KR20190110361A - The method of active vibration control for helicopter - Google Patents

Abstract

The present invention relates to an active vibration control method of a helicopter comprising: a signal processing step of receiving information from an accelerometer arranged on a helicopter to extract a sensor output value including an N/rev value; a step of selecting a helicopter property model matched to the sensor output value among preset helicopter property models; a parameter selection step of determining a control speed of a control algorithm in accordance with a vibration level of the sensor output value and a mode selection of a user; a vibration control model updating step of updating a vibration control algorithm from the parameter and the helicopter property model; and a control input generation step of generating a control input to drive one or more load generators from the updated vibration control model. The active vibration control method of a helicopter can select a helicopter property model in accordance with vibration of a main rotor to generate a control input to reduce vibration of a helicopter.

Description

Helicopter active vibration control method {THE METHOD OF ACTIVE VIBRATION CONTROL FOR HELICOPTER}

The present invention relates to a method for controlling active vibration of a helicopter, and more particularly, to an active vibration control and a trouble-detectable control method for reducing vibration generated in a helicopter.

The main rotor of the helicopter generates a significant vibration as it rotates, as a way to reduce this has been developed passive and active vibration reduction device. Passive vibration reduction devices are constructed of spring-mass and are designed to have a natural frequency that matches the excitation frequency for vibration reduction at the local position of the helicopter to reduce vibration. The active vibration reducing device reduces the vibration by adjusting the output of the load generator according to the main vibration. Republic of Korea Patent Publication No. 10-2014-0068875 is disclosed in relation to the conventional active vibration reduction technology.

However, this conventional technology has a problem that instability occurs when controlling at a high speed.

Republic of Korea Patent Publication No. 10-2014-0068875

Disclosure of Invention The present invention aims to provide a method for controlling an active vibration of a helicopter, which solves the problems of an active vibration reduction technology of a conventional helicopter at high speed and ensures sufficient stability even in a sudden vibration change.

As a means for solving the above problems, a signal processing step of receiving information from the accelerometer provided in the helicopter according to the present invention to extract the sensor output value including the N / rev value, helicopter characteristics matching the sensor output value of the preset helicopter characteristics model Selecting a model, selecting a parameter to determine the control speed of the control algorithm according to the vibration level of the sensor output value and the user's mode selection, updating the vibration control model to update the vibration control algorithm from the parameter and the helicopter characteristic model, and There may be provided an active vibration control method for a helicopter including a control input generation step of generating a control input for driving one or more load generators from an updated vibration control model.

Here, the helicopter characteristic model may be calculated based on information received from the accelerometer when the load generator is driven while the helicopter's main rotor is rotating.

The signal processing step performs bandpass filtering on the information received from the accelerometer, modulates the cosine and sin signals synchronized with the phase of the tachometer by the filtered information, and modulates the modulated information. Can be configured to extract a sensor output value from the N / rev acceleration component including the magnitude and phase information.

The parameter selection step corresponds to any one of a plurality of sections according to the sensor output value, and the control convergence rate is differently applied to have different control speeds for each section, and the mode selection is applied to the vibration reduction target portion. Can be configured to select weights accordingly.

In addition, the vibration control model update step updates the objective function of the vibration control algorithm, calculates the gradient of the control load based on the convergence rate, the helicopter characteristic model, and the acceleration reflecting the weight, and the current control load and stability improvement variables. And a control input based on the calculated gradient of the control load.

Furthermore, in the selecting of the helicopter characteristic model, any one model may be selected corresponding to the main frequency of the helicopter among the plurality of helicopter characteristic models.

In the meantime, the selected helicopter characteristic model may include a Hermitian transpose so that the selected helicopter characteristic model may be applied to the vibration control update equation.

The active vibration control method of the helicopter according to the present invention can select the helicopter characteristic model according to the rotational speed of the main rotor to generate a control input, thereby reducing the vibration according to the rotational speed of the main rotor of the helicopter.

1 is an active vibration control law of a helicopter according to a first embodiment of the present invention.
2 is a conceptual diagram of a helicopter characteristic generation step of the first embodiment.
3 is a flowchart of a helicopter characteristic generation step of the first embodiment.
4 is a flowchart of a signal processing step of the first embodiment.
5 is a block diagram of a helicopter characteristic model selection step of the first embodiment.
6 is a block diagram of a control variable selection step of the first embodiment.
7 is a conceptual diagram for determining the step size in the control variable selection step.
8 is a block diagram of the vibration control step of the first embodiment.
9 is a flowchart of an active vibration control method of a helicopter according to a second embodiment of the present invention.
10 is a flowchart of the high vibration diagnosis step of the second embodiment.
Fig. 11 is a flowchart of the rotor speed diagnosis step of the second embodiment.
12 is a flowchart of a control state diagnosis step of the second embodiment.
13 is a flowchart of a control input / output diagnosis step of the second embodiment.

Hereinafter, an active vibration control method of a helicopter according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings. And the names of each component in the description of the following embodiments may be called other names in the art. However, if their functional similarity and identity, even if the modified embodiment can be seen as an equivalent configuration. In addition, the symbols added to each component is described for convenience of description. However, the contents shown in the drawings in which these symbols are described do not limit each component to the ranges in the drawings. Similarly, even if an embodiment in which the configuration on the drawings is partially modified is employed, it can be regarded as an equivalent configuration if there is functional similarity and identity. In addition, in view of the general level of those skilled in the art, if it is recognized as a component to be included naturally, the description thereof will be omitted.

'Helicopter' or 'helicopter' described below means the same meaning as the HELICOPTER.

1 is an active vibration control law of a helicopter according to a first embodiment of the present invention.

As shown, the active vibration control law of the helicopter according to the present invention is a helicopter characteristic model generation step (S100), signal processing step (S200), helicopter characteristic model selection step (S300), control parameter selection step (S400), It may be configured to include a vibration control step (S500). Helicopter characteristic model is generated before the flight of the helicopter, the helicopter characteristic model is selected according to the vibration information received during the flight and reflects this to generate a control input to operate the load generator provided in the helicopter.

2 is a conceptual diagram of the helicopter characteristic model generation step of the first embodiment, and FIG. 3 is a flowchart of the helicopter characteristic model generation step of the first embodiment.

The helicopter characteristic model generation step (S100) is a step of generating a helicopter characteristic model which is an input value of vibration control. The helicopter characteristic model is constructed to calculate the characteristic model of the helicopter while the main rotor is rotating.

Helicopter characteristic model generation step (S100) is to generate a control command for driving the actuator (S110), actuator control command transmission step (S120), acceleration signal acquisition step (S130), using the acceleration signal and the actuator control command Computing the characteristic model (S140), repeatedly performing as many as the number of actuators to accumulate and temporarily store the characteristic model (S150), storing the calculated characteristic model in a matrix form (S160), the characteristic model in the memory It may be configured to include a step (S170) for storing.

The helicopter characteristic model drives the load generator generating dynamic load and can be calculated using the acceleration data obtained from the accelerometer when the vibration caused by the driving of the load generator and the vibration generated in the main rotor are combined. Vibration may differ between the level of vibration expected in theory or design and the vibration that is actually generated, due to structural influences and various factors generated during the assembly process. Therefore, the model is measured experimentally by measuring inputs and outputs in a real helicopter. The helicopter characteristic model, which is a model of the helicopter system, can be generated in plural according to the number of load generators and the vibration level of the main rotor.

The signal processing step S200 is configured to extract the N / rev component of the signal of the accelerometer. Signal processing step (S200) is the second IIR low pass filtering step (S210), the second IIR high pass filtering step (S220), cosine step (cos), sin signal step (sin) signal modulation step (S230), acceleration component of Size, phase extraction step (S240), filter phase error correction step (S250) and size limiting step (S260) can be configured.

The second-order IIR low pass filtering step S210 is a step of applying a second-order Infinite Impulse Responce (IIR) low-pass filter to a signal received from an accelerometer.

Secondary IIR high pass filtering step (S220) is a step of applying a second IIR high-pass filter (High-pass filter) to the signal received from the accelerometer.

A band-pass filter function may be implemented through second-order IIR low frequency and high frequency filtering.

A cosine and sin signal modulation step S230 corresponds to a step of modulating the signal by multiplying the cosine and sine signals synchronized with the phase of the dachometer by the filtered acceleration signal.

The magnitude and phase extraction step of the acceleration component (S240) corresponds to the step of extracting the magnitude and phase information of the N / rev acceleration component from the modulated acceleration signal.

In the filter phase error correction step S250, a function for compensating for the phase error according to the characteristics of the bandpass filter is performed.

The size limiting step S260 adjusts the maximum value of the output N / rev acceleration component so as not to exceed the limit.

The helicopter characteristic model selection step (S300) selects a helicopter characteristic (C-model) to be applied to the vibration control equation according to the acceleration signal and corresponds to the step of deforming to apply to the vibration control.

5 is a block diagram of a helicopter characteristic model selection step of the first embodiment. As shown, the helicopter characteristic model selection step includes the helicopter characteristic model index selection step (S310) according to the main excitation frequency (N / rev frequency), the Hermit preposition step (S320) of the helicopter characteristic model, the accelerometer / load generator state information It may be configured to include a calculation and output step (S330).

The helicopter characteristic model index selection step (S310) according to the main excitation frequency may calculate the helicopter characteristic model according to the N / rev frequency. The helicopter characteristic model selects the model index according to the main frequency based on the input frequency setting range and the range number value, and calculates the helicopter characteristics through this.

The Hermit transposition step S320 of the helicopter characteristic model is a step of performing the Hermit transposition so that the Helicopter characteristic model for the main frequency can be applied to the vibration control update equation.

Accelerometer / load generator state information calculation and output step (S330) corresponds to the step of calculating and outputting the state information to be applied to the control by comparing the current state and the configuration information mounted on the accelerometer and load generator.

Control variable selection step (S400) selects the convergence rate (convergence rate) according to the average vibration level, and corresponds to the step of selecting the sensor weight (accelerometer weighting) according to the user's input (discrete input).

6 is a block diagram of a control variable selection step of the first embodiment. As shown in the drawing, the control variable selection step may include a convergence rate selection step S410 and a mode selection step according to a user input S420.

7 is a conceptual diagram for determining the step size in the control variable selection step. Convergence rate selection step (S410) corresponds to the step of selecting the convergence rate in consideration of the control speed and stability at the current vibration level. As shown in FIG. 7, the control convergence rate is a parameter that determines the control speed of the vibration control algorithm, and the control convergence rate may be determined based on a plurality of threshold values. If the control convergence rate is high (the step size is small so that the control input is generated at shorter intervals), the control speed is high, but the control state may become unstable due to the sudden change of the control input, and conversely, if the control convergence rate is low (relatively, In order to generate control inputs at long intervals, the control speed is low, but stability is ensured, so appropriate trade-off is required. In the present invention, as an example, three sections are classified according to the average vibration level, and the control convergence rate is selected to be high for zone 1 with a high vibration level, the control convergence rate is low for zone 3, and the control is performed for zone 2, which is an intermediate value. Select the convergence rate to be medium. However, the three convergence rates are just examples and may be determined by being divided into appropriate numbers according to hardware performance and control algorithms.

The mode selection step S420 is a step of selecting a region in which the user, that is, the pilot concentrates vibration reduction. Mode selection step (S420) is composed of a total of three modes, the general mode does not apply weight as a parameter according to the mode selection input of the pilot, the pilot mode mainly for reducing the vibration of the cockpit and the cabin mode for reducing the vibration of the cabin mainly on the VIP boarding Can be. The control parameter is selected according to the mode selection and reflected in the control input to concentrate the vibration reduction.

The vibration control step S500 corresponds to the step of calculating the control load and the objective function by updating the vibration control equation based on the helicopter state model.

8 is a block diagram of the vibration control step of the first embodiment. As shown, the vibration control model is configured to calculate the control load and the objective function of the leaky Filtered-x least Mean Square (leakly F-x LMS) algorithm according to the vibration control update equation.

The vibration control update equation is constructed as follows.

는 Command,

는 Current Command이고,

는 Convergence rate,

는 Leak(안정성 향상 factor),

는 C-model,

는 Hermitian transpose,

는 Accelerometer weighting,

는 error (acceleration)을 뜻한다.here

Command,

Is the Current Command,

Is the convergence rate,

Is the Leak (stability improvement factor),

The C-model,

Hermitian transpose,

Accelerometer weighting,

Means error (acceleration).

및

는 복소 matrix이며,

및

는 scalar이며,

는 복소 matrix이며,

는 실수 matrix이고,

는 복소 matrix이다.And

And

Is a complex matrix,

And

Is scalar,

Is a complex matrix,

Is a real matrix,

Is a complex matrix.

Vibration control algorithm The objective function equation is as follows.

는 Cost function of leaky F-x LMS Algorithm,

는 error (acceleration),

는 Hermitian transpose,

는 Accelerometer weighting,

는 Leak,

는 Current Command이다. here

The cost function of leaky Fx LMS Algorithm,

Error (acceleration),

Hermitian transpose,

Accelerometer weighting,

Leak,

Is the Current Command.

및

는 scalar이며.

및

는 복소 matrix이고,

는 실수 matrix이다.And,

And

Is scalar.

And

Is a complex matrix,

Is a real matrix.

) 계산 단계(S520), 제어하중 수식을 통한 제어하중 계산단계(S530), 하중발생기 사이 연관행렬 적용(Force Mapping)단계(S540), 제어하중에 하중발생기 상태정보 적용단계(S550), 제어하중 한정(Limitation)단계(S560), 현재 제어하중 계산단계(S570), 제어하중 지연효과 적용(Slew rate)단계(S580), 목적함수 계산단계(S590)를 포함하여 구성될 수 있다.The vibration control step is an acceleration sensor weight application step (S510), the control load gradient portion (

) Calculation step (S520), control load calculation step through the control load equation (S530), the relationship between the load generator (Force Mapping) step (S540), applying the load generator status information to the control load (S550), control load It may include a limitation step (S560), the current control load calculation step (S570), the control load delay effect (Slew rate) step (S580), the objective function calculation step (S590).

Acceleration sensor weighting step (S510) is to apply the weight and status information by multiplying the acceleration by the sensor weight output in the control parameter selection step (S400) and the state information of the accelerometer output in the helicopter characteristic model selection step (S300), respectively. Step.

) 계산 단계(S520)는 수렴률(convergence rate), 헬기특성모델 및 가중치 적용된 가속도를 입력받아 제어하중의 그래디언트(gradient,

성분)를 계산하는 단계이다.Gradient part of control load (

The calculation step (S520) receives a convergence rate, a helicopter characteristic model, and a weighted acceleration and receives a gradient of a control load.

Component).

The control load calculation step (S530) through the control load equation is a step of calculating the control load by receiving the control load gradient outputted from the control load gradient portion calculation step 520, the current control load, and the leak. The control load may be calculated according to the control load equation described above.

The force matrixing step (S540) between load generators is a step of performing force mapping by multiplying the force mapping matrix by a control load matrix. Force mapping is configured to control the direction and size of control loads by mounting two load generators in multiple ways.

The step of applying the load generator state information to the control load (S550) corresponds to the step of applying the state information by multiplying the load generator state information output in the helicopter characteristic model selection step (S300) to the control load.

The control load limitation step (S560) is configured to limit the control load according to the set control load amplitude limit and to output the control load in the form of complex amplitude and phase. The complex type control load can be calculated with real and imaginary parts using proportional expression so that the phase does not change.

The amplitude of the control load may output a smaller control value among the control load amplitude limit and the calculated control load amplitude, and the phase of the control load may be compensated for the negative phase to output a value of 0 to 2 pi.

The present control load calculation step (S570) is a step of applying a unit delay to use the control load to be output as the current control load in the next step.

Applying the control load delay effect (Slew rate) step (S580) is a step to limit the control load by the value set for each step by applying the delay effect for the stable change of the control load.

)에 따라 알고리즘 자체진단시험(CBIT)에 사용할 목적함수를 계산하는 단계이다.The objective function calculating step (S590) is the objective function formula of the algorithm (

In this step, the objective function to be used for the algorithm self-diagnosis test (CBIT) is calculated.

On the other hand, the above-described helicopter characteristic model generation step (S100) may be set in advance before the flight of the helicopter, the signal processing step (S200) to vibration control step (S500) may be performed repeatedly during the flight of the helicopter.

Hereinafter, a helicopter active vibration control method according to a second embodiment of the present invention will be described with reference to FIGS. 9 to 13. The second embodiment may be configured to include the same components as the first embodiment, and the same components will be omitted in order to avoid overlapping descriptions, and the differences will be described in detail.

9 is a flowchart of an active vibration control method of a helicopter according to a second embodiment of the present invention.

As shown in the second embodiment according to the present invention, after the vibration control step S500, the high vibration diagnosis step S600, the rotor rotation speed diagnosis step S700, the control state diagnosis step S800, and the control input / output diagnosis step S900. It may be configured to include).

The second embodiment includes a high vibration diagnosis step S600 to a control input / output diagnosis step S900 to comprehensively diagnose the current control state.

10 is a flowchart of the high vibration diagnosis step of the second embodiment. As shown, the high vibration diagnosis step determines whether high vibration is generated for the individual signals and the average signal of the accelerometer signal. Here, Accel_RMS_Avg_Calculation calculates an average value of the accelerometer signal, and High_Vibration_Check determines whether high vibration occurs for each individual accelerometer signal and accelerometer average signal.

Fig. 11 is a flowchart of the rotor speed diagnosis step of the second embodiment. As shown, the rotor rotation diagnosis model includes a oneP frequency Average Calculation module, an Acceleration RMS Average Calculation module, a Tach Status module, and a Tach Valid Check module.

The oneP frequency Average Calculation module calculates the average value of the oneP frequency value of 25 signals per step. The Acceleration RMS Average Calculation module calculates the average value of the accelerometer signal magnitudes. The Tach Status module calculates the average of the oneP frequency signal for 1 second and determines the tachometer signal status as Low, Normal, or High based on the result.The Tach Valid Check module compares the average value of the accelerometer signal size with the tachometer signal status. Determine whether there is an abnormal tachometer signal.

12 is a flowchart of a control state diagnosis step of the second embodiment. As shown, the control state diagnosis model receives the objective function as an input and checks the value of the objective function in real time. There is only one type of event that occurs through the control diagnostic module, and the risk level is defined as Level 0. Accordingly, the objective function value is output as event data when an event occurs. The control diagnosis module is divided into ControlAlgorithm CBIT step to check the objective function and Control Bit Log step to generate output value. The ControlAlgorithm CBIT step performs diagnostic wait, event declaration, and event clearing functions. The Control BIT Log step adjusts the output value so that the CNT Detect and objective function values are output only when the event is declared.

13 is a flowchart of a control input / output diagnosis step of the second embodiment. As shown, the control input diagnostic model receives the CFG (load generator) load command value (size / phase), load output value (size / phase), and CFG status value as inputs, and the difference between the command value and the output value for the activated CFG. Calculate the size by phase and phase, and check the difference value. Events that occur through the control input diagnosis module can be 24 types according to the size / phase of 12 CFGs as an example, and the risk level is defined as Level 0. Since all events can occur at the same time, the event code does not identify the CFG ID where the event occurred. Therefore, the event code is defined to distinguish how many events are duplicated and the corresponding CFG ID is defined through event data.

As described above, the active vibration control method of the helicopter according to the present invention can select the helicopter characteristic model according to the vibration of the main rotor to generate a control input, thereby reducing the vibration of the helicopter.

S100: helicopter characteristic model generation step
S200: Signal Processing Step
S300: Selecting Helicopter Characteristic Model
S400: Steps to select control parameters
S500: vibration control step
S600: High Vibration Diagnostic Step
S700: Rotor rpm diagnostic step
S800: Control Status Diagnostic Steps
S900: Control Input / Output Diagnosis Step

Claims (7)

A signal processing step of receiving information from an accelerometer provided in the helicopter and extracting a sensor output value including an N / rev value;
Selecting a helicopter characteristic model matching the sensor output value from a preset helicopter characteristic model;
A parameter selecting step of determining a control speed of a control algorithm according to the vibration level of the sensor output value and a user's mode selection;
A vibration control model updating step of updating a vibration control algorithm from the parameter and the helicopter characteristic model; And
And a control input generating step of generating a control input for driving one or more load generators from the updated vibration control model. According to claim 1,
The helicopter characteristic model,
Helicopter active vibration control method of the helicopter, characterized in that calculated based on the information received from the accelerometer when the load generator is driven. The method of claim 2,
The signal processing step,
Performing bandpass filtering on the information received from the accelerometer,
Performs modulation by multiplying the cosine and sin signals synchronized with the phase of the tachometer by the filtered information,
And a sensor output value including magnitude and phase information of an N / rev acceleration component is extracted from the modulated information. The method of claim 2,
The parameter selection step,
Corresponding to any one of a plurality of sections according to the sensor output value, the control convergence rate is differently applied to have a different control speed for each section,
And the mode selection selects a weight according to the vibration reduction target portion. The method of claim 4, wherein
The updating of the vibration control model updates the object function of the vibration control algorithm.
Calculating a gradient of a control load based on the convergence rate, the helicopter characteristic model, and the weighted acceleration;
A method for controlling the active vibration of a helicopter, comprising calculating a control input based on a gradient of a current control load, a stability enhancing variable, and a calculated control load. The method of claim 2,
The selecting of the helicopter characteristic model may include any one of a plurality of helicopter characteristic models selected according to the main frequency of the helicopter. The method of claim 6,
The helicopter characteristic model selected in the step of selecting the helicopter characteristic model, the active vibration control method of the helicopter, characterized in that the Hermit transpose is applied to be applied to the vibration control update equation.

Images