JP2015024705A - Automatic landing/taking-off control method of small electric helicopter - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an automatic landing/taking-off control method of a small electric helicopter which stabilizes motion in altitude and horizontal direction of a small electric helicopter at landing and taking off.SOLUTION: In an automatic landing/taking-off control method of an automatic helicopter, horizontal acceleration is estimated by using a steady kalman filter 11 from an output of an acceleration sensor mounted on the small electric helicopter, and by using the estimated value, performs an acceleration feedback control for stabilizing horizontal movement at the time of landing and taking-off. The ground altitude and a speed in altitude direction are estimated by using a kalman filter from an output of an ultrasonic sensor mounted on the small electric helicopter, and the estimated values are used for altitude control at landing and taking-off.

Description

  The present invention relates to an automatic take-off and landing control method for a small electric helicopter that is a small unmanned aerial vehicle.

  Over the past decade, unmanned aerial vehicle (UAV) technology has made tremendous progress, and UAVs are increasingly being applied to practical missions as well as research and development sites. . UAV is superior in cost, safety and convenience compared to manned aircraft, and its application range is expected to expand further in the future.

  When applying UAV to an actual mission, autonomous control technology is essential from the viewpoint of reducing the burden on the operator, and research on UAV autonomous control has been conducted all over the world against this background (Non-Patent Document 1, 2). In recent years, a small UAV with a weight of about 10 kg or less has attracted a great deal of attention because of its ease of transportation and handling, and the research object of autonomous control has been steadily decreasing year by year (Non-patent Document 3). ).

  The research group of the present inventors is conducting research on autonomous control of a small unmanned helicopter, which is a kind of small UAV, and aims to make the autonomous electric helicopter of about 2 kg that can be carried with one hand. (In this specification, an unmanned electric helicopter having a weight of about 10 kg or less is referred to as a “small electric helicopter.”) Autonomous control of a small machine is much more difficult than control of a medium or larger machine. This is because in addition to being easily affected by disturbances due to its small size, the limit on the weight that can be mounted is severe, so that the sensors used for control are significantly limited. In fact, there is a problem that a sensor that can be mounted on a 2 kg electric helicopter does not have sufficient accuracy for use in autonomous control.

  In the autonomous control system of a small electric helicopter developed by previous research by the present inventors, highly accurate aircraft attitude, position, and velocity are estimated by integrating multiple low-precision sensor data (combined navigation system) Using the obtained estimation data, the helicopter attitude control, hovering control, guidance control and the like have been successfully performed (Non-Patent Documents 4 and 5).

  The prior art documents (non-patent documents 1 to 12) related to the present invention are listed below.

Oh, C.K. and Barlow, G.J., “Autonomous controller design for unmanned aerial vehicles using multi-objective genetic programming,” in Proceedings of Congress on Evolytional Computation 2004, 2004, Vol. 2, pp. 1538-1545. Eric, N. Johnson, Suresh, K. Kannan, “Adaptive Trajectory Control for Autonomous Helicopters,” in AIAA Journal of Guidance Control and Dynamics, Vol. 28, No. 3, pp. 524-538, 2005. Shaojie Shen, Nathan Michael, Vijay Kumar, “Autonomous Multi-Floor Indoor Navigation with a Computationally Constrained MAV,” in Proceedings of IEEE International Conference on Robotics and Automation 2011, TuA101.4, 2011. Satoshi Suzuki and Kenzo Nonami, “Nonlinear Adaptive Control for Small-Scale Helicopter,” in Journal of System Design and Dynamics, Vol. 5, No. 5, pp.866-880, 2011. Satoshi Suzuki, “Low Accuracy Sensor-based Navigation and Fully Autonomous Guidance Control of Small Electric Helicopter,” in Proceedings of IEEE International Conference on Robotics and Automation 2011, ThA101.4, 2011. R.M.Rogers, Applied Mathematics in Integrated Navigation System Second Edition, AIAA Education Series, 2003. Takayasu Sakai, Kazunobu Kajimura, Kenji Niimi, “Altitude Measurement Error by Barometric Altimeter and Its Correction,” Research Report of Electronic Navigation Research Institute, No. 114, 2005. Kenzo Nonami, Hidekazu Nishimura, Fundamentals of Control Theory with MATLAB, Tokyo Denki University Press, 1998. Kenzo Nonami, Farid Kendoul, Satoshi Suzuki, Wei Wang and Daisuke Nakazawa, Autonomous Flying Robots Unmanned Aerial Vehicles and Micro Aerial Vehicles, Springer 2010. Jack B. Kuipers, Quaternions and Rotation Sequences, Princeton University Press, 2002. Tajima Hiroshi, Basics of Multibody Dynamics-How to Establish 3D Equations of Motion, Tokyo Denki University Press, 2006. Kanichiro Kato, Yuki Imanaga, Introduction to Helicopter, University of Tokyo Press, 1985.

  As described above, autonomous flight of a small electric helicopter can be realized by the autonomous control system developed by the research group of the present inventors. However, the current situation is that automation of helicopter takeoff and landing has not yet been realized. The biggest reason why take-off and landing cannot be automated is that the GPS and barometer installed in the helicopter have errors during take-off and landing.

  With respect to GPS, errors in positioning data increase due to a phenomenon called multipath in which radio waves are reflected by trees and buildings near the ground. On the other hand, regarding the barometer, the air pressure is increased between the rotor and the ground by blowing down the main rotor, so that the atmospheric pressure rises and a large error estimated based on the atmospheric pressure occurs. For this reason, it is considered difficult to apply the conventional autonomous control system as it is to the automatic take-off and landing of a small electric helicopter.

  An object of the present invention is to propose a new control method that can stabilize altitude and horizontal movement of a small electric helicopter during take-off and landing and that does not use a GPS and a barometer.

In order to solve the above problems, an automatic take-off and landing control method for a small electric helicopter of the present invention includes:
A small electric helicopter is equipped with an acceleration sensor,
From the output of the acceleration sensor, using a stationary Kalman filter, the horizontal acceleration of the small electric helicopter is estimated,
Using the estimated horizontal acceleration, acceleration feedback control is performed to stabilize horizontal movement during take-off and landing of the small electric helicopter,
The stationary Kalman filter is designed using the following equation of state:

In the state equation,
a _x and a _y are forward and rightward accelerations of the aircraft,
p and q are roll, pitch angular velocity,
a to d are flap angles of the main rotor and the stabilizer,
L _b and M _a are appropriate factors determined by the center of gravity of the fuselage, the geometrical arrangement of the rotor head, the bending rigidity of the rotor blades,
I _{xx and} I _yy are the moments of inertia of each axis,
τ _b and τ _s are time constants related to the flapping motion of the main rotor and stabilizer,
K _{1 to} K ₃ are coefficients determined by the link mechanism around the rotor, and
δ _air , δ _ele represent aileron and elevator steering inputs,
It is characterized by that.

Moreover, the automatic take-off and landing control method of the small electric helicopter of the present invention is
The small electric helicopter is equipped with an ultrasonic sensor whose measurable distance is at least the diameter of the main rotor,
From the output of the ultrasonic sensor, the ground altitude and altitude direction speed of the small electric helicopter are estimated using a Kalman filter,
Perform altitude control during takeoff and landing using the estimated ground altitude and altitude direction speed,
The Kalman filter is designed using the following equation of state:

In the above state equation, h _g corresponds to the ground altitude, v _d corresponds to the vertical downward velocity, b _a corresponds to the bias error of the acceleration sensor output, and r ₃ corresponds to the row of the coordinate transformation matrix from the body fixed coordinate system to the ground fixed coordinate system. The vector ^ab represents the output vector of the acceleration sensor, and g represents the gravitational acceleration.

In the automatic take-off and landing control method of the present invention, automatic take-off and landing control is performed by acceleration feedback control and altitude control using an ultrasonic sensor. According to the present invention, it has been confirmed by experiments that the problem that a large error occurs when the sensor used for autonomous control makes automatic takeoff and landing can be solved. Therefore, by using the method of the present invention, it is possible to construct a fully autonomous control system including automatic take-off and landing control of a small electric helicopter.

It is an image which shows an example of the small electric helicopter which is the control object of the present invention. It is a graph which shows the positioning data of a horizontal position. It is a graph which shows altitude positioning data. It is a block diagram which shows a guidance control system. It is explanatory drawing which shows a coordinate system. It is a graph which shows the result of a hovering control experiment. It is a graph which shows the altitude data estimated using the compound navigation system. It is a block diagram which shows the acceleration feedback control system by this invention. It is a graph which shows the result of a hovering flight experiment. It is a graph which shows the estimation result of the ground height by this invention. It is a graph which shows the estimation result of the altitude direction speed by this invention.

  With reference to the drawings, an embodiment of an automatic take-off and landing control method for a small electric helicopter (hereinafter sometimes simply referred to as “helicopter”) to which the present invention is applied will be described. In the following, after describing the outline of the autonomous control system for small electric helicopters developed in the previous research, we will clarify the problems of the current system for realizing automatic take-off and landing and solve the problems The control system will be described. Specifically, first, an example of the experimental system of the present invention including a small electric helicopter will be described, and then an outline of the autonomous control system consisting of a composite navigation system and a guidance control system constructed in the previous research And its effectiveness. Next, the problems of the current system for realizing automatic takeoff and landing will be described, the control system of the present invention as a solution will be designed, and its effectiveness will be verified by experiments. Finally, I will give a conclusion.

[Example of experimental system]
A small electric helicopter to be controlled by the method of the present invention is, for example, Lepton (registered trademark) -Ex (trade name), and its overall image is shown in FIG. 1, and main specifications are shown in Table 1 (Table 1). Each is shown. The Lepton-Ex is a high maneuvering helicopter developed for hobbies and can output a large thrust as compared with its own weight, so it can perform not only static flight such as hovering but also aerobatics.

Table 2 (Table 2) shows the configuration of the control device mounted on the small electric helicopter (Lepton-Ex). The control device includes an FPGA board, a wireless module, a small attitude sensor, a GPS module, and a barometer. A feature of this control device is that only a light sensor of about several tens of grams is used so that it can be mounted on a 2 kg small electric helicopter. In general, the accuracy and weight of a sensor are proportional to each other, and the lighter the sensor, the lower the accuracy. In particular, a small GPS module used for helicopter position / velocity measurement does not have sufficient accuracy to be used for autonomous control of a small helicopter, and it is necessary to improve accuracy by integrating a plurality of sensor data.

[Autonomous control system for small electric helicopters]
Next, as an outline of the autonomous control system of a small electric helicopter developed by the research group of the present inventors, the design of a composite navigation system and a guidance control system will be described.

(Composite navigation system)
In consideration of applying an autonomous unmanned helicopter to aerial photography, remote sensing, etc., it is desirable that the hovering accuracy when no wind is at least about 0.5 m in the horizontal direction and about 1 m in the vertical direction. Therefore, the position measurement accuracy of the helicopter needs to be at least higher. However, as described above, the small GPS module used for position / velocity measurement does not have sufficient accuracy for this required specification. Table 3 (Table 3) shows the error of the small GPS module obtained from the comparison test with the true value. From Table 3, it can be seen that both horizontal and vertical data have large errors and it is difficult to use them directly for autonomous control of a small electric helicopter.

  In order to compensate for the weak points of such a small GPS module, a compound navigation system is constructed that corrects errors by using an extended Kalman filter with the basic equations of inertial navigation and atmospheric pressure altitude measurement as process models.

First, the state equation and observation equation of the system are derived. Now, _assuming that v _n , v _e , and v _d are the speeds of a small electric helicopter facing north, east, and vertically downward, respectively, from the basic equation of inertial navigation (Non-Patent Document 6), their time derivatives are as follows: It is represented by the formulas (1) to (3).

Here, a ^b vector output vector from the acceleration sensor mounted on the machine body, r ₁ ~r ₃ is corresponding to each row of the coordinate transformation matrix to terrestrial fixed coordinate system from the body-fixed coordinate system, Omega is the earth's rotation angular velocity , L is the latitude of the current position of the helicopter, and g is the gravitational acceleration.

  Generally, an acceleration sensor output includes a bias error, which becomes a big problem when performing inertial navigation calculation. Therefore, the error estimation is performed by correcting the bias error dynamics of the acceleration sensor by assuming the following equation (4) and incorporating the dynamics into the extended Kalman filter.

Here, w _{x to} w _z are white noise.

Now, when the vector b _a = [b _ax ; b _ay ; b _az ] ^T is introduced and the equations (1) to (3) are rewritten into equations including bias error correction, the following equations (5) to (7) It becomes.

  Subsequently, the time differential of the latitude L, longitude λ, and altitude h of the current position of the electric small helicopter can be expressed by the following equations (8) to (10).

Here, R _n and R _e represent the polar radii in the north-south direction and the east-west direction of the earth, respectively.

  Furthermore, from the basic equation used for atmospheric pressure altitude measurement (Non-Patent Document 7), the relationship between the time differential of the atmospheric pressure P and the vertical speed can be expressed by the following equation (11).

However, R represents a gas constant, T represents temperature, and M represents the molar mass of the atmosphere.

When the equations (4) to (11) are put together, the following nonlinear equation of state is obtained.

Next, the observation equation is derived. Since the observation amount of the system is the latitude, longitude, altitude and horizontal pressure that can be obtained from the GPS, and the atmospheric pressure measured using a barometer, the output vector is y = [L l h v _n v _e P] If ^T , then the observation equation is:

In the equation, v is a vector representing observation noise.

  A compound navigation system was constructed by applying the extended Kalman filter algorithm to the process model obtained above, and a verification experiment was conducted. The results of the verification experiment are shown in FIGS. In this experiment, all equipment was mounted on a cart, and the position and speed were estimated while moving. Moreover, RTK-GPS is mounted on the same vehicle, and the RTK-GPS positioning data is handled as a true value. FIG. 2 shows the positioning data at the horizontal position, and FIG. 3 shows the positioning data at the altitude. In each figure, the broken line is the positioning data of the small GPS module, the solid line is the estimated value estimated by the navigation system, and the chain line is the true value obtained using RTK-GPS.

  From each figure, the data of the small GPS module has a large error of about 3m in the horizontal direction and about 10m in the altitude direction with respect to the true value, and does not satisfy the above-mentioned required specifications at all. The estimated data enables accurate position estimation with an error of about 0.3 m at the maximum in both the horizontal and altitude directions, and shows the effectiveness of the constructed composite navigation system.

(Guidance control system)
Next, a guidance control system for guiding the small electric helicopter to an arbitrary target point using the estimated data obtained by the composite navigation system is constructed. As shown in FIG. 4, the guidance control system 1 includes a guidance controller 3 that generates an acceleration vector necessary for guidance of the small electric helicopter 2, and a rotor collective pitch angle for realizing a desired acceleration vector using the rotor thrust. And a thrust generator 4 for calculating the target body posture, and a posture control system 5 for causing the body posture to follow the calculated target posture. Estimated data is input to the guidance controller 3 from the compound navigation system 6 described above. Since the attitude control system 5 has already been constructed and verified in the previous research (Non-Patent Document 4), the guidance controller and the thrust generator are designed below.

  First, a translational motion model of a small electric helicopter is derived. Now, when the target acceleration in each axis direction in the three-dimensional space is used as the system input, the dynamics from the input to the aircraft speed are considered as follows, taking into account the damping of the air resistance and the delay of the attitude control system of the inner loop. Approximation is performed using a third-order transfer function expressed by Expression (17).

Here, the parameters in the transfer function are all the same for each axis.

Next, an induction controller is designed using the obtained model. First, when the target speed of each axis calculated by proportional control using the deviation between the target point and the current position is set to v _nd, v _ed, v _dd , an optimal controller for causing the body speed to follow the target speed To design. Realization of the transfer function of Equation (17) was sought, a servo expansion system (Non-Patent Document 8) was constructed, and a state feedback controller was designed by applying optimal control theory to the expansion system. Now, the state quantity of the servo expansion system
In this case, the state feedback control is expressed by the following equation (18).

Here, a _id is a target acceleration, F _i is a state feedback gain, and subscripts in the formula indicate components in the north direction, the east direction, and the vertically downward direction.

Ei represents the integral of the deviation between the target speed and the aircraft speed expressed by the following equation (19).

Here, the state quantities that cannot be directly observed during the experiment are estimated by the same-dimensional observer. Finally, a vector a _d = [a _nd ; a _ed ; a _dd + g] ^T constructed using the generated target acceleration of each axis
Then, a thrust generator that realizes the target acceleration of each axis represented by this vector using each axis component of the rotor thrust is designed. The structure of the small electric helicopter, by setting the inclination and the size of the rotor thrust appropriately, it is possible to implement any of a _d. Therefore, the rotor collective pitch angle determines the magnitude of the thrust, by obtaining the aircraft attitude to determine the slope, it can be realized a _d.

Now, when the rotor collective pitch angle and theta _pit, by utilizing that the relationship of theta _pit and rotor thrust under the assumption that the rotor rotational speed is constant is linear (non-patent document 9), a _d The relationship between the norm of the motor and the rotor collective pitch angle for realizing it is expressed by the following equation (20) with K _pit as an appropriate gain.

  Subsequently, the target body posture is obtained. In the previous research (Non-Patent Document 4), since the quaternion (Non-Patent Document 10) is used for the attitude representation of the small electric helicopter, the quaternion representing the target body attitude is calculated here. In order to facilitate the calculation, a coordinate system shown in FIG. 5 is defined. First, N-frame is a coordinate system having an origin at the center of gravity of the small electric helicopter, the Zn axis being in the gravity direction, and the Xn axis being in the true north direction. Subsequently, R-frame is a coordinate system in which N-frame is rotated by a target azimuth angle yd [rad] around the Zn axis.

If the quaternion representing the attitude of the R-frame with respect to the N-frame is q ^r _n , this can be calculated by the following equation (21).

a _d is expressed as a vector on the N-frame, which when a material obtained by coordinate conversion on the R-frame and ^a _{r d,} the relationship of _{a d} and ^a _{r d} is using ^q _{r n} the following formula (22)

Here, if we define a vertically upward reference vector as
T _d = [0 0 −K _pit || a _d ||] ^T
T is the rotation axis lambda _d and the rotation angle phi _d of Simple Rotation (Non-Patent Document 11) representing the rotational position required to achieve _{a d} by _d using _{T d} and ^a _{r d} the following equation (23) Can be calculated.

Further, the quaternion q _r ^d representing the rotation posture similar to this Simple Rotation is expressed by the following equation (24).

However, s (x) = sinx, c (x) = cosx, and λdx−λdz represents each axis component in the R-frame of λd.

By using the above q ^d _r and q ^r _n , the quaternion q _d representing the target posture for realizing a _d can be calculated by the following equation (25).

  FIG. 6 shows the results of the hovering control experiment performed using the equations (18), (20) and (25) obtained above and the attitude control system of the previous study. During the experiment, there was almost no wind, and the hovering flight was performed with the origin set at a height of about 5 m above the ground as the target point. The solid line in the figure shows a three-dimensional flight trajectory of about 2 minutes. Ninety percent of the flight trajectory is within a radius of 0.5m centered on the origin, and we were able to obtain results that satisfactorily met the required specifications for autonomous unmanned helicopter operation.

[Automatic takeoff and landing control]
Next, we will design a control system for realizing automatic take-off and landing of a small electric helicopter. As described above, the autonomous flight of a small electric helicopter was realized using a compound navigation system and a guidance control system. However, the current situation is that automation of takeoff and landing has not yet been realized. The biggest reason why take-off and landing cannot be automated is that, as described above, the GPS and barometer installed in the helicopter have an error during take-off and landing. This is because with respect to GPS, errors in positioning data increase due to a phenomenon called multipath in which radio waves are reflected by trees and buildings near the ground. As for the barometer, the air is compressed between the rotor and the ground by blowing down the main rotor, so that the atmospheric pressure rises and a large error estimated based on the atmospheric pressure occurs.

  Fig. 7 shows the altitude data estimated using the compound navigation system when the helicopter actually took off. Although the small electric helicopter is taking off in the vicinity of 5 to 10 seconds in FIG. 7, it can be seen that the altitude data is reduced for the reason described above. From the above, it is considered difficult to apply the above-mentioned autonomous control system as it is to automatic take-off and landing. Therefore, in this invention, the new control system which does not use GPS and a barometer is proposed, and it aims at stabilizing the altitude and horizontal motion of a helicopter at the time of takeoff and landing.

(Acceleration feedback control)
Since GPS cannot be used, only the acceleration sensor can be used to stabilize the horizontal movement of the small electric helicopter. Here, the horizontal movement of the small electric helicopter is stabilized using acceleration feedback control.

In general, the output of an acceleration sensor contains a large amount of noise and is difficult to use directly for feedback control. Therefore, the horizontal acceleration is estimated using a Kalman filter based on a motion model of a small electric helicopter. First, a process model used for filter design is derived. To simplify the model, we make several assumptions:
(1) The roll and pitch angle of the helicopter are sufficiently small, and the yaw angular velocity is stabilized and can be regarded as zero.
(2) The flap angle of the main rotor and stabilizer is sufficiently small.
(3) The magnitude of the thrust of the main rotor is equal to its own weight in a steady state such as hovering.

First, regarding the first assumption, the attitude inclination during take-off / landing and hovering is very small, and the yaw direction is considered to be a reasonable assumption because it is stabilized by a stabilizing device called a rate gyro. Regarding the second assumption, it can be said that the flap angle of the rotor and the stabilizer is about several degrees at most, so it is still valid. The third assumption is self-evident.

Under these assumptions, the horizontal acceleration applied to the small electric helicopter is expressed by the following equations (26) and (27).
Here, a _x and a _y are aircraft forward and rightward acceleration, T _mr main rotor thrust, m is the mass of the helicopter, f and q represent the roll and pitch angles, respectively.

Subsequently, the relationship between the roll and pitch angles and the angular velocities in each direction is expressed by the following equations (28) and (29).
Here, p, q, and r are roll, pitch, and yaw angular velocity, respectively.

  The following state equations (30) and (31) can be obtained by combining the rotary motion model of the small unmanned helicopter derived in the previous research (Non-patent Document 9) with the formulas (26) to (29). .

Here, a to _d are flap angles of the main rotor and the stabilizer, L _b and M _a are appropriate coefficients determined by the center of gravity of the fuselage, the geometric arrangement of the rotor head, the bending rigidity of the rotor blades, etc., and I _xx and I _yy are Moment of inertia of each axis, τ _b and τ _s are time constants related to the flapping motion of the main rotor and stabilizer, K _{1 to} K ₃ are coefficients determined by the link mechanism around the rotor, δ _ali and δ _ele are aileron and elevator steering inputs Represents.

  A stationary Kalman filter was designed using the equation of state obtained here. As shown in FIG. 8, the acceleration feedback control system 10 is configured by combining the coordinate conversion 12 of acceleration and the PI controller 13 with the steady Kalman filter 11.

  FIG. 9 shows the result of a hovering flight experiment performed using only the acceleration feedback control by the acceleration feedback control system 10. The solid and broken lines in the figure represent the horizontal speed during the experiment, and are obtained using the autonomous navigation system described above. FIG. 9 shows that the horizontal speed can be stabilized and the designed acceleration feedback is effective to some extent for stabilizing the horizontal movement of the small electric helicopter.

(Advanced control using ultrasonic sensors)
Since it is difficult to use GPS and a barometer for altitude measurement at the time of automatic takeoff and landing, a sensor for measuring the ground altitude is required separately. Necessary sensor specifications include a sufficiently small size and light weight. In addition, since it is known that the problem of barometers generally occurs when the distance between the rotor surface of the helicopter and the ground is less than the diameter of the main rotor (Non-patent Document 12), it is possible to measure at least more distances. Must. In the case of this small electric helicopter, a measurable distance of 1 m or more is required. In consideration of the above specifications, an ultrasonic sensor was selected this time. The specifications of the selected ultrasonic sensor are shown in Table 4 (Table 4).

  It can be seen that the selected ultrasonic sensor is small and light, and has a sufficient measurable distance. However, since the ultrasonic sensor has a characteristic that it is easily affected by acoustic noise, a large measurement error occurs due to the effect of rotor wind noise of a small electric helicopter. In addition, for altitude control of a small electric helicopter, not only altitude data but also altitude direction speed is required. Therefore, the influence of acoustic noise is reduced using a Kalman filter, and the altitude direction velocity is also estimated.

The equation of state used for the design of the Kalman filter is shown in the following equation (32).
Here, h _g represents the ground altitude, and b _a represents the bias error of the acceleration sensor output.

  A Kalman filter was designed using this equation of state, and the ground altitude and altitude direction velocity were estimated. For the estimation, data obtained by flying a small electric helicopter at an altitude sufficient to reduce the error between the GPS and the barometer is used. The estimation results are shown in FIGS.

  First, regarding FIG. 10, the solid line represents the ground height estimated by the Kalman filter, and the broken line represents the raw data of the ultrasonic sensor. In the vicinity of 40 seconds and 80 seconds, a large error occurs in the data of the ultrasonic sensor due to the influence of acoustic noise, but the estimated value can reduce the influence of this noise. In FIG. 11, the solid line represents the estimated value of the altitude direction speed, and the broken line represents the altitude speed estimated using the navigation system. The two estimated values almost coincided with each other, indicating that the designed Kalman filter can perform accurate velocity estimation.

  By using the ground altitude and altitude direction speed estimated in this way and the guidance control system described above, altitude control during automatic take-off and landing of a small electric helicopter can be performed.

[Conclusion]
As explained above, we constructed a fully autonomous control system including automatic take-off and landing control for the purpose of autonomy of small electric helicopters. New automatic take-off and landing control including acceleration feedback control and altitude control using ultrasonic sensors to solve the problem that a large error occurs when sensors used for autonomous control perform automatic take-off and landing A method was proposed. Experiments confirmed the effectiveness of the automatic take-off and landing control method according to the present invention.

DESCRIPTION OF SYMBOLS 1 Guidance control system 2 Small electric helicopter 3 Guidance controller 4 Thrust generator 5 Attitude control system 10 Acceleration feedback control system 11 Stationary Kalman filter 12 Acceleration coordinate conversion 13 PI controller

Claims (2)

A small electric helicopter is equipped with an acceleration sensor,
From the output of the acceleration sensor, using a stationary Kalman filter, the horizontal acceleration of the small electric helicopter is estimated,
Using the estimated horizontal acceleration, acceleration feedback control is performed to stabilize horizontal movement during take-off and landing of the small electric helicopter,
The stationary Kalman filter is designed using the following equation of state:
In the state equation,
a _x and a _y are forward and rightward accelerations of the aircraft,
p and q are roll, pitch angular velocity,
a to d are flap angles of the main rotor and the stabilizer,
L _b and M _a are appropriate factors determined by the center of gravity of the fuselage, the geometrical arrangement of the rotor head, the bending rigidity of the rotor blades,
I _{xx and} I _yy are the moments of inertia of each axis,
τ _b and τ _s are time constants related to the flapping motion of the main rotor and stabilizer,
K _{1 to} K ₃ are coefficients determined by the link mechanism around the rotor, and
δ _air , δ _ele represent aileron and elevator steering inputs,
An automatic take-off and landing control method for a small electric helicopter. The small electric helicopter is equipped with an ultrasonic sensor whose measurable distance is at least the diameter of the main rotor,
From the output of the ultrasonic sensor, the ground altitude and altitude direction speed of the small electric helicopter are estimated using a Kalman filter,
Perform altitude control during takeoff and landing using the estimated ground altitude and altitude direction speed,
The Kalman filter is designed using the following equation of state:
In the above state equation, h _g corresponds to the ground altitude, v _d corresponds to the vertical downward velocity, b _a corresponds to the bias error of the acceleration sensor output, and r ₃ corresponds to the row of the coordinate transformation matrix from the body fixed coordinate system to the ground fixed coordinate system. , A ^b is the output vector of the acceleration sensor, g is the gravitational acceleration,
The automatic take-off and landing control method for a small electric helicopter according to claim 1.