JP2006123014A - Inverted two-wheeled robot - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an inverted two-wheeled traveling robot capable of realizing an intuitive interface for causing autonomous operation by positively utilizing the compliance of an aircraft with respect to posture change.
An unstable body 3, a posture sensor 4, actuators 2 a and 2 b that generate a posture-holding force, a posture control unit 7 that stabilizes the posture, and human intentions based on changes in the posture sensor. And a motion control unit 9 that determines an operation mode from the output of the motion variation detection unit 13.
[Selection] Figure 2

Description

  The present invention is an inverted two-wheeled vehicle that realizes an intuitive and easy-to-understand operation instruction interface for humans, among robots that work in coexistence with humans and that balance the originally unstable aircraft by the power of control. It relates to a traveling robot.

  As a conventional automatic machine coexisting with a human who balances an unstable airframe by control, a coaxial two-wheel vehicle Segway (registered trademark) developed in the United States is famous. The basic configuration is shown in FIG. FIG. 16 is described in Patent Document 1.

  In FIG. 16, wheels 100a and 100b are connected by a frame 102, and DC servomotors 104a and 104b are fixed to the frame to drive the wheels. A handle bar 108 is attached to the frame 102 via a stem 106. A box 110 is provided at an appropriate position on the frame 102, and an inclination sensor and a control unit (both not shown) are accommodated therein. Note that switches 112a and 112b are provided on the grip of the handle bar 108, and for example, the DC servo motor 104a is turned to the right when the switch 112a is turned on.

  A person (not shown) rides on the frame 102 while holding the handle lever 108. A control unit (not shown) performs feedback control on the motors 104a and 104b based on the inclination of the frame 102 detected from the box 110 so that the inclination becomes zero. When a person moves his / her body weight on the frame 102 in a desired front-rear direction, the vehicle can travel in that direction along with the balance control performed by the box 110.

  In FIG. 17, a robot 210 is a communication robot that was developed mainly for the purpose of communicating with children, and is intended to be able to perform body communication such as contact and hugging in addition to communication by voice or the like. . The robot 210 prevents the robot 210 body from overturning by controlling the two wheels 218a and 218b mounted on the same axis. Furthermore, in order to prevent a fall due to an unexpected external force due to contact or the like in communication with a child, a hip joint 228 and a hip motor 256 that drives the hip joint 228 are provided. For example, when the collision sensor (not shown) attached to the surfaces of the chests 232 and 240a detects contact, the waist joint 228 moves the waist joint 228 in the pushed direction to release the force. be able to. Further, by connecting the waist joint 228 and the chest joint 262 with the link 264, it is possible to move both joints by one waist motor 256 and to take a larger escape operation. As described above, the combination of the balance control by the wheels 218a and 218b and the escape operation by the hip joint 228 and the chest joint 262 realizes an inversion control that is robust against contact and collision, and is safe for children who communicate with each other. Coexistence is possible.

Japanese Patent No. 3070015 (FIG. 23) Japanese Patent Laid-Open No. 2003-271243 (FIG. 3)

  However, the above-described conventional configuration has a problem that it is impossible to realize an operation instruction interface that is intuitive and easy to understand for humans to robots and automatic devices that are balanced by control.

  That is, in Patent Document 1, it is as intuitive as an automobile accelerator in the sense that a person can control the movement back and forth by moving his / her weight, but he / she must always keep his / her weight in a desired direction. This vehicle cannot be said to have operated autonomously upon receiving instructions.

  Also, in Patent Document 2, the child's contact with the robot causes the movement of the upper body of the robot to escape, but the robot is merely reflexively operated and the contact is removed. It is hard to say that the robot has returned to its original state again and has autonomously moved upon receiving a motion instruction.

  The present invention solves the above-described conventional problems, and realizes an operation instruction interface that is intuitive and easy to understand for humans with respect to a robot of a type that balances an originally unstable body by the power of control. For this reason, by using compliance in the posture holding direction caused by the robot trying to balance, the force that the human acts on the robot is regarded as a change in posture, and the intention of the human is read out from this change pattern .

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide an inverted two-wheeled robot that causes an autonomous operation in response to a human motion instruction that changes posture.

  In order to achieve the above object, the present invention is configured as follows.

In order to solve the above-mentioned conventional problems, according to one aspect of the present invention, two wheels arranged coaxially and respectively connected to two axles, and a vehicle body that rotatably supports the two axles. A fuselage having
An attitude sensor for detecting the attitude of the aircraft,
An actuator for driving each of the two axles to generate a force for maintaining the posture of the airframe;
An attitude control unit for controlling the operation of the actuator so as to stabilize the attitude of the airframe;
A determination unit for determining the human intention based on a change in posture of the aircraft over time from the posture sensor;
An inverted two-wheeled robot having an operation control unit that changes an operation mode based on an output of the determination unit is provided. Therefore, in this configuration, it is possible to recognize the force acting on the robot by the human as a change in the posture of the robot, read the human intention from the change pattern, and perform an autonomous operation.

  With this configuration, it is possible to realize an operation instruction interface for an autonomous robot that is intuitive and easy to understand for humans.

  According to the inverted two-wheeled traveling robot of the present invention, it is possible to give an operation instruction by changing the posture by applying a force by utilizing compliance due to maintaining the posture by control. Become. For example, it is possible to realize an operation instruction that is easy for a human to understand from daily habits, such as “start by pressing the back” or “stop by pressing the chest”. This is passive, such as in the conventional invention, by continuously stepping on the accelerator of the device (conventional example 1), or taking a reflective action with respect to the force applied by humans (conventional example 2). This is to realize an interface that simply instructs an autonomous operation, which is not a typical operation, in an intuitive and easy-to-understand manner.

  Before describing embodiments of the present invention in detail based on the drawings, various aspects of the present invention will be described below.

According to the first aspect of the present invention, two wheels disposed coaxially and respectively connected to two axles, and an airframe having a vehicle body that rotatably supports the two axles,
An attitude sensor for detecting the attitude of the aircraft,
An actuator for driving each of the two axles to generate a force for maintaining the posture of the airframe;
An attitude control unit for controlling the operation of the actuator so as to stabilize the attitude of the airframe;
A determination unit for determining the human intention based on a change in posture of the aircraft over time from the posture sensor;
An inverted two-wheeled robot having an operation control unit that changes an operation mode based on an output of the determination unit is provided.

  According to the second aspect of the present invention, the operation mode changed by the operation control unit includes a standby mode for waiting for the start of travel, a follow-up mode for traveling following the human, and a stop mode for stopping the travel. The inverted two-wheeled traveling robot according to the first aspect is provided.

  According to the third aspect of the present invention, the determination unit determines that the intention of the human body is present if the temporal posture change of the aircraft from the posture sensor exceeds a threshold value, and if it is equal to or less than the threshold value, the human body is determined. The inversion according to the first or second aspect, wherein the operation mode is changed by the operation control unit only when the determination unit determines that the intention is not human intention and the operation control unit determines that the intention is the human intention. A two-wheeled robot is provided.

According to a fourth aspect of the present invention, the operation mode further includes an ambient condition detection sensor that detects an ambient condition of the robot and inputs a detection result to the motion control unit.
Based on the detection result of the ambient condition detection sensor, it is determined whether or not the follow-up is possible, and the follow-up operation is performed only when the follow-up is possible and stopped when the follow-up is impossible. An inverted two-wheeled traveling robot according to any one of the first to third aspects is provided.

According to the fifth aspect of the present invention, the apparatus further comprises an object detection sensor for detecting a load,
Based on the detection result of the object detection sensor, the operation control unit is configured to start the travel following the human after the load mode is loaded from the standby mode in the standby mode for waiting for the load. The first to fourth aspects are characterized by switching to a follow-up start waiting mode for waiting for an instruction and switching from the follow-up mode to an unloading wait mode for waiting for unloading in the follow-up mode in which the vehicle follows the human. An inverted two-wheeled traveling robot according to any one aspect is provided.

According to the sixth aspect of the present invention, the operation mode to be changed by the operation control unit includes a standby mode for waiting for the loading of the load, and a tracking for waiting for a travel start instruction following the person after the loading of the load. Including a start waiting mode, a follow-up mode for following the person, and an unloading waiting mode for stopping the running and waiting for unloading,
If the determination unit determines that the change in posture of the aircraft over time from the posture sensor has exceeded a threshold, the standby mode, the follow-up start waiting mode, the follow-up mode, the unloading wait mode, and the standby mode The inverted two-wheel traveling robot according to the first aspect is provided in which the operation mode is changed in order by the operation control unit.

According to the seventh aspect of the present invention, the operation mode to be changed by the operation control unit includes a standby mode for waiting for the loading of the load, and a tracking for waiting for a travel start instruction following the human after the loading of the load. Including a start waiting mode, a follow-up mode for following the person, and an unloading waiting mode for stopping the running and waiting for unloading,
When the object is detected by the object detection sensor, the operation control unit changes the operation mode from the standby mode to the follow-up start waiting mode, while the object detection sensor detects that the object has disappeared. In that case, the operation mode is changed by the operation control unit in the standby mode from the unloading standby mode,
If the determination unit determines that the change in posture of the aircraft over time from the posture sensor has exceeded a threshold value, the operation control unit moves from the follow-up start wait mode to the follow-up mode in the follow-up start wait mode. The inverted two-wheeled robot according to the fifth aspect, wherein the operation mode is changed by the operation control unit from the following mode to the unloading waiting mode in the following mode while the mode is changed. provide.

  Embodiments of the present invention will be described below with reference to the drawings.

(First embodiment)
FIG. 1A is a side view of an inverted two-wheeled traveling robot 67 as an example of an autonomous robot in the first embodiment of the present invention. In FIG. 1A, FIG. 1B, and FIG. 1C, 1a and 1b are coaxially arranged, and two wheels connected to two axles 2c and 2d, respectively, 2a and 2b are connected to two axles 2c and 2d, respectively. The two wheels 1a and 1b are driven to rotate independently to run the robot 67 and maintain the posture of the airframe 3 in a predetermined posture (standing posture as shown in FIG. 1B and FIG. 6A). As an example of the actuator, two motors 3 are supported on the axles 2c and 2d so as to be rotatable around the axles 2c and 2d, and a body for holding the motors 2a and 2b, and 4 is an attitude for detecting the attitude of the aircraft 3 It is a sensor.

  Further, the robot 67 has a cargo bed 5 as a part of the airframe 3 in order to carry luggage. Reference numeral 29 denotes a load on the loading platform 5. Reference numeral 15 denotes a human (user) who is giving an operation instruction, and gives an operation instruction by touching the robot 67. Further, the robot 67 detects the position of the user 15 in order to carry the baggage following the user 15 and outputs a detection result to the control unit 100 (the operation control unit 9 described later in detail). A human detection sensor 6 as an example is also provided on the upper part of the body 3, for example. In addition to the position of the user 15, the human detection sensor 6 may detect an object around the robot 67, for example, an obstacle such as an obstacle or other human beings when traveling. However, when an obstacle is detected, the operation control unit 9 can control the motors 2a and 2b so as to avoid the obstacle so that the robot 67 can travel to avoid the obstacle, or the obstacle can be avoided. If not, the motors 2a and 2b can be controlled so as to stop the running of the robot 67 before the robot 67 comes into contact with the obstacle. That is, it is possible to have a follow-up mode in which it is determined whether or not follow-up is possible based on the detection result of the human detection sensor 6 and the follow-up operation is performed only when follow-up is possible and stopped when follow-up is impossible.

  Here, as an example of the human detection sensor 6, one using ultrasonic waves is frequently used with high reliability. By providing a human with an ultrasonic and infrared oscillator and providing two ultrasonic microphones and an infrared light receiving element as the human detection sensor 6 (receiver), the human wave is detected from the time difference between the ultrasonic waves reaching the two ultrasonic microphones. It is possible to measure the position (transmitter position). An example of a product that applies such a principle is mimio (registered trademark) of Virtual Ink Corporation of the United States. In this product, a receiver is attached to the whiteboard, and a transmitter is attached to the whiteboard pen, so that the position of the pen can be detected in a non-contact manner, and the characters and pictures written on the whiteboard by the user can be imported to the PC in real time. It is.

  The human detection sensor 6 can also use radio waves. In that case, it is effective to detect the direction of arrival of radio waves, such as the Esper Antenna (registered trademark) developed at the International Telecommunications Research Institute (ATR). In this case, a human has a radio wave transmitter.

  Other examples of the human detection sensor 6 include those using image recognition of a camera image and those using a heat ray image of a far infrared camera.

  As a sensor for detecting an object around the robot 67, a laser scanner type area sensor of SICK, Germany is often used. This sensor obtains the distance to an object that reflects light by scanning the laser plane, and the relative positional relationship between the obstacle and the robot can be grasped in a plane.

  Whether or not an object such as a luggage 29 is placed on the loading platform 5 is determined by detecting the change of the attitude of the fuselage 3 by the attitude sensor 4 when the luggage 29 is placed on the loading platform 5. 29 can be detected on the loading platform 5. However, in order to more accurately detect the presence of an object on the loading platform 5, proximity sensors 70a and 70b as examples of object detection sensors may be provided. In that case, the detection results of the proximity sensors 70a and 70b are provided. Is input to the control unit 100. As the object detection sensor, a weight sensor that detects the weight of the object on the loading platform 5, an optical sensor that optically detects an object on the loading platform 5, and a mechanical object when the object is placed on the loading platform 5 are used. Various sensors can be used, such as contact sensors that turn on.

  As an example of the attitude sensor 4, a gyro / acceleration sensor is used. By detecting the direction of gravity by the attitude sensor 4, the attitude of the airframe 3 is detected with respect to the direction of gravity, and the traveling operation of the inverted two-wheeled traveling robot 67 and the like. The detection result is output to the control unit 100 that performs control. Based on the detected posture, the control unit 100 gives an appropriate torque command to the motors 2a and 2b to maintain the balance of the airframe 3, so that the airframe 3 has a predetermined posture (FIG. 1B or FIG. 6A). To maintain a standing posture).

  FIG. 2 is a control system block diagram of the inverted two-wheeled traveling robot 67 in the first embodiment of the present invention. In FIG. 2, 3 is the body of the inverted two-wheeled traveling robot 67, 2 a and 2 b are motors for rotationally driving the two wheels 1 a and 1 b of the inverted two-wheeled traveling robot 67, and 4 is for detecting the attitude of the body 3. This is a posture sensor. In FIG. 2, an attitude sensor 4, a situation detection sensor 6, luggage sensors 70a and 70b, and a load sensor 82 are provided for the body 3, and encoders 2e and 2f are also provided. The encoders 2e and 2f are sensors for detecting the rotation angle and rotation angular velocity of the wheel. The outputs of the encoders 2e and 2f are fed back to the controller 7 via the control loop 8. The information transmitted from the body 3 to the attitude sensor 4 and the encoders 2e and 2f is “mechanical motion”, and the result of the body 3 responding to the torque applied to the body 3 from the motors 2a and 2b. It is.

  The control unit 100 receives the output of the posture sensor 4 and outputs the operation mode 11 to the posture variation detection unit 13 that functions as an example of a determination unit, and detects the variation from the posture variation detection unit 13. An operation control unit 9 to which a signal 12 is input and a controller 7 that functions as an example of an attitude control unit that receives an operation command 10 from the operation control unit 9 and also receives an output of the attitude sensor 4 are configured. ing. Therefore, the output of the attitude sensor 4 is fed back and becomes torque input to the motors 2 a and 2 b via the controller 7. With the control loop 8 described above, the robot 67 can maintain an inverted posture. Reference numeral 9 denotes an operation control unit, which indicates how the robot 67 should move based on the operation scenario 14 prepared in advance according to the task and stored in the operation scenario database 14A (specifically, what operation The operation command 10 is given to the control loop 8. In the case of the robot 67 shown in FIG. 1A, the rotation speed of the left and right wheels 1a and 1b can be used as the command value of the operation command 10, and the translation corresponding to the difference or average of the rotation speeds of the left and right wheels 1a and 1b can be used. The speed or rotation speed can also be used as the command value. The output of the posture sensor 4 is input to the posture variation detection unit 13, where the posture variation is detected by a predetermined determination algorithm corresponding to the operation mode 11 given from the operation control unit 9 of the robot 67. The fluctuation detection signal 12 is returned to the operation control unit 9. The operation control unit 9 interprets this as an operation instruction of the user 15 to advance the operation mode 11 based on the operation scenario 14.

  Here, based on the attitude and wheel rotation angle of the airframe 3 detected by the attitude sensor 4 and the encoders 2e and 2f, the control unit 100 gives an appropriate torque command to the motors 2a and 2b to maintain the balance of the airframe 3. The control algorithm used in the control unit 100 when performing the above is Y. Ha and S.H. Yuta: Trajectory Tracking Control for Navigation of Self-Contained Mobile Inverse Pendulum, Proceedings of Conference on Intelligent Robots and Systems. 1875-1882, 1994. Hereinafter, a control algorithm used in the control unit 100 will be specifically described.

First, parameters of the aircraft 3 are defined as shown in FIG. The rotation angle of the wheels 1a, 1b is θ, the attitude angle of the airframe 3 is φ, the mass of the airframe 3 is M _b , the moment of inertia of the airframe 3 is I _b , and the mass of the wheels 1a, 1b is M _w (the wheels 1a, 1b are Because there are two, it is displayed as twice the mass of one wheel.), And the inertial moment of the wheels 1a, 1b is displayed as _Iw (since there are two wheels 1a, 1b, it is displayed as twice the inertial moment of one wheel. ), The radius of the wheels 1a and 1b is defined as r, and the height (distance) of the center of gravity of the airframe 3 is defined as L, as shown in FIG. Furthermore, as an example, the measured values of the prototype robot are applied to the parameters of the airframe 3 as follows. The motor 2a, the viscous friction coefficient mu _g of viscous friction coefficient mu _s and the ground and the wheel of 2b substitutes the experience respectively, the motor 2a, the torque constant of 2b tau _t and the reduction ratio η and the moment of inertia I _{M (motor} Since there are two 2a and 2b, it is displayed as twice the moment of inertia of one motor.) Adopts catalog data of the motors 2a and 2b to be used.

M _b = 59.4 Mass of the aircraft (kg)
M _w = 3.257 * 2 Wheel mass (kg)
I _b = 6.05 Aircraft moment of inertia (Kg · m ² )
I _w = 0.027 * 2 Wheel inertia (Kg · m ² )
I _M = 0.0527 * 2 Motor moment of inertia (Kg · m ² )
r = 0.2 Wheel radius (m)
L = 152/1000 Center of gravity distance (m)
μ _s = 0.0001 Motor viscous friction constant (N · m / (rad / sec))
mu _g = 0.0001 ground and the wheels of the viscous friction coefficient (N · m / (rad / sec))
τ _t = 3.69 Motor torque constant (N · m / A)
η = 30 Reduction ratio of motor g = 9.80665 Gravity acceleration (m / s ² )

  When the above variables are used, the equation of motion of the inverted mobile robot 67 (in other words, the above-described inverted two-wheeled traveling robot) 67 shown in FIG. However, it is assumed that the robot 67 is in an inverted state and the attitude angle φ of the body 3 is substantially zero.

（式２）

(Formula 2)

  Furthermore, when the state variables are defined as in Equation 3, Equations 1 and 2 can be summarized as the following state equations.

  The attitude angle φ of the airframe 3 can be measured by the attitude sensor 4, and the rotation angle θ of the wheels 1a and 1b is attached to the motors 2a and 2b, and an encoder 2e that can input information on the rotation angle θ to the control unit 100. , 2f, all the state variables can be measured. For this reason, it is possible to stabilize the robot 67 in an inverted state by determining an appropriate state feedback coefficient by an optimal regulator method or the like.

FIG. 4 shows a block diagram of a control system of the robot 67 as a more specific example. The inverted two-wheeled traveling robot 67 as shown in FIG. 1A is a one-input four-output system that receives a motor torque command 68 and outputs a state variable 69. The state variable 69 is multiplied by the state feedback gain 71 via the summing point 70 and then fed back to the inverted two-wheeled traveling robot 67 as the motor torque command 68 via the summing point 72. The purpose of the summing point 72 is to feed forward the friction of the system that cannot be described as a linear model, and the output of the friction estimator 74 expressed as a function of the angular velocity command value 73 is used. The purpose of the summing point 70 is to give a target value of the state variable 69, and the output of the target state generator 75 expressed as a function of the angular velocity command value 73 is used. The target state generator 75 not only generates the angle command value θ _ref from the angular velocity command value 73 but also generates a target value φ _ref of the posture angle φ that balances with the frictional force based on the value of the friction estimator 74.

  FIG. 5 is a diagram illustrating an example of the operation scenario 14 of the inverted two-wheeled traveling robot 67 in the first embodiment of the present invention. In FIG. 5, steps S16, S18, S20, and S22 are specific examples of the operation mode 11 in FIG. Steps S17, S19, S21, and S23 indicate criteria for determining when the operation mode 11 is advanced. Here, the task of the robot 67 is to carry the baggage 29 of the user 15 on behalf of the user. Based on this task, the change of the operation mode 11 will be described below.

  First, an outline of the operation scenario 14 will be described with reference to FIG. First, the central graph of FIG. 6 will be described with reference to FIG. That is, FIG. 7 is a diagram showing a change in posture of the body 3 detected by the posture sensor 4 of the inverted two-wheeled robot 67 in the first embodiment of the present invention and a change in the operation mode 11 based on the change. 6 and 7, reference numeral 24 is a graph showing the time change of the posture of the posture sensor 4 in the front-rear direction. Steps S16 (standby mode), S18 (start waiting mode), S20 (follow-up mode), S22 (unloading) (Waiting mode) represents a change in the operation mode 11 associated with the operation scenario 14. The operation M25 (loading operation), the operation M26 (back pressing operation), the operation M27 (chest pressing operation), and the operation M28 (unloading operation) represent specific operations of the user 15.

  In the standby mode (step S16), the robot 67 is balance-controlled so that the loading platform 5 is horizontal when the luggage 29 is not placed (see (a) of FIG. 6).

  In the luggage placing operation M25 (see FIG. 6B) for placing the luggage 29 of the user 15 on the loading platform 5, posture change occurs due to the loading of the luggage 29 on the loading platform 5 (the luggage placing operation M25). After waiting for this to stabilize, the operation mode 11 shifts to the start waiting mode (step S18) (see FIG. 6C). The reason for changing the operation mode 11 after waiting for the posture variation to stabilize is to wait until the luggage 29 is placed securely. Here, the luggage 29 mounted on the loading platform 5 acts as a disturbance element for the inversion control of the robot 67. After the disturbance is absorbed by the control system, a constant posture offset 30 is generated.

  In the start waiting mode (step S16) (see (c) in FIG. 6), the balance 29 is continuously controlled while the load 29 is placed on the loading platform 5, although the level may be slightly lost due to the weight of the load 29. .

  The robot 67 is triggered by the posture sensor 4 detecting the forward tilt posture by the operation M26 (see FIG. 6D) of pushing the back of the robot 67 of the user 15 and the posture return by the user 15 releasing the hand. Transits to the follow-up mode (step S20) (see FIG. 6E). The reason for waiting until the hand is released is to surely determine the intention of the user 15 and to minimize the risk of a collision or the like due to the transition to the follow-up mode S20 faster than the user 15 assumes. The robot 67 has a compliance for maintaining inversion by a control force, and a clear intention can be transmitted to the robot 67 when the user 15 firmly presses the robot 67 against this force. Further, as long as the robot 67 can be tilted forward, any part of the back may be pushed, or the entire robot surface may be pulled. Such an instruction operation that requires a certain amount of force and a wide tolerance for the place of contact are different from the operation instruction by a mechanical switch or the like.

  In the steady mode in the follow-up mode (step S20) (see FIG. 6E), the robot 67 travels following the user 15, that is, the robot 67 accelerates or decelerates according to the movement of the user 15, and the posture angle changes. 31a and 31b are generated. However, since there is rolling friction between the wheels 1a, 1b and the ground 90, the robot 67 travels with a forward tilt angle that cancels this. That is, during constant speed traveling, an offset 32 of the posture angle φ corresponding to the rolling friction of the wheels 1a and 1b is generated. Since these fluctuations in the posture angle φ can be predicted from the dynamic equation of the robot shown in Document 1, the posture fluctuation detection unit 13 does not misunderstand these fluctuations as an instruction from the user 15.

  When the user 15 arrives at the destination, and the user 15 pushes the chest of the robot 67 and the robot 67 is tilted backward by the motion M27 (see FIG. 6F), the user 15 moves at the moment when the posture is zero-crossed. It is determined that there is an instruction, and the mode changes to the unloading waiting mode (step S22) (see (g) of FIG. 6). The reason for determining that the user 15 is instructed to move at the moment when the posture is zero-crossed is that even if the stop command is determined early, the robot 67 operates on the safe side, and there is a risk of collision with the user 15. Because it goes down.

  In the unloading waiting mode (step S22) (see (g) of FIG. 6), the balance control is performed in the state where the load 29 is placed, as in the start waiting mode (step S18) (see (c) of FIG. 6). The user 15 waits for a load unloading operation (see FIG. 6H) from the loading platform 5. After the load unloading operation (see (h) in FIG. 6), the process returns to the standby mode (step S16) (see (i) in FIG. 6).

  Hereinafter, the operation scenario 14 will be described in detail.

  First, the robot 67 whose operation mode 11 is step S16 in the standby mode is detected by detecting a change in the posture of the airframe 3 by the posture sensor 4 by placing the load 29 on the loading platform 5 from the user 15 in step S17. A signal is input to the control unit 100 or the presence of a load is detected by proximity sensors 70a and 70b as an example of an object detection sensor (that is, a load detection signal is input to the control unit 100 from the proximity sensors 70a and 70b). Wait).

  When it is confirmed by the posture sensor 4 or the proximity sensors 70a and 70b that the baggage 29 is placed in step S17 (that is, a posture change signal from the posture sensor 4 or a baggage detection signal from the proximity sensors 70a and 70b is sent to the control unit 100. When input, the control unit 100 enters a step S18 waiting for a start, and then, in step S19, an operation start command is received from the user 15 (that is, the attitude sensor 4 detects a change in the attitude of the body 3). Wait).

  When the posture sensor 4 detects that an operation start command is received from the user 15 in step S19, the control unit 100 enters the tracking mode in step S20 from step S19, and follows the user 15 by the operation control of the control unit 100. To start. The motion control algorithm for following is disclosed in Document 2, Akihisa Oya, Yosuke Nagumo: Human following traveling of an autonomous mobile robot using a light emitter as a guide, and the Japan Society of Mechanical Engineers Robotics Mechatronics Lecture 2001.6.9-10 Is available. An example of the follow mode will be described later.

  Step S20, which is the follow-up mode, ends when the stop command of the user 15 is detected by the posture sensor 4 in Step S21 (that is, the change of the posture of the airframe 3 is detected by the posture sensor 4). Waiting for the user 15 to lower the luggage 29 (that is, the attitude sensor 4 detects a change in the attitude of the airframe 3 when the luggage 29 is lowered, or the proximity sensor 70a, 70b causes the luggage 29 to be placed on the loading platform 5. Step S22 is entered in the unloading waiting mode.

  Next, in step S23, it is detected that the posture has been changed by the posture sensor 4, or because the signals from the proximity sensors 70a and 70b that have been detected as having the load 29 on the loading platform 5 disappear, the load 29 is lowered. It is determined whether or not it has been completed. When the end of the unloading of the luggage 29 is detected, a series of tasks are completed, and the process returns to step S16 which is a standby mode.

  Next, the operation scenario 14 shown in FIG. 5 will be described with reference to FIGS. 8 and 9 from the viewpoint of the movement between the robot 67 and the user 15 so as to be easier to understand. 8 and 9 show a case where the robot 67 and the user 15 are viewed from above, where 15 is a user, 29 is a load, 1a and 1b are wheels of the robot 67, 3 is a fuselage, and 5 is a loading platform. A black arrow 83 represents the forward direction of the robot 67.

  In FIG. 8A, the user 15 with the luggage 29 arrives at the robot 67 in the standby mode (step S16).

  In FIG. 8B, when the user 15 puts the luggage 29 on the loading platform 5, the robot 67 detects the attitude change by the attitude sensor 4 or the luggage from the proximity sensors 70a and 70b in step S17. When the detection signal is input to the control unit 100, the control unit 100 recognizes the load placement operation of the user 15 and changes to the start waiting mode (step S18).

  Further, in FIG. 8C, when the user 15 presses the body 3, the robot 67 detects this change in the posture of the body 3 from the rear side to the front side by the posture sensor 4 in step S19. Recognized by the control unit 100 and enters the follow-up mode (step S20). Therefore, when the user 15 turns around in front of the robot 67 as shown in FIG. 8D, the robot 67 moves forward, that is, travels following the user 15 start. Relying on the human sensor 6 as shown in FIG. 8E, the user 15 follows as the user 15 continues walking. In the follow-up mode (step S20), the control unit 100 recognizes the position of the user 15 by the human sensor 6, and controls the motors 2a and 2b so that the distance between the robot 67 and the user 15 is always constant. However, if the user 15 is too close or cannot be seen in front of the robot 67, the robot 67 stops traveling.

  Next, in FIG. 9F, when the user 15 stops, the distance between the robot 67 and the user 15 detected by the human sensor 6 becomes equal to or less than a certain distance, so that the controller 100 drives the motors 2a and 2b. Is stopped, the robot 67 is stopped. Of course, if the user 15 starts walking again, the human sensor 6 causes the distance between the robot 67 and the user 15 to exceed a certain distance, so that the controller 100 starts driving the motors 2a and 2b. The robot 67 performs the following operation again.

  9 (g), when the user 15 looks back and approaches the robot 67, the distance between the robot 67 and the user 15 detected by the human sensor 6 becomes equal to or less than a certain distance. By stopping the driving of the motors 2a and 2b, the robot 67 is stopped on the spot. Further, when the user 15 presses the airframe 3 as shown in FIG. 9 (h), the robot 67 detects the change in the attitude of the airframe 3 from the front side to the rear side by the attitude sensor 4, thereby controlling this operation. The tracking mode (step S20) is terminated after being recognized by 100, and the unloading waiting mode (step S22) is entered.

  Next, in FIG. 9I, when the user 15 removes the luggage 29, the attitude sensor 4 detects the attitude change and inputs it to the control unit 100, or the proximity sensor 70a, 70b detects the luggage detection signal. Is no longer input to the control unit 100, the robot 67 recognizes this operation again by the control unit 100 and changes to the standby mode (step S16).

  Next, in (j) of FIG. 9, the user 15 leaves with the luggage 29 from the robot 67 that has entered the standby mode (step S <b> 16).

The inverted two-wheeled robot 67 according to the first embodiment is first characterized as a basic feature of an inverted two-wheeled robot.
・ Because it can be taller while maintaining balance compared to a four-wheel robot, the bottom area is small and a small turn is effective.
・ Because of the balance at all times, the car body does not have to tilt when climbing over the steps.
・ Because of the constant balance, it is safe to move away in the opposite direction when a collision occurs.
Etc.

For the above reasons, when an inverted two-wheeled robot is adopted as a robot that carries luggage in crowds such as airports,
・ Small turn in a crowd and not taking up space when stored.
・ The cargo bed does not tilt even when passing through steps and slopes, preventing luggage collapse.
・ In the event of a collision with a person or object, the impact can be kept to a minimum.
There is a merit such as.

  However, as described in the conventional example, there has been a problem with an interface when instructing the inverted two-wheeled traveling robot to carry the load. In other words, when constructing a user interface using mechanical switches, the user has to find and press the position of the switch on an unstable body that floats while balancing. This is not only cumbersome to find a switch, but is also unsuitable as a user interface, accompanied by anxiety that at first glance, the body of an unstable robot must be pushed.

  Therefore, in the first embodiment of the present invention, the posture of the inverted two-wheeled traveling robot 67 has compliance in the direction in which the balance is controlled (in other words, the posture is allowed to be displaced in accordance with the pressed force). Even when the user 15 presses an arbitrary place on the body 3 of the robot 67, an interface that can understand the intention of the user 15 can be realized.

  As described in the conventional example, a Segway (registered trademark) system is known as a conventional example of a user interface for an inverted two-wheel mechanism. Therefore, the difference between the interface system and the Segway (registered trademark) system according to the first embodiment of the present invention will be described with reference to FIG. When the user 215 who gets on the Segway (registered trademark) moves his / her weight and tilts the vehicle body 200 (see FIG. 10A), the wheel 201 rotates to compensate for the vehicle body 200 tilt, thereby moving forward ( (Reverse) is performed (see FIGS. 10B and 10C). Although this operation method is very intuitive, the user 215 needs to tilt his / her body while traveling. This is the same as the accelerator operation of a passenger car, and the user 215 must always give instructions. In contrast, in the first embodiment of the present invention, when the user 15 pushes the body 3 of the inverted two-wheeled robot 67 (see FIG. 11F), the robot 67 autonomously changes the mode. Then, the mode shifts from the stop mode (see (e) of FIG. 11) to the follow mode to the user 15 (see (g) of FIG. 11). Furthermore, when the user 15 presses the body 3 again (see (h) of FIG. 11), the mode again shifts to the stop mode (see (i) of FIG. 11). That is, in the first embodiment of the present invention, the operation of the robot 67 is given as a change in the posture of the robot 67 caused by the user 15 pushing the hand. For this reason, it is not necessary for the user 15 to keep giving instructions to the robot 67. Once the instructions are given, it is sufficient to watch the autonomous operation of the robot 67.

  Therefore, according to the inverted two-wheeled traveling robot 67 of the first embodiment, the user 15 applies a force to change the posture by using the compliance caused by trying to maintain the posture by the control. It is possible to give instructions. For example, it becomes possible for the inverted two-wheeled traveling robot 67 to realize an operation instruction that the user 15 can easily understand from daily habits, such as “start by pressing the back” or “stop by pressing the chest”. This is passive, such as in the conventional invention, by continuously stepping on the accelerator of the device (conventional example 1) or taking a reflective action with respect to the force applied by the user (conventional example 2). This is to realize an interface that simply instructs an autonomous operation, which is not a typical operation, in an intuitive and easy-to-understand manner.

  FIG. 12 shows a method for determining the mode transition timing. When the time-series change of the posture angle φ is measured, the amount of variation from a certain stable posture angle φ exceeds a predetermined threshold, and further, the time required for the change is within a predetermined time limit. It is determined that the change is not a disturbance but a change that has occurred as a result of the action of the user 15. The moment when the change exceeds the threshold value is designated as timing A, and the instant when the change is completed and stabilized again is designated as timing B. It is determined in advance at which timing the mode change occurs depending on the operation mode. This is because, depending on the operation mode, it is desirable to make the mode transition earlier, and in another operation mode, it is desirable to make the mode transition later. For example, in the start waiting mode, the mode transitions to the follow-up mode at timing B after waiting for the posture to be stable for safety. In the follow-up mode, a transition to the unloading waiting mode is made at timing A in order to reflect the stop command of the user 15 early.

  In addition, as shown in FIG. 12, the mode switching is detected when both the posture angle change amount exceeds a certain threshold value and the time required for the change is equal to or shorter than a certain time limit. That is, the determination criterion is that the magnitude of the posture angle change is greater than a certain value and the rate of change is greater than a certain magnitude. It is necessary to ensure that the magnitude of the posture angle change is greater than or equal to a certain value so that the fluctuation is not vibration or noise. The reason for limiting the change time is to distinguish it from the autonomous attitude control operation of the control system. In general, an inverted control system has a natural frequency determined by feedback gain, airframe mass, and moment of inertia. For this reason, the change in the attitude angle during acceleration / deceleration becomes a fluctuation near this frequency and becomes relatively slow. On the other hand, when the user 15 pushes the robot 67, there is an impact associated with the contact, and the rate of change of the posture angle φ increases, resulting in a short-time change. Therefore, the operation pressed by the user 15 causes a large change in posture within the time limit. As described above, it is possible to detect the contact motion of the user 15 that is not affected by noise, vibration, or acceleration / deceleration during traveling.

  FIG. 13 and FIG. 14 show simulation results obtained by combining the control system shown in FIG. 4 and the linear model of the robot 67 expressed by Expression 3. In the simulation, in order to model the state in which the robot 67 is pushed by the user 15, the state in which the luggage 29 is placed on the loading platform 5, and the state in which the robot 67 travels on the uneven road, a torque disturbance is set according to each situation. Then, the change of the attitude angle φ of the airframe 3 was calculated. Here, when modeling the action of the user 15 pushing the robot 67, the force of the user 15 pushing the robot 67 is approximated as a triangular wave, and when the pushing force of the user 15 is taken into the simulation, the disturbance from the viewpoint of the robot 67 This is called torque.

First, in FIGS. 13A and 13B, torque disturbance set values 79 of the motors 2a and 2b, which are modeled when the user 15 presses the body 3 of the stopped robot 67,
The graph 80 of the attitude | position angle change of the body 3 under the torque disturbance is shown. FIG. 13A shows a graph of torque disturbance change (a graph of the relationship between the disturbance torque and time) 79 set for 10 seconds. It is assumed that the user 15 pushes a place with a height of 1 m of the airframe 3 with a force of 5N, and the maximum value of the torque of the motors 2a and 2b is 5Nm. FIG. 13B shows a graph of posture change for 10 seconds (graph of relationship between posture angle φ and time) 80. The threshold value V76 of the posture angle φ set for detecting the movement of the user 15 pressing the body 3 by the posture sensor 4 is 0.05 radians, and the predetermined time T78 for determining the rising of the posture change is 0.00. 5 seconds. As one example model, the rising time T77 from the steady state to the set threshold value V76 is 0.3 to 0.4 seconds, shorter than the predetermined time T78 of 0.5 seconds, and the robot 67 This indicates that the control unit 100 can recognize this as an operation from the user 15.

13 (c) and 13 (d), a set value 79 of torque disturbance of the motors 2a and 2b when the user 15 places the load 29 on the loading platform 5 of the robot 67 being stopped,
The graph 80 of the attitude | position angle change of the body 3 under the torque disturbance is shown. In this case as well, when the luggage 29 is shifted from the center position and placed on the loading platform 5, a torque for overturning the robot 67 is applied, and this torque is applied to a staircase as indicated by 79 in FIG. Approximates a shape that changes with time. This torque is called disturbance torque because it appears to the robot 67 as disturbance. The torque setting value in FIG. 13C is based on the assumption that the user 15 places a 40 kg load 29 at a place where the center of gravity is on the loading platform 5 and 2.5 cm off the axle. As described above, it is assumed that a disturbance torque of a maximum of 10 Nm is generated by placing the load 29 on the loading platform 5. At this time, the posture angle change graph 80 in FIG. 13D reaches the set threshold value V76 in about 0.2 seconds, which is sufficiently more than the predetermined time T78 of 0.5 seconds. Short. From this, the robot 67 can recognize the loading operation of the user 15 by the control unit 100.

  14 (e) and 14 (f) show a graph of torque disturbance set values (graph of the relationship between disturbance torque and time) 79 and posture angle change when the robot 67 continuously climbs over a projection having a height of 1 cm. The graph (graph of the relationship between the attitude angle φ and time) 80 is shown. The torque setting value shown in FIG. 14 (e) is a pulse disturbance torque of about 5 Nm at maximum when the robot 67 gets over the projection of 1 cm in height, which looks like a cable rolling on the ground, every 2 seconds. Modeling. Since the posture angle change graph 80 at this time does not exceed the set threshold value V76 in the plus direction or the minus direction, the robot 67 detects the posture angle change 80 as an action of the user 15 by the control unit 100. Never do.

  14 (g) and 14 (h), a graph of a speed change when the robot 67 accelerates from a stopped state to 0.6 m / s and then decelerates (a graph of the relationship between speed and time) 81. And a graph of posture angle change (a graph of the relationship between posture angle φ and time) 80. For the acceleration / deceleration simulation of the robot 67, it is not necessary to set a disturbance torque, and it is only necessary to change the speed command value 73. FIG. 14 (g) shows a graph 81 of the speed change of the robot 67 calculated by the simulation. FIG. 14H shows that the posture angle change graph 80 at this time has a rise time T77 of about 0.8 seconds before the set threshold value V76 is exceeded. Since the rising time T77 is longer than 0.5 seconds, which is a predetermined time T78, the posture angle change 80 is not recognized by the control unit 100 as the operation of the user 15 by the robot 67.

  As described above, FIGS. 13A and 13B are recognized as user actions by the robot 67, and this is an opportunity to change the operation mode, but FIG. 13C and FIG. (D) is not recognized by the robot 67. Thanks to this, it is possible to prevent the robot 67 from erroneously recognizing the operation of the user 15 due to the state of the floor or its own acceleration / deceleration.

  13 (a) and 13 (b), the user 15 presses the airframe 3 of the stopped robot 67, and the user 15 places the load 29 on the loading platform 5 of the stopped robot 67. However, it may be difficult to clearly distinguish between the two.

  Therefore, when an object detection sensor such as the proximity sensor is not arranged, in the operation of the robot 67, the detected change in posture is placed on the load of the user 15 according to the current operation mode of the robot 67. It is determined whether the response corresponds to the user 15 pressing. That is, the control unit 100 determines that the detected change in posture corresponds to the load placement in the standby mode, determines that it corresponds to pressing by the user 15 in the start waiting mode, and determines that the user 15 in the follow-up mode. It is determined that the button is pressed, and in the unloading waiting mode, it is determined that it corresponds to the unloading.

  On the other hand, when an object detection sensor such as the proximity sensor is arranged, the control unit 100 determines whether the detected hair is detected. Specifically, in the modification of the first embodiment shown in FIG. 1B and FIG. 1C, a new sensor is provided for distinguishing between the load of the user 15 and the pushing operation of the user 15. 1B and 1C, 1a and 1b are coaxially arranged and two wheels connected to two axles 2c and 2d, and 2a and 2b are connected to two axles 2c and 2d, respectively. Two motors 3 as an example of an actuator for independently rotating and driving 1a and 1b are supported on the axles 2c and 2d so as to be rotatable about the axles 2c and 2d, and the motors 2a and 2b are driven. The body 4 to be held is an attitude sensor that detects the attitude of the body 3. Reference numeral 29 denotes a load on the loading platform 5. Reference numeral 15 denotes a user who wants to give an operation instruction, and the operation instruction is given by the user 15 coming into contact with the robot 67. Reference numerals 70a and 70b are proximity sensors for detecting the presence of an object on the loading platform 5. The proximity sensors 70a and 70b can detect whether the luggage 29 is present on the loading platform 5. By using the proximity sensors 70a and 70b, it is possible to determine the difference between the loading operation of the user 15 and the pressing operation of the user 15, and it is possible to reliably prevent the malfunction of the robot 67 due to a determination error of the control system. Further, since the presence / absence of the load 29 can be determined even by using the load sensor 82 that measures the weight of the load 29 mounted on the loading platform 5, the same effect as the case where the proximity sensors 70a and 70b are arranged. Can be obtained. The detection results of the proximity sensors 70a and 70b and the load sensor 82 are respectively input to the operation control unit 9 of the control unit 100 (see FIG. 2).

(Second Embodiment)
The present invention is not limited to the above-described embodiment, and can be implemented in various other aspects. For example, the present invention is not limited to an inverted two-wheeled traveling robot that performs loading and unloading operations, and can also be applied to an inverted two-wheeled traveling robot that does not perform loading and unloading operations. FIG. 15 shows an example of the interaction with the user 15 in the task of an inverted two-wheeled traveling robot that does not have loading and unloading operations. Here, an example will be described in which the guidance operation of the user 15 is performed by the inverted two-wheeled traveling robot 67 in the premises of a building where it is not necessary to assume the luggage 29. FIG. 15 shows a situation when the robot 67 and the user 15 are viewed from above. 15 is a user, 1 a and 1 b are wheels of the robot 67, 3 is a body, and 6 is a human sensor. A black arrow 83 represents the forward direction of the robot 67.

  In FIG. 15A, the user 15 arrives at the robot 67.

  When the user 15 pushes the airframe 3 from the back in FIG. 15B, the attitude sensor 4 detects a change in the attitude of the airframe 3 from the rear side to the front side, and the controller 100 of the robot 67 recognizes this and guides it. Change to mode.

  In the guidance mode, the robot 67 travels in front of the user so as to lead the user 15 to the destination while taking a certain distance between the robot 67 and the user 15 by using the human sensor 6 (see FIG. 15 (c)).

  When the user 15 stops, the robot 67 is stopped in order to prevent the user 15 from losing sight, and waits for an instruction from the next user 15. Therefore, when the user 15 stops in FIG. 15D, the distance between the robot 67 and the user 15 detected by the human sensor 6 exceeds a certain distance. By stopping the drive, the robot 67 stops.

  Here, when the user 15 presses the airframe 3 again as shown in (e-1) of FIG. 15, the attitude sensor 4 detects the attitude change of the airframe 3 from the rear side to the front side, and the control unit 100 detects the motors 2a, In order to drive 2b again, the robot 67 starts moving forward. However, as shown in (e-2) of FIG. 15, when the user 15 turns forward and pushes the airframe 3 from the front, the attitude sensor 4 detects a change in the attitude of the airframe 3 from the front side to the rear side, and the control unit Since the driving of the motors 2a and 2b is stopped by 100, the guidance mode is ended.

  In the second embodiment, the loading / unloading of the luggage 29 of the user 15 is not detected, but the user 15 recognizes that the user 15 has pushed from the front and back. This makes it possible to start and end guidance with an intuitive and easy-to-understand interface.

  It is to be noted that, by appropriately combining arbitrary embodiments of the various embodiments described above, the effects possessed by them can be produced.

  An inverted two-wheeled robot according to the present invention includes an unstable body, an attitude sensor that detects the attitude of the aircraft, and an actuator that generates a force for rotating the two axles to maintain the attitude of the aircraft. A posture control unit that controls the operation of the actuator so as to stabilize the posture of the aircraft, and a determination unit that determines the human intention based on a change in posture of the aircraft over time from the posture sensor. And an operation control unit that changes the operation mode from the output of the determination unit. In addition, by actively using the compliance generated by the balancing operation of an inverted two-wheeled robot, the robot is instructed by an intuitive human action such as pushing or pulling the approximate part of the robot, and the robot moves autonomously. Can be useful. This technology can also be applied to intuitive switching of operation modes of a vehicle or the like using a balance operation by control other than a robot.

1 is a side view of an inverted two-wheeled traveling robot in a first embodiment of the present invention. It is a side view of the inverted two-wheeled traveling robot in the modified example of the first embodiment. It is a front view of the inverted two-wheeled traveling robot in the modified example of the first embodiment. It is a block diagram of the control system of the robot in the first embodiment. It is explanatory drawing for demonstrating the definition of the parameter of the body of the inverted two-wheeled traveling robot of the said 1st Embodiment. It is a more specific block diagram of the control system of the inverted two-wheeled traveling robot of the first embodiment. It is a flowchart showing an example of the operation | movement scenario of the robot in the said 1st Embodiment. It is explanatory drawing which shows the outline of the operation | movement scenario of the robot in the said 1st Embodiment. It is a figure showing the mode of a change of the posture sensor of the robot in the said 1st Embodiment. (A)-(e) is the top view which looked at the robot and user in the said 1st Embodiment from the upper direction, respectively. (F)-(j) is the top view which looked at the robot and the user in the said 1st Embodiment from the upper direction, respectively. (A)-(d) is explanatory drawing which shows operation | movement of the robot of the Segway (trademark) system for a comparison with the robot in the said 1st Embodiment, respectively. (E)-(i) is explanatory drawing which shows operation | movement of the robot in the said 1st Embodiment, respectively. It is explanatory drawing which shows the determination method of the transition timing of the mode of the robot in the said 1st Embodiment. (A) and (b) are a graph of the relationship between the user's pushing force and time, which is regarded as a disturbance torque when the user pushes the body of the robot that is stopped in the robot in the first embodiment, and the attitude angle φ. It is a graph of the relationship with time, and (c) and (d) are the rotational torque due to the load that is regarded as a disturbance torque when the user places the load on the loading platform of the robot that is stopped in the robot in the first embodiment. It is the graph of the relationship between time, and the graph of the relationship between posture angle (phi) and time. (E) and (f) are graphs of the relationship between wheel torque and time required for getting over, which is regarded as disturbance torque when the robot continuously gets over the protrusion of 1 cm height in the robot in the first embodiment, and It is a graph of the relationship between the attitude angle φ and time, and (g) and (h) are the speed and time when the robot in the first embodiment is changed from a stopped state to a running state of 0.6 m / s. It is a graph of a relationship, and a graph of the relationship between posture angle (phi) and time. (A)-(e-2) is the top view which looked at the robot and user in 2nd Embodiment of this invention from the upper direction, respectively. It is a figure showing the vehicle balanced by the conventional control. It is a side view of the robot balanced by the conventional control.

Explanation of symbols

1a, 1b Wheel 2a, 2b Motor 2c, 2d Axle 2e, 2f Encoder 3 Airframe 4 Attitude sensor 5 Loading platform 6 Human detection sensor 7 Controller 8 Control loop 9 Operation control unit 10 Operation command 11 Operation mode 12 Variation detection signal 13 Attitude variation Detection unit 14 Operation scenario 14A Operation scenario database 15 Human (user)
24 Time variation of posture 29 Luggage 67 Inverted two-wheeled traveling robot 68 Motor torque command 69 State variable 70 Summation point 70a, 70b Proximity sensor 71 State feedback gain 72 Summation point 73 Angular velocity command value 74 Friction estimator 75 Target state generator 82 Load Sensor 90 Ground 100 Control unit

Claims (7)

Two wheels arranged on the same axis and connected to two axles respectively, and a vehicle body having a vehicle body rotatably supporting the two axles,
An attitude sensor for detecting the attitude of the aircraft,
An actuator for driving each of the two axles to generate a force for maintaining the posture of the airframe;
An attitude control unit for controlling the operation of the actuator so as to stabilize the attitude of the airframe;
A determination unit for determining the human intention based on a change in posture of the aircraft over time from the posture sensor;
An inverted two-wheeled traveling robot comprising: an operation control unit that changes an operation mode based on an output of the determination unit.   2. The operation mode to be changed by the operation control unit includes a standby mode for waiting to start running, a follow-up mode for running following the person, and a stop mode for stopping the running. An inverted two-wheeled robot described in 1.   The determination unit determines that the human body intends if the change in posture of the aircraft over time from the posture sensor exceeds a threshold, and determines that the human intention is not included if the change is equal to or less than the threshold. 3. The inverted two-wheeled traveling robot according to claim 1, wherein the operation mode is changed by the operation control unit only when the determination unit determines that the intention is the person. 4. The operation mode further includes an ambient condition detection sensor that detects a situation around the robot and inputs a detection result to the motion control unit,
Based on the detection result of the ambient condition detection sensor, it is determined whether or not the follow-up is possible, and the follow-up operation is performed only when the follow-up is possible and stopped when the follow-up is impossible. The inverted two-wheeled traveling robot according to any one of claims 1 to 3. It further comprises an object detection sensor for detecting the luggage,
Based on the detection result of the object detection sensor, the operation control unit is configured to start the travel following the human after the load mode is loaded from the standby mode in the standby mode for waiting for the load. 5. A follow-up start waiting mode for waiting for an instruction is switched, and in the follow-up mode in which the vehicle follows the person, the follow-up mode is switched to an unloading wait mode for waiting for unloading. The inverted two-wheeled traveling robot according to any one of the above. The operation mode to be changed by the operation control unit includes a standby mode for waiting for a load, a follow-up start waiting mode for waiting for a travel start instruction following the human after the load is placed, and following the human. Including a follow-up mode for traveling, an unloading waiting mode for stopping the above-mentioned traveling and waiting for unloading,
If the determination unit determines that the change in posture of the aircraft over time from the posture sensor has exceeded a threshold, the standby mode, the follow-up start waiting mode, the follow-up mode, the unloading wait mode, and the standby mode The inverted two-wheel travel robot according to claim 1, wherein the operation control unit sequentially changes the operation mode. The operation mode to be changed by the operation control unit includes a standby mode for waiting for a load, a follow-up start waiting mode for waiting for a travel start instruction following the human after the load is placed, and following the human. Including a follow-up mode for traveling, an unloading waiting mode for stopping the above-mentioned traveling and waiting for unloading,
When the object is detected by the object detection sensor, the operation control unit changes the operation mode from the standby mode to the follow-up start waiting mode, while the object detection sensor detects that the object has disappeared. In that case, the operation mode is changed by the operation control unit in the standby mode from the unloading standby mode,
If the determination unit determines that the change in posture of the aircraft over time from the posture sensor has exceeded a threshold value, the operation control unit moves from the follow-up start wait mode to the follow-up mode in the follow-up start wait mode. 6. The inverted two-wheeled traveling robot according to claim 5, wherein, in the follow-up mode, the operation mode is changed by the operation control unit from the follow-up mode to the unloading waiting mode in the follow-up mode.

Images