KR101193116B1 - Method of recognizing flight pattern of rotor craft, apparatus for recognizing flight pattern implementing the same and recording medium - Google Patents

Abstract

A method of recognizing a flight pattern of a rotorcraft and a recording medium embodying the same are provided. The method of recognizing a flight pattern of the rotorcraft aircraft includes: measuring flight data of the aircraft to generate a plurality of measurement data; basic flight states including ground flight, low speed / hover flight, and high speed flight based on the measurement data; Determining a first flight state of the aircraft during the second flight state among the detailed flight states related to the detailed altitude, movement type, and attitude of the aircraft based on the measurement data within the conditions of the first flight state. Determining, when the aircraft is abnormally flyed within the conditions of the second flight state, determining a third flight state among abnormal flight states associated with rapid flight control and the first, second and first Outputting a comprehensive flight status of the aircraft including three flight statuses.

Description

Recognition method of flight pattern of a rotorcraft aircraft, flight pattern recognition device and recording medium implementing the same {METHOD OF RECOGNIZING FLIGHT PATTERN OF ROTOR CRAFT, APPARATUS FOR RECOGNIZING FLIGHT PATTERN IMPLEMENTING THE SAME AND RECORDING MEDIUM}

The present invention relates to a method for recognizing a flight pattern of a rotary wing aircraft, a flight pattern recognition apparatus and a recording medium for implementing the same, and more particularly, a method of recognizing a flight pattern for determining the detailed flight status of a rotorcraft aircraft, flight pattern recognition An apparatus and a recording medium.

Flight pattern recognition is an important algorithm that is fundamental to the development of rotorcraft systems. Such flight pattern recognition becomes an important factor in determining the life of the aircraft in the case of the life management system, and in the case of the HUMS or the structural monitoring system, it is the basis for analyzing the signals obtained from the main vibration signal or the main structure. For example, in the case of the HUMS vibration signal, by analyzing the vibration signal obtained by flying the aircraft in a stable state, it is possible to analyze the defects of parts of the aircraft, such as a gearbox. On the other hand, when using the vibration signal acquired in the aircraft's transition flight state (turning down, rising, etc.), it is necessary to analyze whether the excessive vibration value is due to the aircraft's transition flight state or gearbox defect. Make it difficult Therefore, flight status recognition is an important algorithm that is the basis for accurate data analysis in the development of main system of rotorcraft.

Recently, various algorithms have been introduced and used. As an example, a pattern recognition technique is used, and this technique mainly grasps the takeoff and landing state and the equilibrium state of the aircraft based on the aircraft's altitude and speed. However, the pattern recognition technique applied to the rotorcraft aircraft is vulnerable to noise such as transient vibration values caused in the aircraft's transition flight state, and thus has an error in recognition of flight patterns.

As such, errors in the analysis of flight patterns affect the life expectancy of the aircraft. Specifically, excessively short life expectancy leads to shortened parts replacement cycles resulting in increased maintenance costs, or excessively long life expectancy leading to fatal losses in the event of an emergency. . That is, the flight pattern recognition applied to the existing aircraft is mainly based on an algorithm based on a simple pattern matching technique, which is difficult to analyze the flight pattern in the transition state of the aircraft, and various studies to improve it Is being tried.

The technical problem to be achieved by the present invention is to recognize the flight pattern of the rotorcraft aircraft to output a more accurate flight pattern by identifying various detailed flight conditions generated in the transition state of the aircraft as well as the basic state of the aircraft, flight pattern recognition An apparatus and a recording medium having recorded thereon a program for implementing the same are provided.

The object of the present invention is not limited to the above-mentioned object, and other objects not mentioned will be clearly understood by those skilled in the art from the following description.

According to an aspect of the present invention for achieving the above technical problem, a method for recognizing a flight pattern of a rotary wing aircraft, generating a plurality of measurement data by measuring the flight data of the aircraft, ground flight, low-speed / Determining a first flight state of the aircraft during basic flight conditions including hover flight and high speed flight, the detailed altitude, movement pattern and attitude of the aircraft based on the measurement data within the conditions of the first flight condition Determining a second flight state during detailed flight states associated with the second flight state; determining a third flight state during abnormal flight states associated with rapid flight control within the conditions of the second flight state; and the first and second And outputting a comprehensive flight status of the aircraft including a third flight status.

Specific details of other embodiments are included in the detailed description and the drawings.

According to the present invention, on the basis of the plurality of measurement data, not only the basic state of the aircraft, but also various detailed flight states, that is, the detailed flight state related to the detailed altitude, movement type and attitude of the aircraft, and the abnormal flight state related to the rapid flight control You can output the correct flight pattern by grasping.

In addition, the detailed flight conditions and abnormal flight conditions may be determined using fuzzy inference and neural network techniques, thereby minimizing the influence of noise generated in the aircraft's transitional flight state. Can be provided to the user.

1 is a flight pattern recognition apparatus according to an embodiment of the present invention.
2 is a block diagram showing a detailed configuration of a second flight state determination unit.
3 is a configuration diagram of a neural network performed by a third flight state determination unit.
4 is a flowchart illustrating a method of recognizing a flight pattern according to an embodiment of the present invention.
5 is a flowchart for configuring a learned fuzzy inference system.
6 is a flowchart for constructing a learned neural network.
7 is a diagram illustrating an example of measurement data.
8 is a correlation table showing the correlation between the measurement data and the basic flight state and the correlation between the measurement data and the detailed flight state.
9 is a view showing a basic flight state and a detailed flight state accordingly.
10 and 11 are diagrams classifying detailed flight states and abnormal flight states according to basic flight states.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings and the contents described below. However, the present invention is not limited to the embodiments described herein and may be embodied in other forms. Rather, the embodiments disclosed herein are being provided so that this disclosure will be thorough and complete, and will fully convey the concept of the invention to those skilled in the art. Like numbers refer to like elements throughout. It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. In this specification, the singular forms also include the plural unless specifically stated otherwise in the text. &Quot; comprises "and / or" comprising ", as used herein, unless the recited element, step, operation, and / Or additions.

Hereinafter, a flight pattern recognition apparatus according to an embodiment of the present invention will be described in detail with reference to FIG. 1. 1 is a device for recognizing a flight pattern according to an embodiment of the present invention.

The flight pattern recognition apparatus 100 is a device that determines various flight states based on a plurality of measurement data which are flight data obtained in actual flight of a rotorcraft. In detail, the flight pattern recognition apparatus 100 may include a flight data measurement unit 110, a correlation table providing unit 120, a first flight state determination unit 130, a second flight state determination unit 140, and a third flight state. The determination unit 150 and the output unit 170 is included.

First, the flight data measuring unit 110 monitors flight data in an actual flight of a rotorcraft aircraft to obtain a plurality of measurement data. As shown in FIG. 7, the measurement data includes altitude change rate, indicated air velocity, X-axis acceleration, Y-axis acceleration, Z-axis acceleration, pitch angle, propagation altitude, roll angle, YAW rate and Weight On Wheel (WOW) and the like. The enumerated measurement data is data commonly used in the art to which the present invention pertains, and a detailed description thereof will be omitted.

Correlation table providing unit 120 is a first correlation table and test flight between the measurement data of the test flight (hereinafter referred to as test flight data) obtained by the flight data measurement unit 110 and the actual flight state before the actual flight of the aircraft The correlation table including the second correlation table between the data and the detailed flight status is based on the flight status determination performed by the first, second, and third flight status determination units 130, 140, and 150, and includes a plurality of measurement data. With, can be used.

This correlation table can be created, as in FIG. In this correlation table, the first correlation table is associated with basic flight conditions such as On the Ground, In the Air & Slow Flight (Hover), and In the Air & Fast Flight. Correlation between the measurement data, the second correlation table is Taxi, Take off, Touch down, Forward Flight, Rear Flight, Sideward Correlation between detailed flight status and measurement data such as flight, spot turn flight, climb, descent, level flight, and vertical approach can be shown.

In addition, the basic flight state and the detailed flight state (in FIG. 8, Flight Regimes includes basic and detailed flight states, etc.) may be classified based on the plurality of measurement data listed above, and each of the plurality of measurement data is averaged. (Mean), standard deviation (Std. Dev), upper limit (Upper) and lower limit (Lower) can be calculated, and the basic and detailed flight conditions can be classified using the calculated results.

In FIG. 8, a correlation table including all of the second correlation table related to the detailed flight state as well as the first correlation table related to the basic flight state is illustrated as an example, but the first and second correlation tables may be separately created.

The first flight state determination unit 130 is based on the measurement data of the actual flight obtained by the flight data measurement unit 110 and the first correlation table of the correlation table providing unit 120, ground flight, low-speed / hover flight and It is possible to determine a first flight state of the aircraft during basic flight states, including high speed flight. For example, the first flight state determination unit 130 may include On the Ground, In the Air & Slow Flight (Hover), and In the Air shown in FIGS. 9 to 11. ≪ RTI ID = 0.0 > a < / RTI > fast flight.

The second flight state determination unit 140 based on the measurement data and the second correlation table of the actual flight acquired by the flight data measurement unit 110 within the conditions of the first flight state, the detailed altitude of the aircraft, the movement type and The second flight state can be determined among the detailed flight states associated with the attitude. In this case, the detailed flight state is the same as the detailed flight state of the second correlation table described with reference to FIG. 8, such as slide, takeoff, shortly before landing, forward flight, rear flight, sideward flight and spot turn. Flight, ascent, descent, altitude maintenance and up and down approaches.

In this case, for example, the second flight state is determined within the conditions of the first flight state performed by the second flight state determination unit 140. As shown in FIGS. 9 and 10, the first flight state Second flight status plate among detailed flight conditions (Forward slight, Sideward Flight, Spot turn, etc.) classified in the ground / hover flight conditions (In the Air & Slow Flight (Hover)) determined by the determination unit 130 The government 150 determines the second flight state.

The second flight state determination unit 140 may be implemented using a fuzzy inference technique. The fuzzy inference technique can determine various flight conditions of a rotorcraft in more detail, and contribute to accurate and error-free determination of various flight conditions by excluding the effect of noise caused by the aircraft's transition flight status. The fuzzy inference technique is constructed by learning the fuzzy inference system based on the test flight data and the second correlation table before the actual flight, and inputting the measurement data during the actual flight into the fuzzy inference system to determine the second flight state. It can proceed to.

Referring to FIG. 2, a detailed configuration of performing a fuzzy inference technique in the second flight state determination unit 140 will be described. 2 is a block diagram showing a detailed configuration of a second flight state determination unit. The second flight state determination unit 140 may include a fuzzy rule generator 142, a fuzzy inference system component 144, and a defuzzy unit 146.

The fuzzy rule generator 142 may generate a fuzzy rule based on the test flight data and the second correlation table. The fuzzy rule can be written as a table in the form of an If / Then rule, for example If: (ALT_RATE is MF1) AND (IND_AIRSPD is MF2) AND ..., THEN : (On the Ground is MF1) AND (Light on Wheel is MF2) OR ...

The fuzzy inference system component 144 may fuzzy the fuzzy rule to learn and configure a fuzzy inference system between test flight data and detailed flight conditions. The fuzzy inference system can be generated using a membership function and MAMDANI inference technique suitable for system configuration. The fuzzy inference system configured in the fuzzy inference system configuration unit 144 receives the measurement data during the actual flight and outputs the result to the depurator 146.

The deferr unit 146 may determine the second flight state using the learned fuzzy inference system, and in detail, calculates the detailed flight state corresponding to the measurement data within the conditions of the first flight state and determines the second flight state. Can be. The depurator 146 may determine the second flight state by applying a centroid technique to a result output from the fuzzy inference system.

Meanwhile, the third flight state determination unit 150 may determine the third flight state among abnormal flight states related to rapid flight control based on the measurement data at the time of the actual flight acquired by the flight data measurement unit 110. . In this case, the abnormal flight conditions include the sudden acceleration, the sudden deceleration, the sharp rise, the descent, the sudden avoidance maneuver, the sudden stop, the pullout_rolling, the pullout_symmetric indicated at the rightmost side in FIGS. 10 and 11. , Sideslip, and collective pop-up.

The third flight state determination unit 150 may be implemented using a neural network technique. The neural network technique can determine the various flight conditions of the rotorcraft in more detail, and in particular, contributes to the accurate and error-free determination of various flight conditions by excluding the effects of noise caused by the aircraft's transition flight status. The neural network technique may be configured by learning neural network circuits based on test flight data before actual flight, and inputting measurement data during actual flight into the learned neural network circuit to determine a third flight state.

Using the neural network technique, the third flight state determination unit 160 is configured by the neural network DB unit 160 having a database used to construct a neural network between the test flight data before the actual flight of the aircraft and the abnormal flight state. You can learn by constructing neural networks.

By the above-described database, the neural network may be a fed forward neural network having, for example, an input layer defined as measurement data, an output layer defined as an abnormal flight state, and a hidden layer, as shown in FIG. 3.

In addition, the relational expression used in the hidden layer may be [Equation 1] below.

[Equation 1]

는 입력층 뉴런,

는 출력층 뉴런,

는 가중치,

는 활성함수, N은 뉴런의 수이며 첨자 i는 선행층, j는 현재층을 의미한다)(From here

Input layer neurons,

Is the output layer neuron,

Is the weight,

Is the active function, N is the number of neurons, subscript i is the preceding layer, j is the current layer)

In this case, the activity function is a tangent sigmoid function, which may be represented by Equation 2 below.

&Quot; (2) "

On the other hand, the output unit 170 outputs the comprehensive flight state determined by the above-described first, second, third flight state determination unit 130, 140, 150 to provide to the pilot, control tower and the like.

Hereinafter, referring to FIG. 4, a method for recognizing a flight pattern of a rotorcraft aircraft according to an exemplary embodiment of the present invention operated by the flight pattern recognition apparatus 100 will be described in detail. 4 is a flowchart illustrating a method of recognizing a flight pattern according to an embodiment of the present invention.

First, the flight data measuring unit 110 monitors the flight data in the actual flight of the rotorcraft to obtain a plurality of measurement data (S400). As shown in Figs. 7, 10, and 11, the measurement data includes altitude change rate, indicated air speed, X-axis acceleration, Y-axis acceleration, Z-axis acceleration, pitch angle, propagation altitude, Roll angle, YAW rate, and weight on wheel (WOW) and the like.

Simultaneously with the acquisition of the measurement data, the correlation table providing unit 120 performs the first correlation table and the test flight data and detailed flight between the test flight data and the basic flight state acquired by the flight data measurement unit 110 before the actual flight of the aircraft. A second correlation table between states can be created. Such a correlation table may be created including the first and second correlation tables, for example, as shown in FIG. 8.

Next, the first flight state determination unit 130 may include ground flight, low speed / hover flight, and high speed flight based on the measurement data and the first correlation table during actual flight acquired by the flight data measurement unit 110. It is possible to determine the first flight state of the aircraft among the basic flight states (S410). As a result, the first flight state may be determined as one of the basic flight states shown on the leftmost side, as shown in FIGS. 10 and 11.

Subsequently, when the first flight state is determined to be one of ground flight, low speed / hover flight, and high speed flight, the second flight state determination unit 140 may select among the detailed flight states within the determined first flight condition. The second flight state is determined (S420). In the case where the determination of the second flight state is made by the fuzzy inference technique, the measurement of the second flight state is inputted to the fuzzy inference system that has already been learned and configured in the fuzzy inference system configuration unit 144 so that the deferred unit 146 Through the second flight state is determined.

As described above, when the second flight state is determined by the learned fuzzy inference system, the fuzzy inference system is configured through the process as shown in FIG. 5. 5 is a flowchart for configuring a learned fuzzy inference system.

The flight measurement data unit 110 acquires test flight data during the flight test process (S500). Subsequently, the first flight state determination unit 120 classifies the ground flight, the low speed / hover flight, and the high speed flight based on the obtained test flight data and the first correlation table of the correlation table providing unit 120 (S510).

According to the classified result, the fuzzy rule generator 142 constructs a matrix by creating a second correlation table between the test flight data and the detailed flight status (S520a, S520b, and S520c). Next, the fuzzy rule generator 144 generates a fuzzy rule according to the classified result based on the correlation table composed of the matrix (S530a, S530b, S530c). For example, the fuzzy rule may be prepared as a table in the form of If / Then rule, and the preparation of the fuzzy rule is the same as that of the fuzzy rule generator 142 described with reference to FIG. 2.

Next, the fuzzy inference system configuration unit 144 fuzzy the fuzzy rule and configured to learn the fuzzy inference system between the test flight data and detailed flight status for each classified results (S540a, S540b, S540c). The fuzzy inference system can be generated using a membership function and MAMDANI inference technique suitable for system configuration.

The learned fuzzy inference system for each of the classified results is provided at the time of determination of the second flight state for determining the second flight state in the actual flight (S550).

Referring to FIG. 4 again, the process after the determination of the second flight state (S420) will be described. The third flight state determination unit 150 is based on the measurement data during the actual flight acquired by the flight data measurement unit 110. In operation S430, the third flight state is determined among the abnormal flight states associated with the rapid flight control. These abnormal flight conditions include the sudden acceleration, deceleration, sudden rise, descent, sudden avoidance maneuver, sudden stop, pullout_rolling, pullout_symmetric, and side slip shown at the rightmost side in FIGS. 10 and 11. (sideslip) and collective pop-ups.

Taking the result determined by the above-described process as an example, when the first flight state is a condition of low speed / hover flight, and the second flight state is sideword flight and it is determined that the flight is abnormal, the rotorcraft is quick stop. It may be determined to any one of the right kick-out & Acceleration_Right and the left kick-out & Acceleration_Left.

Meanwhile, a neural network technique may be applied to the determination of the third flight state made by the third flight state determiner 150. Accordingly, the third flight state determination unit 150 may determine the third flight state based on the database of the neural network DB unit 160 provided for the neural network configuration between the test flight data and the abnormal flight state. Referring to FIG. 6, a process related to the configuration of the database made by the neural network DB unit 160 and the neural network made by the third flight state determination unit 150 will be described. 6 is a flowchart for constructing a learned neural network.

The neural network DB unit 160 configures a database between the test flight data obtained by the flight data measuring unit 110 and the abnormal flight state before the actual flight of the aircraft (S600).

Thereafter, the third flight state determination unit 150 constructs a neural network using a database (S610). For example, as shown in FIG. 3, the neural network may be a fed forward neural network having an input layer defined as measurement data, an output layer defined as an abnormal flight state, and a hidden layer.

 In order to use the configured neural network, it is provided at the time of determination of the third flight state (S620).

Referring to FIG. 4 again, the process after the determination of the third flight state (S430) will be described. When the first, second, and third flight states are determined, the output unit 190 may include the first and the second flights. And outputs a comprehensive flight state of the rotorcraft including a third flight state to a user such as a pilot and a control tower (S440).

According to the present embodiment, based on the plurality of measurement data, not only the basic state of the aircraft, but also various detailed flight states, that is, the abnormal flight related to the detailed altitude, movement type and attitude of the aircraft, and the rapid flight control You can see the status and output the correct flight pattern.

In addition, the detailed flight conditions and abnormal flight conditions may be determined using fuzzy inference and neural network techniques, thereby minimizing the influence of noise generated in the aircraft's transitional flight state. Can be provided to the user.

On the other hand, the program for implementing the flight pattern recognition method according to the present embodiment is stored in a memory of a computer controlling the rotorcraft, read by the CPU or the like to develop the above-described flight pattern recognition method, the program is a flight pattern Can function as a recognition processor.

In addition, the program does not necessarily need to be stored in the memory, and the computer may read and execute the program stored in a storage medium such as a CD-ROM. In addition, the program is stored in another computer (or server) connected to the flight pattern recognition apparatus 100 via a public line, or the like, so that the flight pattern recognition apparatus 100 can read and execute the program from them. You may.

Although the present invention has been described in detail through the representative embodiments, those skilled in the art to which the present invention pertains can make various modifications without departing from the scope of the present invention. Will understand. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by all changes or modifications derived from the claims and the equivalent concepts as well as the following claims.

100: flight pattern recognition device 110: flight data measurement unit
120: correlation table providing unit 130: first flight state determination unit
140: second flight state determination unit 150: third flight state determination unit
160: neural network DB unit 170: output unit

Claims (10)

A method for recognizing a rotorcraft flight pattern automatically performed by a computer,
Measuring flight data of the aircraft to generate a plurality of measurement data,
Determining a first flight state of the aircraft among basic flight states including ground flight, low speed / hover flight, and high speed flight based on the measurement data;
Determining a second flight state among detailed flight states related to a detailed altitude, a movement type, and a posture of the aircraft based on the measurement data within the conditions of the first flight state,
Determining a third flight state among abnormal flight states associated with rapid flight control within the condition of the second flight state, and
Including the first, second and third flight states,
The generating of the measurement data further includes providing a first correlation table between test flight data and a basic flight state and a second correlation table between test flight data and a detailed flight state, wherein the first flight state is determined. And determining the second flight state based on the measurement data and the first correlation table, and determining the second flight state based on the measurement data and the second correlation table. The method of claim 1,
The measurement data includes altitude change rate, indicated airspeed, X-axis acceleration, Y-axis acceleration, Z-axis acceleration, pitch angle, propagation altitude, roll angle, YAW rate and weight on wheel Recognition method of the flight pattern of a rotorcraft aircraft, including). The method of claim 1,
The detailed flight conditions include a rotorcraft aircraft including slide, takeoff, shortly before landing, forward flight, rearward flight, sideward flight, spot turn flight, ascending, descending, altitude maintenance and up and down approach. Recognition method of flight patterns. The method of claim 1,
The abnormal flight conditions include sudden acceleration, sudden deceleration, sudden rise, descent, sudden avoidance maneuver, sudden stop, pullout_rolling, pullout_symmetric, sideslip, and collective popup. -up) method for recognizing flight patterns of a rotorcraft. delete The method of claim 1,
The determining of the second flight state may be performed using a fuzzy inference technique. The method according to claim 6,
The determining of the second flight state
Using the fuzzy inference system in which the fuzzy rule generated between the test flight data and the detailed flight state is fuzzy, the detailed flight state corresponding to the measurement data is calculated within the conditions of the first flight state, and thereby the second flight state. Recognizing a flight pattern of a rotorcraft aircraft comprising the step of determining. The method of claim 1,
The determining of the third flight state may include recognizing a flight pattern of a rotorcraft aircraft using a neural network technique. A recording medium having recorded thereon a program for recognizing a flight pattern of a rotorcraft,
Measuring flight data of the aircraft to generate a plurality of measurement data,
Determining a first flight state of the aircraft among basic flight states including ground flight, low speed / hover flight, and high speed flight based on the measurement data;
Determining a second flight state among detailed flight states related to a detailed altitude, a movement type, and a posture of the aircraft based on the measurement data within the conditions of the first flight state,
Determining a third flight state among abnormal flight states associated with rapid flight control within the condition of the second flight state, and
Outputting a comprehensive flight status of the aircraft including the first, second and third flight status to a computer,
The generating of the measurement data further includes providing a first correlation table between test flight data and a basic flight state and a second correlation table between test flight data and a detailed flight state, wherein the first flight state is determined. The step is determined based on the measurement data and the first correlation table, and the step of determining the second flight state is a computer readable recording medium having recorded a program determined based on the measurement data and the second correlation table. . A flight data measurement unit configured to measure flight data of the rotorcraft to generate a plurality of measurement data,
A first flight state determination unit that determines a first flight state of the aircraft among basic flight states including ground flight, low speed / hover flight, and high speed flight based on the measurement data;
A second flight state determination unit that determines a second flight state among detailed flight states related to detailed altitude, movement type, and attitude of the aircraft based on the measurement data within the conditions of the first flight state,
A third flight state determination unit that determines a third flight state among abnormal flight states related to rapid flight control within the condition of the second flight state, and
And an output unit configured to output a comprehensive flight state of the aircraft including the first, second, and third flight states.
The flight data measuring unit may further include a first correlation table between test flight data and a basic flight state and a second correlation table between test flight data and a detailed flight state, and the first flight state determination unit determining the first flight state And a second flight state determination unit is determined based on the measurement data and the first correlation table, and is determined based on the measurement data and the second correlation table.

Images