Batch evaluation method for unmanned aerial vehicle control law parameters

Abstract

The invention discloses a batch evaluation method for control law parameters of an unmanned aerial vehicle, and the method comprises the following steps: building a six-degree-of-freedom mathematical model of the unmanned aerial vehicle, and obtaining a state-space equation of each working point; establishing a control law design model; according to the control law design model, obtaining control law parameters; and establishing a control law parameter evaluation model, importing the state-space equation obtained in the step 1 and the control law parameters obtained in the step 3, and performing batch evaluation on the control law parameters through an evaluation script. Compared with the prior art, the process control law parameter evaluation method is designed, the problems that the control law parameter evaluation work of the unmanned aerial vehicle flight control system in engineering design is high in repeatability and low in efficiency are solved, batch control law parameter evaluation is achieved, and the efficiency and quality of control law design and parameter adjustment are improved.

Description

Technical Field

[0001] The invention belongs to the field of unmanned aerial vehicle control law design and simulation, and in particular to a batch evaluation method for unmanned aerial vehicle control law parameters.

[0002] The application of the present invention can avoid the defects of repeated work, low work efficiency, inconvenient global evaluation and comparison in the current control law parameter evaluation process, and improve the efficiency and quality of the global evaluation of UAV control law parameters. Background Art

[0003] UAV control is a necessary condition for stable flight of UAV. The quality of control law parameter design is directly related to whether the UAV can maintain stable attitude and track trajectory as required. Therefore, in order to ensure flight safety, it is usually necessary to evaluate the designed control law parameters.

[0004] The present invention relates to the field of unmanned aerial vehicle control law design and simulation, and can be applied to the evaluation of unmanned aerial vehicle control law parameters.

[0005] In the prior art, the usual approach to evaluating the control law parameters of a UAV flight control system is to select several operating points in the full flight envelope, perform nonlinear model linearization and control law design, and then evaluate the control law parameters at different operating points by means of step response, Bode plot, etc. This method has the following problems:

[0006] Problem 1: For fixed-wing UAVs with large flight envelopes, many working points need to be selected to fully design the global control law parameters. For such control law design requirements with many working points, a lot of repetitive work is required to evaluate the control law parameters, that is, a parameter evaluation is performed for each working point, and it is necessary to repeatedly set model parameters, import linear state space equations, import control parameters, run analysis functions, run drawing functions, etc., and then manually splice multiple evaluation images and data, which greatly reduces the efficiency of control law parameter adjustment;

[0007] Problem 2: Because each operating point is evaluated separately, the control law parameter evaluation is not obvious in the global comparison, and the comparison work is not convenient. If you want to compare the control law effect in the global space, you need to manually splice the data;

[0008] Question 3: The comparison between the control law parameter evaluation under nominal and rudder pull disturbance is not obvious. To compare the changes in control effect under nominal and rudder pull disturbance, it is necessary to run the simulation twice and then manually splice the data for comparison. Summary of the invention

[0009] The purpose of the present invention is to disclose a batch evaluation method for control law parameters of unmanned aerial vehicles. First, a mathematical model of six degrees of freedom of unmanned aerial vehicles is established, and each working point is balanced and linearized in batches to obtain the state space equations of each working point and store them in a table; then, a control law design model is established based on the state space equation to obtain the control law parameters at each working point and store them in a table; finally, a control law evaluation model is established, and the state space equations and control law parameters at different working points are automatically imported into the control law evaluation model through a loop table lookup of an evaluation script, and a batch evaluation function is run to realize batch evaluation of step response and frequency domain indicators of control law parameters. The present invention can realize batch evaluation of control law parameters, including step response and frequency domain indicators such as amplitude margin, phase margin, delay margin, bandwidth frequency, etc., which improves the efficiency and quality of control law design and parameter adjustment.

[0010] The technical solution to achieve the purpose of the present invention is: a batch evaluation method for UAV control law parameters, comprising the following steps:

[0011] Step 1: Establish a mathematical model of the UAV with six degrees of freedom, then select global typical working points and perform batch balancing, linearization, and lateral and longitudinal decoupling to obtain the state space equations of each working point and store them in a table;

[0012] Step 2: Establish control law design model;

[0013] Step 3: Design the model according to the control law and obtain the control law parameters;

[0014] Step 4: Establish a control law parameter evaluation model, import the state space equations obtained in step 1 and the control law parameters obtained in step 3, and perform batch evaluation of the control law parameters through the evaluation script.

[0015] Furthermore, the specific steps in step 1 are as follows:

[0016] Step 1.1: Establish a mathematical model of the UAV with six degrees of freedom;

[0017] Step 1.2: Batch linearization and modal analysis to obtain the state space equations of each operating point.

[0018] Furthermore, the specific steps in step 1.1 are as follows:

[0019] Step 1.1.1: Use the Simulink module building method to build the aerodynamic model, weight structure model, engine model, atmosphere model, and gravity model of the UAV, and establish a six-degree-of-freedom nonlinear mathematical simulation model of the UAV;

[0020] Step 1.1.2: The rudder effect deviation module forms a matching control disturbance by adding the deviation coefficient to the rudder surface related items in the aerodynamic force and moment coefficient, thereby simulating the comprehensive uncertainty caused by the inaccurate modeling of aerodynamic data and environmental data superimposed on the control channel.

[0021] Furthermore, the specific steps in step 1.2 are as follows:

[0022] Step 1.2.1: Select a typical working point in the flight envelope, form a working point vector, and set the pull coefficient;

[0023] Step 1.2.2: Determine the loop length according to the vector length of the working point. In the loop, use the Matlab linearization and balancing method to balance and linearize the UAV six-degree-of-freedom model at each working point, and decouple it into two channels, longitudinal and lateral, to obtain the state space equation. Verify the correctness and controllability of the decoupled model through modal analysis.

[0024] Step 1.2.3: Solve the state-space equations and balance results, and name and store them.

[0025] Furthermore, in step 2, based on the Simulink module building method and the designed PID control law structure, unified control law design models of each control channel are designed respectively, including a pitch control design model, an altitude control design model, a roll control design model, and a yaw control design model.

[0026] Furthermore, in step 3, based on the inner and outer loop cascade design idea, the root locus method is used to design the control law for each working point, and the solved control law parameters are stored separately in .mat files according to the naming method of height and speed.

[0027] Furthermore, the specific steps of step 4 are as follows:

[0028] Step 4.1: Establish a unified control law parameter evaluation model for time domain and frequency domain analysis respectively;

[0029] Step 4.2: Create and run the domain assessment script;

[0030] Step 4.3: Build and run the frequency domain evaluation script.

[0031] Further, step 4.2 includes the following steps:

[0032] Step 4.2.1: Enter the operating point vector and the pull coefficient to be analyzed in the frequency domain evaluation script;

[0033] Step 4.2.2: Set the loop length based on the operating point vector length, match the stored results in step 1 and step 3 by looking up the table, obtain the state space equation and control law parameters at the corresponding operating point, and write them into the global variables;

[0034] Step 4.2.3: Automatically import the state space equations and control law parameters into the control law parameter evaluation model in the loop;

[0035] Step 4.2.4: Run the domain evaluation function in the loop, obtain the evaluation index results, write the evaluation index results into the storage space, and start the next loop;

[0036] Step 4.2.5: After the cycle is completed, a drawing of the step response curves of all operating points is formed, completing the batch time domain evaluation of the control law parameters.

[0037] Further, step 4.3 includes the following steps:

[0038] Step 4.3.1: First, enter the operating point vector and the pull coefficient to be analyzed in the frequency domain evaluation script;

[0039] Step 4.3.2: Set the loop length based on the working point vector length, and establish a frequency domain indicator output table based on the frequency domain indicator and working point configuration;

[0040] Step 4.3.3: Set the loop length based on the length of the operating point vector, match the stored results in steps 2 and 3 by looking up the table, obtain the state space equation and control law parameters at the corresponding operating point, and write them into the global variables;

[0041] Step 4.3.4: Automatically import the state space equations and control law parameters into the control law parameter evaluation model in the loop;

[0042] Step 4.3.5: Run the frequency domain index evaluation function in the loop, obtain the evaluation index results, write the evaluation index results into the storage space and table, and start the next loop;

[0043] Step 4.3.6: After the cycle is completed, an Excel table containing the frequency domain indicators of all operating point control parameters is generated, completing the batch frequency domain evaluation of this control law parameter.

[0044] Compared with the prior art, the present invention has the following significant advantages:

[0045] (1) The batch evaluation method of UAV control law parameters designed in the present invention improves the problems of high repetition and low efficiency in the control law parameter evaluation work in the engineering design of the control law of the UAV flight control system, realizes batch control law parameter evaluation, and improves the efficiency and quality of control law design and parameter adjustment.

[0046] (2) The batch evaluation method of the control law parameters of the UAV designed by the present invention designs an integrated UAV six-degree-of-freedom model with batch modal analysis and linearization functions, which can simultaneously improve the efficiency of control law design.

[0047] (3) The batch evaluation method of UAV control law parameters designed by the present invention has a unified control law design model and a control law evaluation model, so that the design and evaluation of control laws of different models of UAVs can be realized in a process-based design scheme. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is the flow chart of the batch evaluation method of control law parameters.

[0049] Figure 2 This is the effect diagram of batch step response evaluation of control law parameters. DETAILED DESCRIPTION

[0050] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those generally understood by those of ordinary skill in the art to which the present invention belongs.

[0051] The engineering method for batch evaluation of the control law parameters of the UAV has the following specific steps:

[0052] Step 1: Establish a mathematical model of the UAV with six degrees of freedom, and then select global typical working points to perform batch balancing, linearization, and lateral and longitudinal decoupling to obtain the state space equations of each working point and store them in a table.

[0053] Step 1.1: Establish a six-degree-of-freedom mathematical model of the UAV.

[0054] Step 1.1.1: Use the Simulink module building method to build the UAV's aerodynamic model, weight structure model, engine model, atmosphere model, and gravity model respectively, and establish a six-degree-of-freedom nonlinear mathematical simulation model of the UAV.

[0055] The constructed UAV six-degree-of-freedom nonlinear mathematical simulation model takes as input the control rudder deviation and engine speed, and outputs the UAV altitude, speed, attitude, attitude angular rate, acceleration, overload, etc.

[0056] Step 1.1.2: The rudder effect deviation module forms a matching control disturbance by adding the deviation coefficient to the rudder surface related items in the aerodynamic force and moment coefficient, thereby simulating the comprehensive uncertainty caused by the inaccurate modeling of aerodynamic data and environmental data superimposed on the control channel.

[0057] Step 1.2: Batch linearization and modal analysis to obtain the state space equations of each operating point.

[0058] Step 1.2.1: Select a typical working point (altitude, speed) in the flight envelope, form a working point vector, and set the pull coefficient.

[0059] Step 1.2.2: Determine the loop length according to the vector length of the working point. In the loop, use the Matlab linearization and balancing method to balance and linearize the UAV six-degree-of-freedom model at each working point, and decouple it into two channels, longitudinal and lateral, to obtain the state space equation. Verify the correctness and controllability of the decoupled model through modal analysis.

[0060] Step 1.2.3: The solved state-space equations and trim results (including nominal and trimmed) are stored in .mat files according to the naming method of height, speed, and trim.

[0061] Step 2: Establish the control law design model.

[0062] Based on the Simulink module building method and the designed PID control law structure, a unified control law design model for each control channel is designed, including pitch control design model, altitude control design model, roll control design model, and yaw control design model.

[0063] The state space equations in the control law design model are introduced by calling the workspace, that is, the state space equations of each operating point of each channel are obtained by loading the state space equations in the .mat file obtained in step 1.

[0064] Step 3: Design the model according to the control law and obtain the control law parameters.

[0065] Based on the inner and outer loop cascade design idea, the root locus method is used to design the control law for each working point. The solved control law parameters are stored separately in .mat files according to the naming method of height and speed.

[0066] Step 4: Establish a control law parameter evaluation model, import the state space equations obtained in step 1 and the control law parameters obtained in step 3, and perform batch evaluation of the control law parameters through the evaluation script.

[0067] Step 4.1: For time domain and frequency domain analysis, a unified control law parameter evaluation model is established, in which the state space equations and control law parameters are set to be readable from global variables.

[0068] Step 4.2: Create and run the domain assessment script.

[0069] Step 4.2.1: First, enter the height, velocity vector and pull coefficient of the working point to be analyzed in the frequency domain evaluation script.

[0070] Step 4.2.2: Set the loop length based on the operating point vector length, match the stored results in steps 1 and 3 by looking up the table, obtain the state space equations and control law parameters at the corresponding operating point, and write them to the global variables.

[0071] Step 4.2.3: Automatically import the state space equations and control law parameters into the control law parameter evaluation model in the loop.

[0072] Step 4.2.4: Run the time domain evaluation function in the loop. First, run the linearize function based on the control law parameter evaluation model to get the linear description. Then call the tf function to get the transfer function. Finally, get the evaluation index result through step, write the evaluation index result into the storage space, and start the next loop.

[0073] Step 4.2.5: After the cycle is completed, a drawing of the step response curves of all operating points is formed, completing the batch time domain evaluation of the control law parameters.

[0074] Step 4.3: Build and run the frequency domain evaluation script.

[0075] Step 4.3.1: First, enter the height, velocity vector and pull coefficient of the working point to be analyzed in the frequency domain evaluation script.

[0076] Step 4.3.2: Set the loop length based on the working point vector length, and create a frequency domain indicator output table based on the frequency domain indicators and working point configuration.

[0077] Step 4.3.3: Set the loop length based on the length of the operating point vector, match the stored results in steps 2 and 3 by looking up the table, obtain the state space equations and control law parameters at the corresponding operating point, and write them to the global variables.

[0078] Step 4.3.4: Automatically import the state space equations and control law parameters into the control law parameter evaluation model in the loop.

[0079] Step 4.3.5: Run the frequency domain index evaluation function in the loop. First, run the linearize function based on the control law parameter evaluation model to obtain the linear description. Then call the tf function to obtain the transfer function. Finally, get the evaluation index result through allmargin, write the evaluation index result into the storage space and table, and start the next loop.

[0080] Step 4.3.6: After the cycle is completed, an Excel table containing the frequency domain indicators of all operating point control parameters is generated, completing the batch frequency domain evaluation of this control law parameter.

[0081] In the batch evaluation method for the control law parameters of the UAV, the state space equation obtained in step 1 will be used for both the control law design and the control law evaluation, the control law parameters obtained in step 3 will be imported into the control law parameter evaluation model together with the state space equation, and finally the batch evaluation function of the control law parameters will be realized based on the control law evaluation model and the evaluation script.

[0082] Example

[0083] Assume that a certain UAV is at a height (H) of 500m and a speed (V) of 118m / s, 186m / s, and 254m / s respectively; Figure 2 The following are the step responses of the roll angle control at speeds (V) of 118m / s, 186m / s, and 254m / s. Figure 2 The time domain evaluation effects of the control law parameters at different working points can be easily compared, which makes it convenient for designers to adjust the parameters globally to enhance the consistency of the steady-state performance and dynamic performance of the UAV roll angle control in the entire flight envelope.

[0084] Table 1 shows the frequency domain evaluation results of the pitch angle control of a certain type of UAV at different altitudes (H) and speeds (V). In the table, Kq and K_theta are the pitch angle rate gain and pitch angle gain to be evaluated, respectively. Mag indicates the amplitude margin, Phase indicates the phase margin, Freg indicates the cutoff frequency, ξ is the damping ratio, and "nominal" and "1.5 times" in the table indicate the evaluation results under no rudder effect and 1.5 times rudder effect. From Table 1, it is convenient to compare the frequency domain evaluation effects of the control law parameters at different working points, summarize the correctness and regularity of each indicator, and facilitate the adjustment and evaluation of designers.

[0085] Table 1 Frequency domain evaluation results of pitch angle control of a certain type of UAV at different altitudes and speeds

[0086]

Claims

1. A batch evaluation method for UAV control law parameters, characterized in that: The steps include: Step 1: Establish a mathematical model of the UAV with six degrees of freedom, then select global typical working points and perform batch balancing, linearization, and lateral and longitudinal decoupling to obtain the state space equations of each working point and store them in a table; Step 2: Establish control law design model; Step 3: Design the model according to the control law and obtain the control law parameters; Step 4: Establish a control law parameter evaluation model, import the state space equations obtained in step 1 and the control law parameters obtained in step 3, and perform batch evaluation of the control law parameters through the evaluation script.

2. The batch evaluation method of UAV control law parameters according to claim 1 is characterized in that: The specific steps in step 1 are as follows: Step 1.1: Establish a mathematical model of the UAV with six degrees of freedom; Step 1.2: Batch linearization and modal analysis to obtain the state space equations of each operating point.

3. The batch evaluation method of UAV control law parameters according to claim 2 is characterized in that: The specific steps in step 1.1 are as follows: Step 1.1.1: Use the Simulink module building method to build the aerodynamic model, weight structure model, engine model, atmosphere model, and gravity model of the UAV, and establish a six-degree-of-freedom nonlinear mathematical simulation model of the UAV; Step 1.1.2: The rudder effect deviation module forms a matching control disturbance by adding the deviation coefficient to the rudder surface related items in the aerodynamic force and moment coefficient, thereby simulating the comprehensive uncertainty caused by the inaccurate modeling of aerodynamic data and environmental data superimposed on the control channel.

4. The batch evaluation method of UAV control law parameters according to claim 2 is characterized in that: The specific steps in step 1.2 are as follows: Step 1.2.1: Select a typical working point in the flight envelope, form a working point vector, and set the pull coefficient; Step 1.2.2: Determine the loop length according to the vector length of the working point. In the loop, use the Matlab linearization and balancing method to balance and linearize the UAV six-degree-of-freedom model at each working point, and decouple it into two channels, longitudinal and lateral, to obtain the state space equation. Verify the correctness and controllability of the decoupled model through modal analysis. Step 1.2.3: Solve the state-space equations and balance results, and name and store them.

5. The batch evaluation method of UAV control law parameters according to claim 1 is characterized in that: In step 2, based on the Simulink module building method and the designed PID control law structure, a unified control law design model for each control channel is designed, including a pitch control design model, an altitude control design model, a roll control design model, and a yaw control design model.

6. The batch evaluation method of UAV control law parameters according to claim 1 is characterized in that: In step 3, based on the inner and outer loop cascade design idea, the root locus method is used to design the control law for each working point, and the solved control law parameters are stored separately in .mat files according to the naming method of height and speed.

7. The batch evaluation method of UAV control law parameters according to claim 1 is characterized in that: The specific steps of step 4 are as follows: Step 4.1: Establish a unified control law parameter evaluation model for time domain and frequency domain analysis respectively; Step 4.2: Create and run the domain assessment script; Step 4.3: Build and run the frequency domain evaluation script.

8. The batch evaluation method of UAV control law parameters according to claim 7 is characterized in that: Step 4.2 includes the following steps: Step 4.2.1: Enter the operating point vector and the pull coefficient to be analyzed in the frequency domain evaluation script; Step 4.2.2: Set the loop length based on the operating point vector length, match the stored results in step 1 and step 3 by looking up the table, obtain the state space equation and control law parameters at the corresponding operating point, and write them into the global variables; Step 4.2.3: Automatically import the state space equations and control law parameters into the control law parameter evaluation model in the loop; Step 4.2.4: Run the domain evaluation function in the loop, obtain the evaluation index results, write the evaluation index results into the storage space, and start the next loop; Step 4.2.5: After the cycle is completed, a drawing of the step response curves of all operating points is formed, completing the batch time domain evaluation of the control law parameters.

9. The batch evaluation method of UAV control law parameters according to claim 8 is characterized in that: Step 4.3 includes the following steps: Step 4.3.1: First, enter the operating point vector and the pull coefficient to be analyzed in the frequency domain evaluation script; Step 4.3.2: Set the loop length based on the working point vector length, and establish a frequency domain indicator output table based on the frequency domain indicator and working point configuration; Step 4.3.3: Set the loop length based on the length of the operating point vector, match the stored results in steps 2 and 3 by looking up the table, obtain the state space equation and control law parameters at the corresponding operating point, and write them into the global variables; Step 4.3.4: Automatically import the state space equations and control law parameters into the control law parameter evaluation model in the loop; Step 4.3.5: Run the frequency domain index evaluation function in the loop, obtain the evaluation index results, write the evaluation index results into the storage space and table, and start the next loop; Step 4.3.6: After the cycle is completed, an Excel table containing the frequency domain indicators of all operating point control parameters is generated, completing the batch frequency domain evaluation of this control law parameter.

Images