time k.

[0023] Optionally, the update process of the data update model includes the following steps:

[0024] Obtain the estimated flow from time k-1 to time k:

[0025] ,

[0026] Where k is the current time, k-1 is the time before k, and k+1 is the time after k. for Estimated flow rate at any time for The difference in fan speed at any given moment, for The pseudo-partial derivative at time t;

[0027] updating the filter gain based on the Kalman filter to obtain the filter gain at time k ;

[0028] obtaining the estimated flow at time k:

[0029] ,

[0030] obtaining the pseudo partial derivative at time k:

[0031] ,

[0032] wherein, is a step factor, and takes a value in the range of (0, 1]; = ; is a penalty factor, and takes a value greater than 0.

[0033] Optionally, after obtaining the pseudo partial derivative at time k, the method further comprises the step of:

[0034] when or or sign , resetting the pseudo partial derivative , so that .

[0035] Optionally, updating the filter gain based on the Kalman filter comprises the steps of:

[0036] obtaining the estimation error variance at time k-1 for time k:

[0037] ,

[0038] wherein, is the optimal error variance at time k-1, and W is the flow measurement noise variance;

[0039] obtaining the filter gain at time k:

[0040] ,

[0041] wherein, is the flow process noise variance;

[0042] obtaining the optimal error variance at time k:

[0043] .

[0044] Optionally, based on the predicted output trajectory, obtaining the fan speed difference value at the current time ​​comprising steps of:

[0045] The MPC model predictive control method is used to calculate the predicted output trajectory to obtain the fan speed difference value at the current time .

[0046] Optionally, the MPC model predictive control method is used to calculate the predicted output trajectory to obtain the fan speed difference value at the current time comprising steps of:

[0047] The estimated flow at the k time is taken as a starting point, and the output item at the past time is known and the input item The output item from k time to k+N P time is predicted and planned to form a predicted output trajectory, and the predicted output trajectory is made to approach the reference output trajectory with the target flow ;

[0048] The predicted output trajectory is converted into a state space equation:

[0049] ,

[0050] Where x is the output item, and u is the input item;

[0051] The input / output local time domain linear relationship function is compared:

[0052] ,

[0053] At this time, x is equal to , is equal to 1, and B is equal to , is equal to ;

[0054] The MPC model predictive control method is used to calculate the predicted output trajectory to obtain the system optimal input control sequence containing the starting point At this time is the fan speed difference value at the k time .

[0055] Optionally, based on the fan speed difference value at the current time , the fan speed is adjusted, and the estimated flow at the next time approaches the target flow , further comprising steps of:

[0056] The above steps are repeated at future times to obtain the fan speed difference value at future times , and based on the fan speed difference value at future times Real-time adjustment of fan speed to ensure accurate predicted flow rate To target traffic Approaching.

[0057] Secondly, this application provides an intake control system for simulating engine intake, including a pipeline, a flow meter, a centrifugal fan, and a control module. The flow meter and the centrifugal fan are sequentially connected to the pipeline along the air intake direction. The control module is connected to the flow meter and the centrifugal fan. The control module is used to acquire the flow rate of the flow meter and the fan speed of the centrifugal fan, and execute an intake control method for simulating engine intake as described in any of the above schemes.

[0058] Thirdly, this application provides a storage medium storing a computer program that executes the intake control method as described in any of the above schemes.

[0059] The intake control method, system, and storage medium for engine intake simulation according to this application have at least the following advantages compared to the prior art:

[0060] 1. This application constructs a local time-domain linear relationship function between the system input and output based on pseudo-partial derivatives, and combines it with a Kalman filter to construct a data update model, thereby enabling the updating of data based on the difference in wind turbine speed at the previous moment. and the flow rate measured at the next moment Get the estimated traffic at the current moment. Based on this, the estimated flow rate at the current moment is used. Starting from the system's input / output local time-domain linear relationship function, by planning the input terms, the output terms at future times can be predicted and planned. This allows the output trajectory formed by the output terms to approximate the reference output trajectory required for theoretical model planning and control, thereby obtaining the desired predicted output trajectory. Based on the values ​​of the output terms at each time step in this predicted output trajectory, the values ​​of the corresponding input terms at each time step can be quantitatively obtained, thus yielding the value of the input terms at the current time step, i.e., the difference in fan speed at the current time step. The difference in fan speed at the current moment. Adjust the fan speed to predict the flow rate at the next moment. This changes accordingly, and naturally shifts towards the target traffic. Approaching.

[0061] Simultaneously, as the fan speed is adjusted, the measured flow rate at the next moment... This also changes accordingly. At this point, it can continue to be based on the difference in wind turbine speed at the current moment. and the flow rate measured at the next moment The model is updated using data to continue obtaining the estimated flow for the next time step. and continue to establish the predicted output trajectory of the next moment to obtain the fan speed difference value of the next moment , thereby forming a complete control loop for the fan. Through the continuous iteration of the control loop, the fan speed difference value of different moments can be obtained, so that the centrifugal fan of the system can automatically adjust the speed, and in the iteration process, the fan speed difference value is optimized, so that the output item continuously approaches the target flow , achieving the purpose of flow control.

[0062] In summary, the present application can iteratively optimize the fan speed difference value of the centrifugal fan based on the target flow by constructing the input / output local time domain linear relationship function and the data updating model, and quantitatively obtain the fan speed difference value of the centrifugal fan at different moments , so that the system can automatically adjust the speed of the centrifugal fan, and continuously approach the target flow in the pipeline , thereby achieving the purpose of controlling the intake flow of the system.

[0063] In this process, since the fan speed difference value is a parameter quantitatively obtained by the system at each moment for the purpose of approaching the target flow , the speed of the centrifugal fan is adjusted at each moment to approach the target flow , so as to have the constraint ability of the output data, and the response speed is timely, and since the estimated flow always has the trend of approaching the target flow , the fluctuation of the output data is small, and the intake flow is stable.

[0064] 2. The present application can process the measured flow by using the Kalman filtering principle to improve the interference problem, estimate the actual output value of the flow, and obtain a more accurate estimated flow; at the same time, in the data updating model, the estimated flow , filter gain and pseudo partial derivative can be updated in real time based on the two parameters objectively existing in the test process of the fan speed difference value of the previous moment and the measured flow of the next moment, so as to ensure that each parameter at different moments can adapt to the system condition at the moment, thereby ensuring the accuracy of each parameter in the control method, and further ensuring the accuracy of the input item and the output item obtained finally, and solving the problem of poor flow control accuracy. BRIEF DESCRIPTION OF DRAWINGS​

[0065] Figure 1 is a method flow chart of embodiment 1 of the present application;

[0066] Figure 2 is a working schematic diagram of a data updating model in embodiment 1 of the present application;

[0067] Figure 3 is a predicted output trajectory schematic diagram of an MPC at time k in embodiment 1 of the present application;

[0068] Figure 4 is a working schematic diagram of an MPC at time k in embodiment 1 of the present application;

[0069] Figure 5 is a working schematic diagram of a control module of the present application;

[0070] Figure 6 is a control effect diagram obtained by a control method of embodiment 1 of the present application;

[0071] Figure 7 is a system schematic diagram of embodiment 1 of the present application.

[0072] In the figure, the marks are: 100-pipeline, 200-flow meter, 300-centrifugal fan, 400-control module. DETAILED DESCRIPTION

[0073] The specific embodiments of the present application are further specifically described below in combination with the drawings, so as to have further understanding of the concept of the present application, the solved technical problems, the technical features constituting the technical scheme and the brought technical effects. However, it is to be noted that the description of these embodiments is illustrative and does not constitute specific limitation of the present application.

[0074] In order to facilitate the description of the inventive concept of the present application, the engine intake simulation system is briefly described below.

[0075] When the engine intake simulation system is applied to icing wind tunnel test, spray icing can cause data interference and model dynamic change problems, wherein the data interference is that the spray changes the gas entering the engine intake simulation system, affecting the accuracy of the flow meter measuring flow, and the model dynamic change refers to the continuous ice accumulation on the surface of the intake component, the shape is constantly changing, the intake flow resistance changes, and the model of the intake control also presents dynamic change.

[0076] In the traditional scheme, the engine intake simulation system adopts the PID closed-loop control mode of flow mean value feedback, extracts the dynamic flow data value output by the flow meter within a certain time for mean value processing, and obtains the mean flow as the PID control feedback. At this time, the traditional PID control scheme has the characteristics of simple principle, programmatic easy implementation, and strong universality, but PID only uses the flow output data as feedback in the control process, and when the engine intake simulation system sprays, it will affect the accuracy of the flow meter measuring flow, which will cause the system to have deviation and increased fluctuation amplitude when using the flow output data as feedback for control.

[0077] Therefore, in the patent No. 202211244282.0, a stable flow intake control method and system for engine intake simulation and storage medium are disclosed, which calculates the flow measurement noise variance W and the flow process noise variance V, and performs Kalman filtering calculation on the relationship function of the flow Q and the centrifugal fan speed R through the Kalman filter principle, outputs the estimated flow, and then adjusts the speed of the centrifugal fan according to the difference between the estimated flow and the target flow. Thus, the system inputs a more realistic flow deviation feedback, reducing the deviation and fluctuation amplitude of the system in the control process.

[0078] However, the researchers found in their research that for the above scheme, the estimated flow needs to reach a certain threshold (i.e. the target flow) before the speed is adjusted, the whole process is passive, the constraint ability of the output data is weak, there are problems of large output data fluctuation, slow response speed and poor flow control precision. And in practical application, there is a lack of quantitative judgment of speed adjustment, which is difficult to provide reference for the adjustment of system speed.

[0079] Based on this, the present application innovatively proposes an intake control method, system and storage medium for engine intake simulation, which can quantitatively obtain the numerical value of the centrifugal fan speed for adjustment, so as to actively realize the control of the centrifugal fan speed for the purpose of making the estimated flow always approach the target flow, thereby reducing the output data fluctuation of the engine intake simulation system, improving the response speed and flow control precision.

[0080] Embodiment 1

[0081] The present application provides an intake control method for engine intake simulation, as shown in Figure 1 , comprising the steps of:

[0082] S100, setting a target flow ;

[0083] Wherein, the target flow is set according to the experimental requirements.

[0084] S200, obtaining a measured flow rate ;

[0085] The measured flow rate is obtained by a flow meter. During the icing wind tunnel test, the spray will change the gas entering the engine intake simulation system, affecting the accuracy of the measured flow rate of the flow meter, so the measured flow rate exists certain error and cannot be directly regarded as the real entering flow rate through the pipeline in the icing wind tunnel test. The measured flow rate needs to be processed to obtain an estimated value of the real flow rate, and the estimated value is the estimated flow rate .

[0086] S300, establishing an input / output local time domain linear relationship function based on a pseudo partial derivative , wherein the input term is the fan speed difference and the output term is the estimated flow rate ;

[0087] In one or more embodiments, the specific steps of establishing the input / output local time domain linear relationship function include:

[0088] An input / output nonlinear relationship of a general discrete-time nonlinear system of the system is established:

[0089] ,

[0090] In the formula, k is the current time, f(…) is an unknown nonlinear function, is the estimated flow rate at time k, is the fan speed at time k, and are two unknown orders;

[0091] Since the input term and the output term are both observable and controllable in the entire control system, by setting a suitable input term, the corresponding output term can reach a target value, that is, by setting a suitable fan speed difference , the corresponding estimated flow rate can necessarily reach a set value. At this time, the partial derivative of the nonlinear part to the input exists and is continuous, satisfying the Lipschitz condition, wherein the Lipschitz condition is specifically: for any time k1, k2, when k1>0, k2>0, k1≠k2, ≠ , they all have:

[0092] .

[0093] ​Therefore, it can be concluded that there must exist a bounded parameter that can be used as the pseudo-partial derivative at time k. This makes the system represent as:

[0094] .

[0095] This function can be used as an input / output local time-domain linear relationship function.

[0096] In the formula, k represents the current time, and k+1 represents the time after k. The estimated flow rate at time k to time k+1, For the estimated flow at time k, The difference in fan speed at time k, It is the pseudo-partial derivative at time k.

[0097] S400, based on the input / output local time-domain linear relationship function and Kalman filter, obtains the filter gain. The data update model is based on the difference in wind turbine speed at the previous moment. and the flow rate measured at the next moment Data update model such as Figure 2 As shown, this is used to calculate the estimated flow rate from the previous time step. Filter gain and pseudo-partial derivatives Updated to the estimated flow for the next time step. Filter gain and pseudo-partial derivatives ;

[0098] In one or more embodiments, the data update model update process includes the following steps:

[0099] Based on the input / output local time-domain linear relationship function, obtain the estimated flow rate from time k-1 to time k:

[0100] ,

[0101] Where k is the current time, k-1 is the time before k, and k+1 is the time after k. for Estimated flow rate at any time for The difference in fan speed at any given moment, for The pseudo-partial derivative at time t;

[0102] Based on the Kalman filter, the filter gain is adjusted. Update the filter gain at time k. Specifically, it includes the following steps:

[0103] Obtain the estimation error variance of k-1 time to k time:

[0104] ,

[0105] Wherein, is the optimal error variance of k-1 time, and W is the flow measurement noise variance;

[0106] Obtain the filtering gain of k time:

[0107] ,

[0108] Wherein, is the flow process noise variance;

[0109] Obtain the optimal error variance of k time:

[0110] .

[0111] Wherein, the obtaining method of flow measurement noise variance W can be:

[0112] According to the measurement accuracy range [-A, A] of the flow meter and the target flow , the target flow fluctuation range , ] is obtained; B is randomly sampled in the target flow fluctuation range, and the target flow Q is set as the average value, the sample variance of the B samples is calculated, and the flow measurement noise variance W is recorded.

[0113] The obtaining method of flow process noise variance can be:

[0114] Under the set test environment, run the engine intake simulation system; adjust the rotating speed of the centrifugal fan, so that the flow value measured by the flow meter is stable near the target flow ; record the measurement values of multiple flow meters, and calculate the variance based on the measurement values of the multiple flow meters, and record the variance as the flow process noise variance V.

[0115] Obtain the estimated flow of k time:

[0116] ;

[0117] Obtain the pseudo partial derivative of k time:

[0118] ,

[0119] Wherein, is a step factor, and the value is (0, 1]; = ; This is a penalty factor, with a value greater than 0, to avoid algorithmic errors where the denominator is 0.

[0120] Based on this, in order to ensure that the pseudo-partial derivatives always play a role and satisfy... ≠0. After obtaining the pseudo-partial derivative at time k, a reset algorithm is set, which can be:

[0121] when or or sign At that time, with respect to pseudo-partial derivatives Reset to make .

[0122] S500 obtains the estimated flow rate at the current moment through data update model. ;

[0123] S600, based on the input / output local time-domain linear relationship function, uses the estimated flow rate at the current moment. Starting from this point, a predicted output trajectory is established, and the difference in wind turbine speed at the current moment is obtained based on the predicted output trajectory. Specifically, the calculation method can be to use the MPC model predictive control method to calculate the predicted output trajectory in order to obtain the difference in wind turbine speed at the current moment. .

[0124] It should be noted that MPC is a control variable prediction method based on the controlled object model but not bound by the model form. It predicts the future trend of changes based on the system input, output and state, obtains the next optimal control quantity through algorithm processing, and iterates and optimizes it over time.

[0125] Specifically, the steps include:

[0126] like Figure 3 and Figure 4 As shown, the current time k is used as the dividing line, and the estimated flow at time k is used as the dividing line. As a starting point, based on the output items from known past times and input items For the time interval from the current k time to the future k+N time... P The output items within a given time period are used for predictive planning to form a predicted output trajectory, and this predicted output trajectory is made to approximate the target flow. The reference output trajectory;

[0127] Transform the predicted output trajectory into a state-space equation:

[0128] ;

[0129] in For output items, For input items;

[0130] Compare with the input / output local time-domain linear relationship function:

[0131] ,

[0132] at this time, Equivalent to , B equals 1, B equals , Equivalent to ;

[0133] Estimated flow at time k As the starting point for future predictions, based on the state-space equations, the (k+N)th future... P By deriving the output term at time step, we can obtain:

[0134] ,

[0135] in, This refers to the derived value of the output term at time k, since... Equivalent to ,therefore That is, the estimated flow at time k. .

[0136] Therefore, from time k to k+N P The output items between time points, i.e.:

[0137] .

[0138] Based on this, let the output item be The constant term coefficient is M, and the input term is... With coefficient H, we obtain:

[0139] ,

[0140] A cost function between error and input is established using quadratic programming:

[0141] ,

[0142] in, The error is the difference between the output and the target value. , That is, from time k to k+N P Target flow at any time F∈(0,1], is the error weighting matrix, based on the error The influence weight is determined accordingly; ∈(0,1], is the weighted matrix of the input terms, based on the input terms. The influence weight is used to determine the outcome.

[0143] Extending the cost function yields the extended form of the cost function:

[0144] ,

[0145] As can be seen from the extended form of the cost function, it is a function that includes a known starting point. Regarding input items The quadratic function, therefore let The optimal input control sequence of the system can be obtained by solving the problem. ,in, That is, the output item at the current moment. .

[0146] It should be noted that the optimal input control sequence obtained here is based on the prediction and planning of subsequent time steps, derived from the system state at time k. This ideal sequence is based on the system state at time k. However, the system state changes in future time steps, and the trend of these changes is difficult to capture. Therefore, the sequence obtained here, excluding... External control sequence It is not necessarily the optimal input at the corresponding future time. Therefore, when solving for the optimal input at the corresponding future time, the above steps need to be repeated to iteratively optimize the input, thereby ensuring that the optimal input at each time is obtained continuously.

[0147] S700, based on the current turbine speed difference Adjust the fan speed and the estimated flow rate at the next moment. To target traffic Approaching.

[0148] Based on this, the above steps are repeated at various future times to obtain the difference in wind turbine speed at each future time. And based on the difference in wind turbine speed at various future times. Real-time adjustment of fan speed to ensure accurate predicted flow rate To target traffic Approaching.

[0149] At each future time step, the data involved in the control method are iterated based on all the above steps to obtain the parameters at the corresponding time step, thereby obtaining the optimal output term at each future time step. At this time, the centrifugal fan is based on the data at each moment. Adjust the speed.

[0150] like Figure 6As shown, the control method of the application and the PID control method are simulated for 800s of the intake flow, wherein the curve represented by PPD-KF MPC is the simulation curve of the control method of the application, the curve represented by PID is the simulation curve of the PID control method, the target flow is 4.5kg / s, and the whole process is divided into three parts, namely, the pre-spray interval, the spray interval and the post-spray interval, wherein the spray interval is 350s-500s, that is, the spray starts at 350s and ends at 500s. By comparing the two curves, it can be seen that the control of the intake flow realized by the application and the PID control method is obviously better than the latter, and specifically, in the pre-spray interval, the flow of the application directly tends to be stable after reaching the target flow, while the flow under the PID control method has an overshoot in the early stage, even reaching nearly 5.5kg / s, and tends to be stable at nearly 250s, so the time consumed by the application to promote the flow to be stable is much lower than that of the PID control method; in the critical spray interval, the intake is affected by the spray, the disturbance increases, and the model dynamically changes, the curve of the application shows that the flow control fluctuation amplitude is small, which is better than the PID control method, and has better anti-interference and flow control high-precision ability; the post-spray interval is similar to the stable situation after the pre-spray interval, and in the whole process, the curve of the application is closer to the target flow than the curve of the PID, and the data fluctuation is small. In summary, the control of the intake flow of the engine intake simulation system used in the icing wind tunnel test of the application is feasible, and has the characteristics of reducing the output data fluctuation of the engine intake simulation system, anti-interference, flow control precision and the like.

[0151] Embodiment 2

[0152] The embodiment provides an engine intake simulation intake control system, as shown in the figure, comprising a pipeline 100, a flowmeter 200, a centrifugal fan 300 and a control module 400, the flowmeter 200 and the centrifugal fan 300 are connected in sequence along the air inlet direction of the pipeline 100, the control module 400 is connected with the flowmeter 200 and the centrifugal fan 300, the control module 400 is used for acquiring the flow of the flowmeter 200 and the fan speed of the centrifugal fan 300, and executes an engine intake simulation intake control method as described in embodiment 1. Figure 7

[0153] In the process of executing the control method, as shown in the figure, the control module 400 acquires the fan speed difference value Figure 5 at the current time by executing the control method of embodiment 1 , adjusts the speed of the centrifugal fan 300 according to the fan speed difference value , and acquires the measured flow at the next time through the flowmeter 200 . ​

[0154] The present embodiment also provides a storage medium storing a computer program for executing the air intake control method according to the embodiment 1. The storage medium is a computer-readable storage medium, which can be an electronic storage such as a flash memory, an EEPROM (Electrically Erasable Programmable Read-Only Memory), an EPROM, a hard disk or a ROM. Alternatively, the computer-readable storage medium includes a non-transitory computer-readable storage medium. The computer-readable storage medium has a storage space for storing program codes for executing any of the method steps of the above method. The program codes can be read from or written to one or more computer program products. The program codes can be compressed in a suitable form, for example.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than limit the same; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalent ones; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An intake control method for simulating engine intake, characterized in that, Including the following steps: Set target traffic ; Obtain Flow Measurement ; Based on pseudo-partial derivatives Establish a system that includes pseudo-partial derivatives. The input / output local time-domain linear relationship function includes the input term being the difference in fan speed. The output item is the estimated flow rate. ; Based on the input / output local time-domain linear relationship function and the Kalman filter, the filter gain is obtained. The data update model is based on the difference in wind turbine speed at the previous moment. and the measured flow rate at the next moment The data update model is used to update the estimated flow rate from the previous time step. Filter gain and pseudo-partial derivatives Updated to the estimated flow for the next time step. Filter gain and pseudo-partial derivatives ; The estimated flow rate at the current moment is obtained by updating the model with data. ; Based on the input / output local time-domain linear relationship function, the estimated flow rate at the current time is used. Starting from this point, a predicted output trajectory is established, and the difference in wind turbine speed at the current moment is obtained based on the predicted output trajectory. ; Based on the current wind turbine speed difference Adjust the fan speed and the estimated flow rate at the next moment. To target traffic Approaching.

2. The intake control method for engine intake simulation according to claim 1, characterized in that, The aforementioned pseudo-partial derivative Establish a system that includes pseudo-partial derivatives. The input / output local time-domain linear relationship function includes the following steps: Establish the input / output nonlinear relationship for a general discrete-time nonlinear system: ， In the formula, k is the current time, and f(...) is an unknown nonlinear function. It is the estimated flow at time k. It is the fan speed at time k. and For two unknown orders; If the input / output nonlinear relationship satisfies the Lipschitz condition, then it is determined that it has a bounded parameter, which is used as a pseudo-partial derivative. And form a local time-domain linear relationship function between input and output: ， In the formula, k represents the current time, and k+1 represents the time after k. The estimated flow rate at time k to time k+1, For the estimated flow at time k, The difference in fan speed at time k, It is the pseudo-partial derivative at time k.

3. The intake control method for engine intake simulation according to claim 1 or 2, characterized in that, The update process of the data update model includes the following steps: Obtain the estimated flow from time k-1 to time k: ， Where k is the current time, k-1 is the time before k, and k+1 is the time after k. for Estimated flow rate at any time for The difference in fan speed at any given moment, for The pseudo-partial derivative at time t; Based on the Kalman filter, the filter gain is adjusted. Update the filter gain at time k. ; Obtain the estimated traffic at time k: ， Obtain the pseudo-partial derivative at time k: ， in, This is the step size factor, with a value of (0,1]. = ; This is the penalty factor, and its value is >

0.

4. The intake control method for engine intake simulation according to claim 3, characterized in that, After obtaining the pseudo-partial derivative at time k, the following steps are also included: when or or sign At that time, with respect to pseudo-partial derivatives Reset to make .

5. The intake control method for engine intake simulation according to claim 3, characterized in that, Based on the Kalman filter, the filter gain is adjusted. To perform an update, the steps include: Obtain the variance of the estimation error from time k-1 to time k: ， in, Let W be the optimal error variance at time k-1, and W be the variance of the flow measurement noise. Obtain the filter gain at time k: ， in, It is the variance of noise in the flow process; To obtain the optimal error variance at time k: 。 6. The intake control method for engine intake simulation according to claim 3, characterized in that, Based on the predicted output trajectory, the difference in wind turbine speed at the current moment is obtained. Including the following steps: The MPC model predictive control method is used to calculate the predicted output trajectory to obtain the difference in wind turbine speed at the current moment. .

7. The intake control method for engine intake simulation according to claim 6, characterized in that, The MPC model predictive control method is used to calculate the predicted output trajectory to obtain the difference in wind turbine speed at the current moment. The steps include: Estimated flow at time k As a starting point, the output items from past times are known. and input items For time k to k+N P The output items within a given time period are used for predictive planning to form a predicted output trajectory, and this predicted output trajectory is made to approximate the target flow. The reference output trajectory; Transform the predicted output trajectory into a state-space equation: ， Where x is the output term and u is the input term; Compare with the input / output local time-domain linear relationship function: ， At this point, x is equivalent to , B equals 1, B equals , Equivalent to ; The MPC model predictive control method is used to calculate the predicted output trajectory and obtain the optimal input control sequence of the system, including the starting point. ,at this time That is, the difference in fan speed at time k. .

8. The intake control method for engine intake simulation according to claim 1, characterized in that, Based on the difference in wind turbine speed at the current moment Adjust the fan speed and the estimated flow rate at the next moment. To target traffic After approaching, the following steps are also included: Repeat the above steps at various future times to obtain the difference in wind turbine speed at each future time. And based on the difference in wind turbine speed at various future times. Real-time adjustment of fan speed to ensure accurate predicted flow rate To target traffic Approaching.

9. An intake control system for simulating engine intake, characterized in that, The system includes a pipe (100), a flow meter (200), a centrifugal fan (300), and a control module (400). The flow meter (200) and the centrifugal fan (300) are connected sequentially to the pipe (100) along the air intake direction. The control module (400) is connected to the flow meter (200) and the centrifugal fan (300). The control module (400) is used to acquire the flow rate of the flow meter (200) and the fan speed of the centrifugal fan (300), and to execute an intake control method for simulating engine intake as described in any one of claims 1-8.

10. A storage medium, characterized in that, It stores a computer program that performs the intake control method as described in any one of claims 1-8.

Images