JP4143723B2 - Walking control method for multi-legged robot - Google Patents

Description

  The present invention relates to a method for controlling walking of a multi-legged robot having four or more legs on its body.

  Conventionally, as a kind of walking robot, a six-legged robot having six legs on its body is known (see Patent Document 1). A six-legged robot can walk with three legs raised at a time, and is excellent in stability.

However, the six-legged robot has a problem that the walking speed may be slow.
JP 2000-271350 A

  The present invention has been made to solve the above problems, and an object of the present invention is to provide a walking control method for a multi-legged robot capable of improving walking speed.

In order to solve the above-mentioned problem, a walking control method for a multi-legged robot according to claim 1 of the present invention includes:
A thigh joined to the body by a first turning joint, four or more legs having shins joined to the thigh by a second turning joint, and the first turning joint of each leg is driven to turn First driving means for rotating, second driving means for driving to turn the second turning joint of each leg, and a grounding end for detecting a position coordinate (x, z) of a grounding end which is a lower end of the shin part of each leg. A position detecting means; a ground detecting means for detecting that the ground end is grounded; and a support leg selected as three legs for supporting the body among the four or more legs. A walking control method for a multi-legged robot that controls the walking of any of the non-supporting legs excluding the legs, wherein the shin part is removed from a contracted leg state in which the non-supporting leg to be controlled is folded. The first movable limit that is the locus of the ground contact edge when turning and the non-supporting leg in a substantially linear shape At the time before walking in the grounded end movable range, which is a region sandwiched by the second movable limit that is the locus of the grounded end when the entire unsupported leg is turned from the extended leg state that is in the extended state. The first driving is performed so that the grounding end position detecting means detects the first rear leg position coordinate (x ₁ , z ₁ ) where the grounding end exists, and then raises the grounding end vertically upward. And the second drive means are operated, and the third rear leg position coordinate (x) which is the position of the ground contact end when the shin axis direction which is the longitudinal direction of the shin part is the vertical vertical direction in the contracted leg state ₃ , z ₃ ) The ground contact end has reached the second rear leg position having the position coordinate (x ₁ , z ₃ ) having the same vertical coordinate as the _third rear leg vertical coordinate z _3, which is the vertical coordinate of ₃ , z ₃ ). When the grounding end position detecting means detects the first driving means and the front The second driving means is stopped, and then the ground contact end is horizontally directed toward the third rear leg position, the first rear leg position coordinates (x ₁ , z ₁ ) and the third rear leg position coordinates (x ₃ , z ₃ ) actuating the first driving means and the second driving means so as to move horizontally by a movement amount corresponding to the stride value B ₂ which is the absolute value of the difference between the horizontal coordinates x ₁ and x ₃ , When the grounding end position detecting means detects that the grounding end has moved by the stride value B ₂ and reached the third rear leg position, the first driving means and the second driving means are stopped, and The first driving means and the second driving means are operated so as to lower the grounding end vertically downward, and when the grounding detection means detects that the grounding end is grounded, Control is performed to stop the second driving means.

In the walking control method for a multi-legged robot according to the present invention, firstly, three legs that support the body, consisting of the foremost leg that is foremost in the traveling direction and two legs that are located behind the foremost leg. The leg of the back leg, which is the non-supporting leg excluding the supporting leg, is controlled, and the leg shin is in the contracted leg state in which the controlled leg, which is the back leg to be controlled, is folded. Is a trajectory of the grounding end when the entire controlled leg is swung in the extended leg state in which the controlled leg is extended in a substantially straight line. The advantage is that the walking speed of the multi-legged robot can be increased because control is performed so that the stride to be stepped in the traveling direction is maximized within the movable range of the ground contact edge that is sandwiched between the second movable limits. have. Further, the walking control method for a multi-legged robot according to the present invention is secondly composed of the last leg that is the rearmost in the traveling direction and the two legs that are located in front of the last leg and that supports the body. The walking of one of the front legs, which are non-supporting legs, excluding the supporting leg, which is the leg of the leg, is controlled, and the height of the ground contact edge, which is the difference between the vertical coordinates of the trunk and the ground contact edge, is set for each of the four or more legs. The difference between the minimum ground end height H _min that is the minimum absolute value of the ground end height and the maximum ground end height H _max that is the maximum absolute value of the ground end height (H _max − H _min ) is calculated as the maximum ground unevenness value ΔH, and the first front leg position coordinates (the coordinates of the first position, which is the position of the controlled leg, which is the front leg to be controlled, before walking) x ₅ , z ₅ ) is detected by the ground contact edge position detecting means, the ground contact is vertically upward and the ground maximum unevenness value ΔH is set. Since the control is performed so as to increase to a value (ΔH + α) to which the margin value α is added, it is possible to respond to changes in the unevenness of the ground and to increase the walking speed of the multi-legged robot. have.

  In the embodiment described below, an example in which the number of legs is six among the multi-legged robot is described. The same mechanical mechanism as that of the conventional multi-legged robot is not changed, and the logic of the walking control is described. (Algorithm) makes it possible to increase the walking speed, and is the best mode for implementing the present invention.

  Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a perspective view showing the overall configuration of a multi-legged robot that is a first embodiment of the present invention.

  As shown in FIG. 1, the multi-legged robot 101 of the first embodiment is an underwater exploration robot including a moving body 1 and an arithmetic processing device 2. The moving body 1 has a watertight body 3 and six watertight legs 4A and 4B and 4C and 4D and 4E and 4F are provided on the side of the body 3 so as to be movable in water. ing. The body 3 is provided with a head 13a that can be rotated by a drive source such as an electric motor (not shown), and a bendable camera arm 13b is attached to the head 13a. A TV camera 5 is attached. The TV camera 5 has a watertight structure and can shoot a TV image underwater.

  Although not shown in FIG. 1, the body 3 is provided with a tilt sensor. This tilt sensor is a sensor that measures and outputs how much the mounted body 3 is tilted from the horizontal plane. Various known sensors can be used as the tilt sensor. As the inclination sensor, for example, a servo inclinometer or the like is used. Although not shown, the servo inclinometer suspends a pendulum from the frame and measures the angle formed by the frame and the pendulum. In this case, a coil is provided on the frame, a permanent magnet is attached to the tip of the pendulum, and a current is passed through the coil, so that the gap value between the pendulum and the coil is measured based on the value of the current flowing through the coil. Thus, the tilt angle of the pendulum can be obtained. In addition, as the tilt sensor, a potentiometer tilt sensor, a torque balance tilt sensor, a capacitance tilt sensor, or the like can be used.

  Although not shown in FIG. 1, the body 3 is provided with an azimuth angle sensor. This azimuth angle sensor is a sensor that measures the direction (azimuth angle) in which the body 3 is moving. Various known azimuth sensors can be used. As the azimuth sensor, for example, a geomagnetic sensor that detects the direction of the earth's magnetic north pole by detecting geomagnetism is used. As another azimuth angle sensor, a gyro sensor using a gyro compass can be used.

  In addition, although not shown in FIG. 1, the body 3 is provided with a water pressure sensor. This water pressure sensor is a sensor that measures the water pressure acting on the body 3. The depth of the fuselage 3 from the sea surface can be obtained from the measured water pressure value. Various known pressure sensors can be used as the water pressure sensor. As the water pressure sensor, for example, a Bourdon tube pressure sensor, a bellows pressure sensor, a diaphragm pressure sensor, or the like can be used.

  As will be described later, a ground sensor is attached to the lower surface of the lower ends (hereinafter referred to as “grounding ends”) of the six legs 4A to 4F. The grounding sensor is a sensor that detects that the grounding end is in contact with the ground.

  Of the multi-legged robot 101 described above, the arithmetic processing device 2 includes a computer main body 10, an input unit 7, an image display unit 8, and a data output unit 9.

  The computer main body 10 has a configuration as shown in the block diagram of FIG. As shown in FIG. 2, the computer main body 10 includes a CPU 41, a ROM 42, a RAM 43, and an input / output interface 11.

A CPU (Central Processing Unit) 41 has an internal bus (not shown) which is a signal line for sending and receiving current (signals) inside the CPU 41. A calculation unit (not shown), a register (not shown), a clock generation unit (not shown), an instruction processing unit (not shown), and the like are connected. The arithmetic unit in the CPU 41 performs four arithmetic operations (addition, subtraction, multiplication, and division) on various data stored in the register, or logical operation (logical product, logical sum, negation, exclusive logical sum). Etc.) or processing such as data comparison or data shift. The processing result is stored in a register or the like. The clock generation unit generates a clock signal (clock signal) for synchronizing the time of each part of the CPU 41. The CPU 41 operates based on this clock signal. The instruction processing unit controls and processes the fetching, decoding, and execution of instructions to be executed by the arithmetic unit or the like. Further, the CPU 41 detects the data sent or the calculation result and the time in real time based on the clock signal, and temporarily stores them in the RAM 43 or the like .

  A ROM (Read Only Memory) 42 stores a control program for controlling the CPU 41, various data used by the CPU 41, and the like. As the ROM 42, one constituted by a semiconductor chip, a hard disk device or the like is used. The control program for the CPU 41 includes processing procedures such as instructions for causing the CPU 41 to execute various processes and analysis operations, in addition to the basic software of the CPU 41 such as an OS (Operating System).

  A RAM (Random Access Memory) 43 temporarily stores data being calculated by the CPU 41 and the like. The RAM 43 is configured by a semiconductor chip or the like.

  A digital electrical signal generated by the CPU 41 or a digital electrical signal input to the CPU 41 is exchanged with the outside via the input / output interface 11. In the input / output interface 11, an A / D converter and a D / A converter which are conversion systems for digital signals and analog signals are provided. The operation of the computer main body 10 and the data input to the computer main body 10 are performed by the input unit 7 electrically connected to the computer main body 10, for example, the keyboard 7a and the mouse 7b. The signal is sent to the CPU 41 via the input / output interface 11.

  The result calculated by the CPU 41 is sent to the image display unit 8 including a cathode ray tube or a liquid crystal display through the input / output interface 11 and displayed on the screen as an image, characters, numbers, or the like. The data is also sent to the data output unit 9 for output. Examples of the data output unit 9 include a printer (not shown) that prints on paper and outputs, and a disk device that records on a recording disk 12 such as a magnetic disk or a magneto-optical disk.

  The mobile body 1 and the computer main body 10 are electrically connected by a remote control cable, and outputs from the TV camera 5 and various sensors of the mobile body 1 are sent to the CPU 41 of the computer main body 10 in real time, It is displayed on the display unit 8 or is calculated or processed as data for controlling the moving body 1. The moving body 1 is controlled by a command signal from the CPU 41 of the computer main body 10.

  FIG. 3 is a perspective view showing the structure of the legs in the multi-legged robot of the first embodiment of the present invention. FIG. 3 shows the configuration of the leg 4A among the six legs, but the other five legs 4B to 4F have the same configuration. As shown in FIG. 3, the leg 4 </ b> A includes a thigh part 15, a shin part 16, a horizontal rotary joint 14, a first turning joint 17, a second turning joint 18, and a foot part 20. . The leg 4A has three joints and means for driving each joint, and is a movable link mechanism that can be bent freely under the control of the CPU 41.

  Of the legs 4 </ b> A, the thigh 15 is joined to the body 3 in a movable state via the first turning joint 17 and the horizontal rotary joint 14. Moreover, the shin part 16 is joined to the thigh part 15 in a movable state via the second turning joint.

  The horizontal rotary joint 14 has a DC (direct current) motor 21 </ b> C, and the DC motor 21 </ b> C is fixed to the body 3. The DC motor 21C is an electric motor driven by direct current, has a good control response, and enables precise rotation control. The DC motor 21 </ b> C is driven and controlled by the CPU 41 of the computer main body 10.

  An input shaft (not shown) of the harmonic gear 22C is attached to a rotation drive shaft (not shown) of the DC motor 21C. Although not shown in the figure, the harmonic gear 22C includes a wave generator in the center, a flex spline disposed outside the center, and an internal gear called a circular spline disposed further outside. The The wave generator has an elliptical shape, and the flex spline is forcibly deformed by the elliptical shape of the wave generator. In addition, the flex spline is configured to engage with the circular spline at two locations whose central angles are separated by 180 degrees. By this harmonic gear 22C, the rotation of the DC motor 21C is decelerated at a certain reduction ratio.

  A joint cover 14a is attached to the shaft (output shaft) decelerated by the harmonic gear 22C. The joint cover 14a is formed in a substantially box shape, and a plate-like attachment portion is provided at the tip. The shaft of the follower bevel gear 26A is fixed to the plate-like attachment portion.

  An encoder 23C for measuring the rotation angle of the rotation drive shaft of the DC motor 21C is attached to the rear portion of the DC motor 21C. For example, an optical rotary encoder is used as the encoder 23C.

  Therefore, when the DC motor 21C rotates in response to a command from the CPU 41, the joint cover 14a rotates with the joint rotation center line 29C as the center of rotation. At this time, the encoder 23C measures the rotation angle of the DC motor 21C and outputs it to the CPU 41.

  The thigh 15 includes a DC motor 21A, a harmonic gear 22A, and an encoder 23A in a hollow cylindrical case 15a. In the thigh 15, the end on the harmonic gear 22 </ b> A side is connected to the first turning joint 17 in a movable state.

  A drive shaft 24A protrudes from the end of the thigh 15 on the first turning joint 17 side, and a drive bevel gear 25A is attached to the drive shaft 24A. A plate-shaped joint cover 15b is provided at the end of the case 15a of the thigh 15 on the first turning joint 17 side. The joint cover 15b at the end of the case 15a of the thigh 15 on the first turning joint 17 side is not fixed to the driven bevel gear 26A. Further, the plate-shaped joint cover 15b is provided with a through hole (not shown), and the shaft of the driven gear 26A is inserted therethrough. For this reason, the joint cover 15b can rotate about the center line (joint rotation center line) 29A of the follower bevel gear 26A.

  The driven bevel gear 26A is engaged with a drive bevel gear 25A attached to a drive shaft 24A perpendicular to the center line 29A of the driven bevel gear 26A. An output shaft that is decelerated by the harmonic gear 22A is coupled to the drive shaft 24A. The harmonic gear 22A has the same configuration and action as the harmonic gear 22C.

  A rotary drive shaft (not shown) of the DC motor 21A is attached to the input shaft of the harmonic gear 22A. The DC motor 21A is fixed to the case 15a of the thigh 15 and has the same configuration and action as the DC motor 21C.

  An encoder 23A that measures the rotation angle of the rotation drive shaft of the DC motor 21A is attached to the rear portion of the DC motor 21A. The encoder 23A has the same configuration and operation as the encoder 23C.

  Accordingly, when the DC motor 21A is rotated by a command from the CPU 41, the drive shaft 24A is rotated by being decelerated by the harmonic gear 22A, and the drive bevel gear 25A is rotated by this rotation. As a result, the entire thigh 15 rotates around the joint rotation center line 29A. At this time, the encoder 23A measures the rotation angle of the DC motor 21A and outputs it to the CPU 41.

  In the above description, the joint cover 14a provided on the horizontal rotary joint 14, the driven gear 26A having a shaft fixed to the joint cover 14a, the DC motor 21A and the harmonic gear built in and fixed to the case 15a of the thigh 15 A bevel gear 25A driven by 22A and a drive shaft 24A, and a joint cover 15b attached to one end of the case 15a of the thigh 15 and capable of rotating around the axis of the driven gear 26A are defined in the claims. The 1st turning joint 17 is comprised. The DC motor 21A, the harmonic gear 22A, the drive shaft 24A, and the drive bevel gear 25A constitute first drive means in the claims.

  A plate-shaped joint cover 15c is provided at the end of the case 15a of the thigh 15 on the second turning joint 18 side (the side opposite to the first turning joint 17). The shaft of the driven bevel gear 26B is fixed to the plate-shaped joint cover 15c.

  The shin part 16 includes a DC motor 21B, a harmonic gear 22B, and an encoder 23B in a hollow cylindrical case 16a. Moreover, in the shin part 16, the end part on the harmonic gear 22B side is connected to the second revolving joint 18 in a movable state.

  A drive shaft 24B protrudes from an end of the shin portion 16 on the second revolving joint 18 side, and a drive bevel gear 25B is attached to the drive shaft 24B. A plate-shaped joint cover 16b is provided at the end of the case 16a of the shin part 16 on the second turning joint 18 side. The joint cover 16b at the end of the case 16a on the side of the second revolving joint 18 of the shin part 16 is not fixed to the driven gear 26B. Further, the plate-shaped joint cover 16b is provided with a through hole (not shown), and the shaft of the driven gear 26B is inserted therethrough. For this reason, the joint cover 16b can rotate about the center line (joint rotation center line) 29B of the driven bevel gear 26B.

  The driven bevel gear 26B is engaged with a drive bevel gear 25B attached to a drive shaft 24B which is perpendicular to the center line 29B of the driven bevel gear 26B. An output shaft that is decelerated by the harmonic gear 22B is coupled to the drive shaft 24B. The harmonic gear 22B has the same configuration and action as the harmonic gear 22C.

  A rotary drive shaft (not shown) of the DC motor 21B is attached to the input shaft of the harmonic gear 22B. The DC motor 21B is fixed to the case 16a of the shin part 16, and has the same configuration and action as the DC motor 21C.

  An encoder 23B that measures the rotation angle of the rotation drive shaft of the DC motor 21B is attached to the rear portion of the DC motor 21B. The encoder 23B has the same configuration and action as the encoder 23C.

  Accordingly, when the DC motor 21B is rotated by a command from the CPU 41, the drive shaft 24B is rotated by being decelerated by the harmonic gear 22B, and the drive bevel gear 25B is rotated by this rotation. Thereby, the whole shin part 16 turns around the joint rotation center line 29B. At this time, the encoder 23B measures the rotation angle of the DC motor 21B and outputs it to the CPU 41.

  In the above, the joint cover 15c provided on the case 15a of the thigh 15; the driven gear 26B whose shaft is fixed to the joint cover 15c; and the DC motor 21B that is built in and fixed to the case 16a of the shin 16 A drive bevel gear 25B driven by the harmonic gear 22B and the drive shaft 24B, and a joint cover 16b attached to one end of the case 16a of the shin portion 16 and rotatable around the axis of the driven bevel gear 26B are as follows. The 2nd turning joint 18 in is comprised. The DC motor 21B, the harmonic gear 22B, the drive shaft 24B, and the drive bevel gear 25B constitute second drive means in the claims.

  Further, a rod-shaped rod portion 16d is provided at an end portion of the case portion 16a of the shin portion 16 opposite to the second turning joint 18. The other end of the rod-shaped rod portion 16d is a spherical portion 16e formed in a spherical shape. The spherical portion 16 e is fitted into a spherical concave portion 27 a formed on the upper portion of the disk-shaped grounding bump 27, and the spherical portion 16 e and the spherical concave portion 27 a constitute the spherical joint 19. The spherical joint 19 has the same function as a so-called universal joint, and the ground contact member 27 can be directed in any direction within the three-dimensional space with the center point of the spherical joint 19 as the center. Yes.

  The rod portion 16d, the spherical portion 16e, and the grounding member 27 described above constitute the foot portion 20.

  A ground sensor 28 is provided on the lower surface of the ground member 27. As the ground sensor 28, a known force measurement sensor such as a strain gauge or a load cell is used. The ground sensor 28 corresponds to the ground detection means in the claims. Further, the lower surface of the ground sensor 28 (the surface in contact with the ground) corresponds to the ground terminal in the claims.

  In addition, encoders 23A, 23B, and 23C provided on the legs 4A to 4F, an inclination sensor (not shown) provided on the body 3, an azimuth sensor (not shown), and a water pressure sensor (not shown). The position coordinates (x, z) of the ground end (the lower surface of the ground sensor 28), which is the lower end of the shin part 16 of each leg 4A to 4F, can be detected. An end position detecting means is configured.

  The above-described camera arm 13b is not shown in the figure, but, like the leg 4A described above, a joint and a movable link mechanism that has a means for driving the joint and can be bent freely under the control of the CPU 41. It has become.

  Next, the content of the walking control method in the first embodiment will be described with reference to FIGS. 4, 5, and 6. FIG.

  FIG. 4 is a top view showing the state of each leg when the moving body 1 of the multi-legged robot 101 of the first embodiment walks forward. In FIG. 4, the traveling direction of the multi-legged robot is a direction from the right to the left in the figure as indicated by an arrow in the figure. Among the legs 4A to 4F shown as rod-shaped members in FIG. 4, those shown in black are legs (hereinafter referred to as “supporting legs”) whose grounding end is in contact with the ground and supports a load such as its own weight applied to the body 3. ) And illustrated in white represents a leg that does not support a load such as its own weight applied to the body 3 (hereinafter referred to as “non-supporting leg”). The non-supporting leg includes a case where the grounding end is not in contact with the ground and exists in the air, and a case where the load such as its own weight applied to the body 3 is not supported even when the grounding end is in contact with the ground.

  In the case of the first embodiment, the leg 4D (hereinafter referred to as the “frontmost leg”) that is the foremost with respect to the traveling direction in FIG. 4 (the direction from right to left in FIG. 4; the same applies to FIGS. )) Or any of the legs 4B or 4F or 4A (hereinafter referred to as “rear legs”) excluding the two legs 4C and 4E (hereinafter referred to as “middle front legs”) adjacent to the rear of the frontmost leg. Handles the case of controlling such walking. As an example of this rear leg, in FIGS. 5 and 6, the leg 4A that is the rearmost in the traveling direction (hereinafter referred to as “last leg”) is a control target (hereinafter referred to as “controlled leg”). An example of a leg will be described. The two legs 4B and 4F adjacent to the front of the last leg are hereinafter referred to as “middle rear legs”.

  4, FIG. 4 (A) shows a state (black) in which the body 3 is supported by the three front legs 4C and 4E and the last leg 4A. Next, in FIG. 4B, in the state where the body 3 is supported by the three front legs 4C and 4E and the last leg 4A, the middle front legs 4C and 4E and the support legs (black) The state after moving the torso 3 in the horizontal direction from right to left in FIG. 4 by changing each angle of each thigh and shin of the last leg 4A of black) is shown. .

  Next, in FIG. 4C, the middle front legs 4C and 4E, which are the support legs (black), and the last leg 4A are supported in a state where the body 3 is supported by the three front legs 4C and 4E and the last leg 4A. In the unsupported state, the front and rear legs 4D of the non-supporting leg (white) and the thighs and shins of the middle rear legs 4B and 4F of the non-supporting leg (white) are changed without moving. The state after moving the position of the tip (grounding end) of the foremost leg 4D and the middle rear legs 4B and 4F forward (leftward in FIG. 4) is shown.

  In FIG. 4D, the position of the foot tip (grounding end) in FIG. 4C is not moved, and the frontmost leg 4D and the middle rear legs 4B and 4F, which were non-supporting legs (white), are large. By changing each angle of the thigh and shin, these legs also support the load of the torso 3, and the load of the torso 3 is supported by all six legs 4A to 4F. Show. Next, in FIG. 4 (E), the position of the toe (grounding end) in FIG. 4 (D) is not moved, and the thighs of the middle front legs 4C and 4E and the last leg 4D which were the support legs (black) It is shown that by changing each angle of the shin part, the toes (grounding ends) of these legs are brought into an unsupported state floating from the ground. In the process from the state shown in FIG. 4C to the state shown in FIG. 4E through the state shown in FIG. 4D, the supporting legs are changed from the middle front legs 4C and 4E and the last leg 4A to the frontmost leg 4D. The middle rear legs 4B and 4F are converted. This is hereinafter referred to as “leg change”.

  Next, in FIG. 4F, in the state where the body 3 is supported by the three legs of the foremost leg 4D and the middle rear legs 4B and 4F, the foremost leg 4D of the support leg (black), the middle rear legs 4B and The state after moving the torso 3 in the horizontal direction from right to left in FIG. 4 by changing each angle of each thigh and shin of 4F is shown.

  Next, in FIG. 4G, in the state where the body 3 is supported by the three legs of the foremost leg 4D and the middle rear legs 4B and 4F, the foremost leg 4D and the middle rear leg 4B which are support legs (black). 4F and 4F are not moved, and the middle front legs 4C and 4E and the last are not supported in the unsupported state by changing the angles of the thighs and shins of the middle front legs 4C and 4E and the last leg 4A of the unsupported leg (white). The state after moving the position of the foot tip (grounding end) of the leg 4A forward (leftward in FIG. 4) is shown.

  4H, the position of the foot tip (grounding end) in FIG. 4G is not moved, and the thighs of the middle front legs 4C and 4E and the last leg 4A which were non-supporting legs (white). By changing each angle of the shin part and the shin part, the load of the torso 3 is also supported by these legs, and the load of the torso 3 is supported by all of the six legs 4A to 4F. ing. Next, returning to FIG. 4 (A), the position of the foot tip (grounding end) in FIG. 4 (H) does not move, and the front legs 4D and the middle rear legs 4B and 4F, which were the support legs (black), are moved. By changing the angles of the thigh and shin, it is shown that the toes (grounding ends) of these legs are in an unsupported state floating from the ground. In the process from the state of FIG. 4 (G) to the state of FIG. 4 (A) through the state of FIG. 4 (H), the supporting leg is moved from the foremost leg 4D and the middle rear legs 4B and 4F to the middle front leg 4C. And 4E and the last leg 4A. This is also referred to as “leg change” hereinafter.

  That is, when the moving body 1 of the multi-legged robot 101 of the first embodiment shown in FIG. 4 is walking, the following control is performed. That is, the trunk 3 is always supported by at least three legs (support legs) among the six legs 4A to 4F, and first, the space of the trunk 3 is changed by changing the angles of the three support legs. The position is horizontally moved forward, and then, the toes of the remaining three legs (unsupported legs) not supporting the body 3 are stepped forward, and the toes of the unsupported legs are grounded forward. Later, walking is performed by repeating the control such that the leg that has been in a supported state is unsupported and the leg that has been in a non-supported state is in a supported state, and “stepping of the leg” is performed. Is to be done.

  Next, walking of the last leg in the multi-legged robot of the first embodiment will be described with reference to FIGS. FIG. 5 is a side view showing a basic procedure for walking the last leg 4A of the moving body 1 in the multi-legged robot 101 of the first embodiment.

First, when the pressure value measured and output by the ground sensor 28 is equal to or greater than the ground discrimination value, the CPU 41 of the computer main body 10 determines that the ground terminal (the bottom surface of the ground sensor 28 is grounded to the ground S1). Is determined. This position P1 is hereinafter referred to as a first initial position. The first initial position P1 corresponds to the first rear leg position in the claims. At this time, the CPU 41 determines that the coordinates of the first initial position P1 where the ground contact E1 is located are (x _a , z _a ) and records them in the RAM 42 or the like. The coordinates (x _a , z _a ) correspond to the first rear leg position coordinates in the claims. Here, the direction of the horizontal axis x is, for example, the horizontal direction from left to right in FIG. 5, and the direction of the vertical axis z is, for example, the vertical direction from bottom to top in FIG. In the state of being grounded at the first initial position P1, the body 3 is supported by at least three legs other than the last leg 4A as shown in FIG. The end (the ground sensor 28) can be easily separated from the ground S1.

  Next, the CPU 41 does not drive the DC motor 21C of the horizontal rotation joint 14 of the last leg 4A, and outputs a first joint operation command signal to the DC motor 21A of the first turning joint 17 of the last leg 4A in this state. . As a result, the DC motor 21A of the first turning joint 17 starts to rotate, the drive shaft 24A starts to rotate in a predetermined rotation direction at a predetermined rotation speed, and accordingly, the thigh 15 has a predetermined rotation direction. Start turning. At the same time, the CPU 41 outputs a second joint operation command signal to the DC motor 21B of the second turning joint 18 of the last leg 4A. As a result, the DC motor 21B of the second turning joint 18 starts to rotate, the drive shaft 24B starts to rotate in a predetermined rotation direction at a predetermined rotation speed, and accordingly, the shin portion 16 moves in a predetermined rotation direction. Start turning. By such control, as shown by the reference numeral 1 in FIG. 5, the grounding end E <b> 1 moves so as to draw a trajectory that rises vertically upward. As a result, the last leg 4A does not support the load of the moving body 1 and is in the state shown in FIG.

  Thereafter, the angle value of the drive shaft 24A measured by the encoder 23A of the last leg 4A is output to the CPU 41, and the angle value of the drive shaft 24B measured by the encoder 23B of the last leg 4A is also output to the CPU 41. The CPU 41 monitors these values, and when it is determined that the angle value of the drive shaft 24A has reached a predetermined value and the angle value of the drive shaft 24B has reached a predetermined value, the first turning joint of the last leg 4A. The first joint stop command signal is output to the 17 DC motor 21A, and the second joint stop command signal is output to the DC motor 21B of the second swing joint 18 of the last leg 4A. By such control, the ground contact E1 reaches a position P2 in FIG. 5 (hereinafter referred to as “first raised position”) and stops. The first raised position P2 corresponds to the second rear leg position in the claims. Actually, as shown in FIG. 5, the movement of the ground contact E1 in the vicinity of the first raised position P2 is curved, but hereinafter, it is assumed that it bends at a right angle in the vicinity of the first raised position P2. explain.

At this time, the CPU 41 determines that the coordinates of the first ascending position P2 where the ground contact E1 is located are (x _b , z _b ), and records them in the RAM 42 or the like. In this case, after the first raised position P2, since a raised position vertically upward from the first initial position P1, the horizontal coordinate x _b of the first raised after position P2 is the horizontal coordinate x _a of the first initial position P1 Is equal to In this case, the value of the difference between the first raised after position P2 vertical coordinates of the first initial position P1 (z _b -z _a) is a vertical first raised after the position P2 in the case where the first initial position P1 as a reference It represents the direction height.

  Next, the CPU 41 outputs a first joint operation command signal to the DC motor 21 </ b> A of the first turning joint 17. As a result, the DC motor 21A of the first turning joint 17 starts to rotate, and the drive shaft 24A starts to rotate at another rotational speed in another rotational direction different from the above-described upward movement. Accordingly, the thigh 15 starts turning in a turning direction different from the above. At the same time, the CPU 41 outputs a second joint operation command signal to the DC motor 21 </ b> B of the second turning joint 18. As a result, the DC motor 21B of the second turning joint 18 starts to rotate, and the drive shaft 24B starts to rotate at another rotational speed in another rotational direction different from the above-described upward movement. Accordingly, the shin part 16 starts to turn in a turning direction different from the above. By such control, as shown by the reference numeral 2 in FIG. 5, the grounding end E <b> 1 moves so as to draw a trajectory that laterally moves from right to left in FIG. 5.

  Thereafter, the angle value of the drive shaft 24A measured by the encoder 23A is output to the CPU 41, and the angle value of the drive shaft 24B measured by the encoder 23B is also output to the CPU 41. The CPU 41 monitors these values, and when it is determined that the angle value of the drive shaft 24A has reached another predetermined value and the angle value of the drive shaft 24B has reached another predetermined value, the first turning joint The first joint stop command signal is output to the 17 DC motor 21 </ b> A, and the second joint stop command signal is output to the DC motor 21 </ b> B of the second swing joint 18. By such control, the ground contact E1 reaches a position P3 in FIG. 5 (hereinafter referred to as “position after first lateral movement”) and stops. Actually, as shown in FIG. 5, the movement of the ground contact E1 in the vicinity of the position P3 after the first lateral movement has a curved shape, but hereinafter, it bends at a right angle near the position P3 after the first lateral movement. I will explain that.

At this time, the CPU 41 determines that the coordinates of the first post-lateral movement position P3 where the ground contact E1 is located are (x _c , z _c ) and records them in the RAM 42 or the like. In this case, since the first post-lateral movement position P3 is a position laterally moved from the first post-lift position P2 in the horizontal direction (the direction from right to left in FIG. 5), the vertical position of the first post-lateral movement position P3 is vertical. The coordinate z _c is equal to the vertical coordinate z _{b of} the first post-rise position P2. In this case, the absolute value of the horizontal coordinate difference (x _c −x _b ) between the first post-lateral movement position P3 and the first post-uplift position P2 is the value after the first lateral movement when the first initial position P1 is used as a reference. This represents the horizontal distance (step length B ₁ ) of the position P3.

  Next, the CPU 41 outputs a first joint operation command signal to the DC motor 21 </ b> A of the first turning joint 17. As a result, the DC motor 21A of the first turning joint 17 starts to rotate, and the drive shaft 24A further starts to rotate in another predetermined rotation direction at another predetermined rotation speed. 15 starts to turn in another predetermined turning direction. At the same time, the CPU 41 outputs a second joint operation command signal to the DC motor 21 </ b> B of the second turning joint 18. As a result, the DC motor 21B of the second turning joint 18 starts to rotate, and the drive shaft 24B starts to rotate in another predetermined rotational direction at another predetermined rotational speed. Starts turning in yet another predetermined turning direction. By such control, as shown by the reference numeral 3 in FIG. 5, the ground contact E <b> 1 moves so as to draw a trajectory descending vertically downward.

  Thereafter, the CPU 41 monitors the pressure value measured by the ground sensor 28, and if the pressure value is equal to or greater than the ground determination value, the CPU 41 determines that “the ground end (the bottom surface of the ground sensor 28) is grounded to the ground S1. It is determined. The contact determination value in this case may be a value smaller than the contact determination value at the first initial position P1, and may be determined as “grounded” if lightly touching the ground surface S1. At the time of this ground contact determination, the CPU 41 outputs a first joint stop command signal to the DC motor 21A of the first swing joint 17 and also outputs a second joint stop command signal to the DC motor 21B of the second swing joint 18. . By such control, the ground contact E1 reaches a position P4 in FIG. 5 (hereinafter referred to as “position after first lowering”) and stops.

At this time, the CPU 41 determines that the coordinates of the first lowered position P4 where the ground contact E1 is located are (x _d , z _d ) and records them in the RAM 42 or the like. In this case, since the first post-lowering position P4 is a position vertically lowered from the first post-lateral movement position P3, the horizontal coordinate _xd of the first post-lowering position P4 is the position of the first post-lateral movement position P3. Equal to the horizontal coordinate x _c . In this case, the absolute value of the vertical coordinate difference (z _d −z _c ) between the first post-lateral movement position P3 and the first post-lowering position P4 is the first lowering with respect to the first post-lateral movement position P3. It represents the vertical height of the rear position P4.

  Thereafter, the CPU 41 individually and simultaneously controls the other legs, for example, the foremost leg 4D and the middle rear legs 4B and 4F shown in FIG. 4 (E), and similarly to the above, the first turning joint 17 The DC motor 21A is operated while the rotation angle is monitored by the encoder 23A, and the DC motor 21B of the second turning joint 18 is operated while the rotation angle is monitored by the encoder 23B. Accordingly, the body 3 moves in the horizontal direction from right to left in FIG. 4 (horizontal direction D1 from right to left in FIG. 5), and is in the state shown in FIG.

Through the above-described control of the CPU 41 and driving of the DC motor 21A of the first turning joint 17 and the DC motor 21B of the second turning joint 18, etc., each step of walking (hereinafter referred to as “walking step”) is performed. The last leg 4A walks one step. Similarly for the other legs shown in FIG. 4, the foremost leg 4D, the middle front legs 4C and 4E, and the middle rear legs 4B and 4F, the control of the CPU 41 and the DC motor 21A of the first turning joint 17 and the second turning joint 18 By driving the DC motor 21B and the like, each leg walks one step through each step of walking (hereinafter referred to as “walking step”). In FIG. 5, the horizontal distance B ₁ between the first initial position P1 and the first lowered position P4 is the step length of the last leg 4A.

  Next, a walking control method for the last leg 4A, which is a characteristic control method for the multi-legged robot 101 according to the first embodiment of the present invention, will be described in detail with reference to FIG. The code | symbol 1 in FIG. 6 has shown the side view of the moving body 1 of the multilegged robot 101 of 1st Example.

  As shown in FIG. 6, when the shin part 16 is turned in a state where the last leg 4A, which is a controlled leg (a leg to be controlled), is folded (hereinafter referred to as a “reduced leg state”), The trajectory T1 drawn by the end is hereinafter referred to as “first movable limit”. In the contracted leg state, the shin part 16 of the controlled leg 4A is brought close to the limit (limit as a joint link mechanism having a certain diameter and length) with respect to the thigh 15. FIG. 6 illustrates a case where the last leg 4A is in the contracted leg state. Further, in FIG. 6, the leg 4D, which is a leg other than the last leg 4A, is also shown in a contracted state, but the other legs other than the last leg 4A may be in a contracted state by chance. However, when the last leg 4A is in the contracted leg state, it is not always necessary to be in the contracted leg state simultaneously.

  Further, the trajectory T2 of the ground contact when the controlled leg 4A is turned in the state where the controlled leg 4A is extended substantially linearly (hereinafter referred to as “extended leg state”) is hereinafter referred to as “second movable It is called “limit”. In the extended leg state, the shin part 16 of the controlled leg 4 </ b> A is straight with respect to the thigh part 15. Accordingly, the longitudinal center line of the shin part 16 is a straight line on the extension of the longitudinal center line of the thigh 15, and the longitudinal center line of the shin part 16 and the longitudinal center line of the thigh 15 The angle between and is 0 degrees or 180 degrees.

  Hereinafter, the region Z sandwiched between the first movable limit T1 and the second movable limit T2 is referred to as a “grounded end movable range”. In FIG. 6, the first movable limit T1 is an arc line drawn by the ground contact end when the second swivel joint 18 is rotated on the paper surface of FIG. (Part of a circle). Further, the second movable limit T2 is a line segment configured such that, in the extended leg state, the first turning joint 17 is the center of rotation and the thigh 15 and the shin 16 are in a straight line, FIG. This is a circular arc line (a part of a circle) drawn by the ground contact when rotated as a radius above. Accordingly, the grounding end movable range Z is a planar region surrounded by the arc line T1 and the arc line T2.

  However, in practice, if the horizontal rotation joint 14 is allowed to rotate at the same time, the first movable limit T1 is 3 with the second turning joint 18 as the center of rotation and the shin portion 16 as the radius in the contracted state. When rotated in the dimensional space, it becomes a part of the spherical surface drawn by the ground contact edge. Further, the second movable limit T2 is a three-dimensional structure in which the first turning joint 17 is the center of rotation and the thigh 15 and the shin 16 are in a straight line in a stretched state, and the radius is a line segment configured to be straight. When rotated in space, it becomes part of the spherical surface drawn by the ground contact edge. Accordingly, in this case, the grounded end movable range Z is a three-dimensional space region surrounded by the spherical portion T1 and the spherical portion T2.

  Next, a walking control method for the last leg 4A, which is the controlled leg of the multi-legged robot 101 according to the first embodiment of the present invention, will be described in detail with reference to FIG.

  The first embodiment is characterized in how the stride is set in the basic control of the last leg 4A described in FIG. In the first embodiment, the control is performed so that the stride to be stepped in the traveling direction is maximized within the contact end movable range Z.

Hereinafter, the position indicated by reference numeral 0 in FIG. 6 corresponds to the first initial position P1, and the position where the ground contact E1 (the lower surface of the ground sensor 28) exists at the time before the last leg 4A walks (the claims) An example of the first rear leg position in the range will be described. In this case, the coordinates of the position O are (x ₁ , z ₁ ), and these coordinates correspond to the first rear leg position coordinates in the claims.

In the first embodiment, the ground contact end is raised to the position G in FIG. 6 as the first raised position P2. At the position G, the vertical coordinate is equal to the vertical coordinate at the position A. The position A is the position of the ground contact when the longitudinal direction of the shin part 16 (hereinafter referred to as “shin axis direction”) is in the vertical vertical direction in the contracted leg state, and the coordinates thereof are (x ₃ , z ₃ ). This position A corresponds to the third rear leg position in the claims, and its coordinates (x ₃ , z ₃ ) correspond to the third rear leg position coordinates. Z ₃ corresponds to the third rear leg vertical coordinate. The first ascending position P2 corresponds to the second rear leg position in the claims, and its coordinates are (x ₁ , z ₃ ), which corresponds to the second rear leg position coordinates in the claims. is doing.

Thereafter, the ground contact is moved laterally from the position G in the horizontal direction (the direction from the right to the left in the drawing) to reach the position P3 after the first lateral movement. The post-first lateral movement position P3 is the position A in FIG. Value of step length at this time is B _2. Thereafter, the ground contact end is lowered vertically downward from the third rear leg position P3 (position A). Then, at any point during the descent, the contact is made with the ground, and the ground contact E2 is stopped at that position.

  In the first embodiment, when the first initial position P1 is above the line segment FG in the grounded end movable range Z in FIG. 6, the position G is raised immediately above as the first raised position, and then After moving horizontally to position A, it is lowered directly below. In addition, when the first initial position P1 is on the line segment IJ in the grounded end movable range Z in FIG. 6, the position J is raised immediately above the first ascended position, and then horizontally to the position A. After moving horizontally, move it down. When the first initial position P1 is on the line segment CD in the ground contact end movable range Z in FIG. 6, the position D is raised immediately above the first ascended position, and then horizontally to the position A. After moving horizontally, move it down.

In the conventional walking control of this type of multi-legged robot, there is no concept of the ground contact end movable range Z, and control is performed so that one walk is performed with a stride value set or selected in advance (or stored in the ROM 42) before walking. Was. However, in the first embodiment of the present invention, as described above, the step length value to be stepped in the traveling direction within the movable contact end movable range Z that the controlled leg 4A can take, for example, starts from the position O in FIG. the stride value as can be seen from the example, which may be a B _2, it is possible to obtain a maximum stride length value at that location when. Therefore, the walking speed of the multi-legged robot 101 can be increased.

  A control program (not shown) expressing the contents of the control of the multi-legged robot 101 of the first embodiment is stored in the ROM 42 in the computer main body 10, and the CPU 41 performs control based on this control program. Execute. In the first embodiment, the CPU 41 corresponds to the control means in the claims.

  Next, a second embodiment of the present invention will be described with reference to the drawings. The multi-legged robot 102 of the second embodiment is different from the multi-legged robot 101 of the first embodiment in that the content of the leg control (control program: not shown) performed by the CPU 41 is different from that of the first embodiment. That is. The control program (not shown) of the second embodiment is stored in the ROM 42 in the computer main body 10, and the CPU 41 executes control based on this control program. Also in the second embodiment, the CPU 41 corresponds to the control means in the claims. In the multi-legged robot 102 of the second embodiment, since the configuration and operation other than the control program are exactly the same as those of the multi-legged robot 101 of the first embodiment, the description thereof is omitted. FIG. 7 is a side view showing a basic procedure for walking the foremost leg 4D of the moving body 1 in the multi-legged robot of the second embodiment.

First, in FIG. 7, when the pressure value measured and output by the ground sensor 28 is equal to or greater than the ground discrimination value, the CPU 41 of the computer main body 10 determines that “the ground end (the bottom surface of the ground sensor 28) is the ground surface S2. Is in contact with the ground. " This position Q1 is hereinafter referred to as a second initial position. The second initial position Q1 corresponds to the first front leg position in the claims. At this time, the CPU 41 determines that the coordinates of the second initial position Q1 where the ground contact E2 is located are (x _e , z _e ) and records them in the RAM 42 or the like. The coordinates (x _e , z _e ) correspond to the first front leg position coordinates in the claims. Here, the direction of the horizontal axis x is, for example, the horizontal direction from left to right in FIG. 7, and the direction of the vertical axis z is, for example, the vertical direction from bottom to top in FIG. In the state of being grounded to the second initial position Q1, the body 3 is supported by at least three legs other than the foremost leg 4D as shown in FIG. The end (ground sensor 28) can be easily separated from the ground surface S2.

  Next, the CPU 41 does not drive the DC motor 21C of the horizontal rotary joint 14 of the foremost leg 4D, and outputs a first joint operation command signal to the DC motor 21A of the first turning joint 17 of the foremost leg 4D in this state. . As a result, the DC motor 21A of the first turning joint 17 starts to rotate, the drive shaft 24A starts to rotate in a predetermined rotation direction at a predetermined rotation speed, and accordingly, the thigh 15 has a predetermined rotation direction. Start turning. At the same time, the CPU 41 outputs a second joint operation command signal to the DC motor 21B of the second turning joint 18 of the foremost leg 4D. As a result, the DC motor 21B of the second turning joint 18 starts to rotate, the drive shaft 24B starts to rotate in a predetermined rotation direction at a predetermined rotation speed, and accordingly, the shin portion 16 moves in a predetermined rotation direction. Start turning. By such control, as shown by the reference numeral 1 in FIG. 7, the ground contact E2 moves so as to draw a trajectory that rises vertically upward. Thereby, the foremost leg 4D does not support the load of the moving body 1 and is in the state shown in FIG.

  Thereafter, the angle value of the drive shaft 24A measured by the encoder 23A of the foremost leg 4D is output to the CPU 41, and the angle value of the drive shaft 24B measured by the encoder 23B of the foremost leg 4D is also output to the CPU 41. The CPU 41 monitors these values, and when it is determined that the angle value of the drive shaft 24A has reached a predetermined value and the angle value of the drive shaft 24B has reached a predetermined value, the first turning joint of the foremost leg 4D. The first joint stop command signal is output to the 17 DC motor 21A, and the second joint stop command signal is output to the DC motor 21B of the second turning joint 18 of the foremost leg 4D. By such control, the ground contact E2 reaches the position Q2 in FIG. 7 (hereinafter referred to as “second raised position”) and stops. The second raised position Q2 corresponds to the second front leg position in the claims. Actually, as shown in FIG. 7, the movement of the ground contact E2 in the vicinity of the second raised position Q2 has a curved shape, but hereinafter, it is assumed that it bends at a right angle in the vicinity of the second raised position Q2. explain.

At this time, the CPU 41 determines that the coordinates of the second post-rise position Q2 where the ground contact E2 is located are (x _f , z _f ), and records them in the RAM 42 or the like. In this case, since the second post-lift position Q2 is a position that has been raised vertically upward from the second initial position Q1, the horizontal coordinate x _f of the second post-rise position Q2 is the horizontal coordinate x _e of the second initial position Q1. Is equal to In this case, the value of the difference between the second raised after position Q2 vertical coordinates of the second initial position Q1 (z _f -z _e), the vertical second raised after position Q2 in the case where the second rest position Q1 to a reference It represents the direction height.

  Next, the CPU 41 outputs a first joint operation command signal to the DC motor 21 </ b> A of the first turning joint 17. As a result, the DC motor 21A of the first turning joint 17 starts to rotate, and the drive shaft 24A starts to rotate at another rotational speed in another rotational direction different from the above-described upward movement. Accordingly, the thigh 15 starts turning in a turning direction different from the above. At the same time, the CPU 41 outputs a second joint operation command signal to the DC motor 21 </ b> B of the second turning joint 18. As a result, the DC motor 21B of the second turning joint 18 starts to rotate, and the drive shaft 24B starts to rotate at another rotational speed in another rotational direction different from the above-described upward movement. Accordingly, the shin part 16 starts to turn in a turning direction different from the above. By such control, as shown by a circled numeral 2 in FIG. 7, the ground contact E2 moves so as to draw a trajectory that laterally moves in the horizontal direction from right to left in FIG.

  Thereafter, the angle value of the drive shaft 24A measured by the encoder 23A is output to the CPU 41, and the angle value of the drive shaft 24B measured by the encoder 23B is also output to the CPU 41. The CPU 41 monitors these values, and when it is determined that the angle value of the drive shaft 24A has reached another predetermined value and the angle value of the drive shaft 24B has reached another predetermined value, the first turning joint The first joint stop command signal is output to the 17 DC motor 21 </ b> A, and the second joint stop command signal is output to the DC motor 21 </ b> B of the second swing joint 18. By such control, the ground contact E2 reaches a position Q3 in FIG. 7 (hereinafter referred to as “position after second lateral movement”) and stops. Actually, as shown in FIG. 7, the movement of the ground contact E2 in the vicinity of the second post-lateral movement position Q3 has a curved shape, but hereinafter, it bends at a right angle in the vicinity of the second post-lateral movement position Q3. I will explain that.

At this time, the CPU 41 determines that the coordinates of the second post-lateral movement position Q3 where the ground contact E2 is located are (x _g , z _g ) and records them in the RAM 42 or the like. In this case, the second post-lateral movement position Q3 is a position laterally moved from the second post-lift position Q2 in the horizontal direction (the direction from right to left in FIG. 7). The coordinate z _g is equal to the vertical coordinate z _{f of} the second ascending position Q2. In this case, the absolute value of the horizontal coordinate difference (x _g -x _f ) between the second post-lateral movement position Q3 and the second post-lift position Q2 is the second horizontal movement after the second initial position Q1 as a reference. This represents the horizontal distance (step length B ₃ ) of the position Q3.

  Next, the CPU 41 outputs a first joint operation command signal to the DC motor 21 </ b> A of the first turning joint 17. As a result, the DC motor 21A of the first turning joint 17 starts to rotate, and the drive shaft 24A further starts to rotate in another predetermined rotation direction at another predetermined rotation speed. 15 starts to turn in another predetermined turning direction. At the same time, the CPU 41 outputs a second joint operation command signal to the DC motor 21 </ b> B of the second turning joint 18. As a result, the DC motor 21B of the second turning joint 18 starts to rotate, and the drive shaft 24B starts to rotate in another predetermined rotational direction at another predetermined rotational speed. Starts turning in yet another predetermined turning direction. By such control, as shown by the reference numeral 3 in FIG. 7, the ground contact E2 moves so as to draw a trajectory descending vertically downward.

  Thereafter, the CPU 41 monitors the pressure value measured by the ground sensor 28. If the pressure value is equal to or greater than the ground determination value, the CPU 41 determines that “the ground end (the bottom surface of the ground sensor 28) is grounded to the ground S2. It is determined. The contact determination value in this case may be a value smaller than the contact determination value at the second initial position Q1, and may be determined to be “grounded” if lightly touching the ground surface S2. At the time of this ground contact determination, the CPU 41 outputs a first joint stop command signal to the DC motor 21A of the first swing joint 17 and also outputs a second joint stop command signal to the DC motor 21B of the second swing joint 18. . By such control, the ground contact E2 reaches a position Q4 in FIG. 7 (hereinafter referred to as a “second lowered position”) and stops.

At this time, the CPU 41 determines that the coordinates of the second post-lowering position Q4 where the ground contact E2 is located are (x _h , z _h ) and records them in the RAM 42 or the like. In this case, after the second lowered position Q4, since a position lowered vertically downward from the second lateral movement after position Q3, the horizontal coordinate x _h of the second falling after position Q4, the second lateral movement after position Q3 Equal to the horizontal coordinate x _g . In this case, the absolute value of the vertical coordinate difference (z _h −z _g ) between the second post-lateral movement position Q3 and the second post-lowering position Q4 is the second downward movement when the second lateral movement position Q3 is used as a reference. This represents the vertical height of the rear position Q4.

  Thereafter, the CPU 41 individually and simultaneously controls the other legs, for example, the middle front legs 4C and 4E and the last leg 4A shown in FIG. 4A, and the DC of the first turning joint 17 is controlled in the same manner as described above. The motor 21A is operated while the rotation angle is monitored by the encoder 23A, and the DC motor 21B of the second turning joint 18 is operated while the rotation angle is monitored by the encoder 23B. As a result, the body 3 moves in the horizontal direction from right to left in FIG. 4 (horizontal direction D1 from right to left in FIG. 7), and is in the state shown in FIG. 4B.

Through the above-described control of the CPU 41 and driving of the DC motor 21A of the first turning joint 17 and the DC motor 21B of the second turning joint 18, etc., each step of walking (hereinafter referred to as “walking step”) is performed. The foremost leg 4D walks one step. In FIG. 7, the horizontal distance B ₃ between the second initial position Q1 and the second post-lowering position Q4 is the step length of one step of the foremost leg 4D.

  Next, a walking control method for the foremost leg 4D, which is the controlled leg of the multi-legged robot 102 of the second embodiment of the present invention, will be described in detail with reference to FIG. FIG. 8 is a side view for explaining walking control of the foremost leg of the moving body in the multi-legged robot of the second embodiment of the present invention.

  The second embodiment is characterized in how the height for raising the ground contact E2 is set in the basic control of the foremost leg 4D described in FIG. In the second embodiment, the height that is raised when the ground contact E2 is stepped in the traveling direction is controlled so as to be the most appropriate value that is not too low and not too high according to the unevenness at that time. .

Hereinafter, the position indicated by the symbol M in FIG. 8 corresponds to the second initial position Q1, and the position where the ground contact E2 (the lower surface of the ground sensor 28) exists at the time before the foremost leg 4D walks (the claims) An example of the first front leg position in the range will be described. In this case, the coordinates of the position M are (x ₅ , z ₅ ), which correspond to the first front leg position coordinates in the claims, and z ₅ is the first front leg vertical coordinate. It corresponds to.

In the second embodiment, the grounding end is raised vertically upward to the position N in FIG. 8 as the second raised position Q2. The position of N corresponds to the position of the second front leg in the claims, and its position coordinate is (x ₆ , z ₆ ), and the vertical coordinate z ₆ is expressed by the following equation (1)

z ₆ = z ₅ + ΔH + α (1)

Is set to the value of

In the above equation (1), ΔH is obtained as follows. First, at the time before walking (when the ground contact E2 is at the position M), the height of the ground contact, which is the difference between the trunk 3 and the vertical coordinate of the ground contact E2 of the leg, is set to the six legs 4A to 4F. Calculate for each. Next, of these six grounding end heights, the one having the smallest absolute value is selected, and this is set as the minimum grounding end height _Hmin . At the same time, of these six grounding end heights, the one having the maximum absolute value is selected, and this is set as the maximum grounding end height H _max . Next, a difference (H _max −H _min ) between the maximum grounding end height H _max and the minimum grounding end height H _min is obtained, and this value is set to ΔH. This ΔH is hereinafter referred to as “the ground maximum unevenness value”.

  In the above equation (1), α is a margin value. As the value of α, a value stored in the ROM 42 in advance may be used, or the CPU 41 may calculate it by an appropriate method. For example, ΔH may be calculated by multiplying ΔH by a predetermined ratio. Examples of this ratio include values such as 5 to 30% with respect to ΔH. Alternatively, the member height value of the ground member 27 may be used. Alternatively, the past data of the unevenness value of the ground that has been walked up to that point may be recorded, and the margin value α may be set so that even the largest unevenness in the past can be overcome based on the data history. Good.

Since the position N is a position immediately above the position M, the horizontal coordinate x ₆ of the position N is equal to the horizontal coordinate x ₅ of the position M. Therefore, the position coordinates of the position N are (x ₅ , z ₅ + ΔH + α).

Thereafter, the ground contact E2 is laterally moved from the position N in the horizontal direction (the direction from right to left in FIG. 8), which is not shown in FIG. The subsequent steps are the same as the walking step illustrated in FIG. That is, it is moved horizontally from the position N to reach a position corresponding to the second post-lateral movement position Q3 (third front leg position in the claims). The stride value at this time is B _{3 as} shown in FIG. Therefore, the position coordinates of the position corresponding to the second post-lateral movement position Q3 (the third front leg position in the claims) is (x ₅ + B ₃ , z ₅ + ΔH + α). Thereafter, the ground contact E2 is lowered vertically downward from a position corresponding to the second post-lateral movement position Q3 (the third front leg position in the claims). Then, at any point during the descent, the contact is made with the ground, and the ground contact E2 is stopped at that position.

  In the second embodiment, by performing the control as described above, it is possible to set the rising value of the foot when the controlled leg 4D steps in the traveling direction to the most appropriate value according to the unevenness of the ground at that time. it can. Therefore, there is an advantage that it is possible to respond to changes in the unevenness of the ground and to increase the walking speed of the multi-legged robot 102.

  In addition, this invention is not limited to each above-mentioned Example. Each of the above-described embodiments is an exemplification, and has any configuration that has substantially the same configuration as the technical idea described in the claims of the present invention and that exhibits the same operational effects. Are also included in the technical scope of the present invention.

  For example, the multi-legged robot according to the present invention is not limited to the case where the number of legs is six, but when the number of legs is six or more, for example, the number of legs is eight, the number of legs is ten. In this case, the present invention can be applied to the case where the number of legs is an even number of 12 or more. Further, the number of legs may be 7 or more and an odd number. In addition, even if the number of legs is less than 6, if the control is performed so that at least three legs are always in the support state, the other legs are set in the unsupported state, and the first embodiment described above. Alternatively, it can be controlled as in the second embodiment. Therefore, the multi-legged robot according to the present invention can be applied to the case where the number of legs is four and the number of legs is five.

  In the first embodiment described above, an example in which the last leg 4A is controlled as a controlled leg when the last leg 4A is in an unsupported state has been described. However, if the control method of the present invention is used, other rear legs, For example, when the middle rear leg 4B is in an unsupported state, it can be controlled as a controlled leg. Further, if the control method of the present invention is used, when another rear leg, for example, the middle rear leg 4F is in an unsupported state, it can be controlled as a controlled leg. Further, in general, when controlling any of the rear legs, which are non-supporting legs excluding the supporting legs, which are the three legs that are composed of two legs located behind the foremost leg and support the body, the present invention is applied. Is applicable.

  Further, in the second embodiment described above, an example in which the foremost leg 4D is controlled as a controlled leg when the frontmost leg 4D is in an unsupported state has been described. However, if the control method of the present invention is used, other front legs, For example, when the middle front leg 4C is in an unsupported state, it can be controlled as a controlled leg. Further, when the control method of the present invention is used, when another front leg, for example, the middle front leg 4E is in an unsupported state, it can be controlled as a controlled leg. Further, in general, when controlling any of the front legs that are non-supporting legs excluding the supporting legs that are the three legs that support the torso consisting of the two legs positioned behind the last leg, the present invention. Is applicable.

  The multi-legged robot according to the present invention can be used not only as an underwater exploration robot as in the first and second embodiments, but also as various exploration robots on the ground, disaster rescue robots, and the like. .

  Further, as the first drive means or the second drive means in the present invention, a DD (direct drive) motor, a stepping motor, or the like may be used in addition to the DC motor described in the above embodiment.

  Further, the foot portion 20 having the spherical joint 19 as in each of the above embodiments may not be provided, and the lower end of the case 16a of the shin portion 16 may be directly grounded to the ground.

  In each of the above embodiments, the position coordinate is (x, z) and the horizontal coordinate is x, but the horizontal coordinate may be y.

  Further, the body 3 or the like may be equipped with a GPS receiver, and the position coordinates on the earth coordinates of the moving body 1 may be detected in real time.

  In addition, an ultrasonic sensor may be provided in the lower part of the body 3 and the like may be configured to walk while measuring the distance between the body 3 and the ground.

  The present invention can be implemented in companies that perform underwater exploration using a multi-legged robot, and can be used in these industries. In addition, multi-legged robots can be manufactured in the manufacturing industry of machines, electrical / electronic devices, etc., and can be used in these industries.

1 is a perspective view showing an overall configuration of a multi-legged robot that is a first embodiment of the present invention. FIG. It is a block diagram which shows the structure of the computer main body in the multilegged robot of 1st Example of this invention. It is a perspective view which shows the structure of the leg in the multilegged robot of 1st Example of this invention. It is a top view which shows the state of each leg at the time of the forward walking of the mobile body in the multilegged robot of 1st Example of this invention. It is a side view which shows the basic procedure of the walk of the last leg of the moving body in the multilegged robot of 1st Example of this invention. It is a side view explaining the walking control of the last leg of the moving body and the grounding end movable range in the multi-legged robot of the first embodiment of the present invention. It is a side view which shows the basic procedure of the walk of the foremost leg of the moving body in the multilegged robot of 2nd Example of this invention. It is a side view explaining walking control of the foremost leg of the moving body in the multi-legged robot of the second embodiment of the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Mobile body 2 Arithmetic processing apparatus 3 Body 4A-4F Leg 5 TV camera 6 Remote control cable 7 Input part 7a Keyboard 7b Mouse 8 Image display part 9 Data output part 10 Computer main body 11 Input / output interface 12 Recording disk 13a Head 13b Camera Arm 14 Horizontal Rotating Joint 14a Joint Cover 15 Thigh 15a Case 15b, 15c Joint Cover 16 Ticks 16a Case 16b Joint Cover 16d Rod Part 16e Spherical Part 17 First Turning Joint 18 Second Turning Joint 19 Spherical Joint 20 Foot Part 21A-21C DC motor 22A-22C Harmonic gear 23A-23C Encoder 24A, 24B Drive shaft 25A, 25B Drive bevel gear 26A, 26B Driven gear 27 Grounding member 27a Spherical recess 28 Ground sensor 29A-2 C joint rotation center line 41 CPU
42 ROM
43 RAM
101,102 multi-legged robot B ₁ .about.B ₃ stride D1, D2 traveling direction E1, E2 ground terminal P1 first initial position P2 first raised after position P3 first rear transverse movement after position P4 first lowered position Q1 second initial Position Q2 Position after second ascent Q3 Position after second lateral movement Q4 Position after second descending S1, S2 Ground T1 First movable limit T2 Second movable limit Z Grounding end movable range

Claims (1)

A thigh joined to the torso by a first pivot joint;
Four or more legs having a shin part joined to the thigh by a second turning joint;
First driving means for driving to turn the first turning joint of each leg;
Second driving means for driving to turn the second turning joint of each leg;
A grounding end position detecting means for detecting a position coordinate (x, z) of a grounding end which is a lower end of the shin portion of each leg;
A ground detection means for detecting that the ground end is grounded;
A support leg selected as three legs for supporting the body among the four or more legs;
A multi-legged robot walking control method for controlling the walking of any of the non-supporting legs excluding the supporting legs,
The first movable limit that is the locus of the ground contact when the shin part is turned from the contracted leg state in which the non-supporting leg to be controlled is folded, and the non-supporting leg is extended substantially linearly. When the entire unsupported leg is turned from the extended leg state, the ground contact end movable range is a region sandwiched by the second movable limit that is a locus of the ground contact end, and the ground contact is performed before walking. The grounding end position detecting means detects the first rear leg position coordinates (x ₁ , z ₁ ), which is the position where the end exists,
Next, the first driving means and the second driving means are operated so as to raise the grounding end vertically upward, and the shin axis direction which is the longitudinal direction of the shin portion in the contracted leg state becomes the vertical vertical direction. Position coordinates (x ₁ , z ₃ ) having a vertical coordinate equal to the third rear leg vertical coordinate z ₃ , which is the vertical coordinate of the third rear leg position coordinate (x ₃ , z ₃ ) which is the position of the ground contact edge at the time When the grounding end position detecting means detects that the grounding end has reached the position of the second rear leg, which is the position, the first driving means and the second driving means are stopped,
Next, the horizontal contact point x of the first rear leg position coordinates (x ₁ , z ₁ ) and the third rear leg position coordinates (x ₃ , z ₃ ) is set so that the ground contact end is horizontally directed toward the third rear leg position. The first driving means and the second driving means are operated so as to move horizontally by a movement amount corresponding to the stride value B ₂ which is the absolute value of the difference between ₁ and x ₃ , and the grounding end is the stride value B _2. The first driving means and the second driving means are stopped when the grounding end position detecting means detects that it has moved only to reach the third rear leg position,
Next, the first drive means and the second drive means are operated so as to lower the ground end vertically downward, and the first drive means is detected when the ground detection means detects that the ground end is grounded. And a control method for walking the multi-legged robot, wherein the second drive means is controlled to stop.

Images