KR20090002398A - Device of pulse command generation for position reference and sensor data acquistion of human riding robot's joint actuator and its control device and method thereof - Google Patents

Abstract

Device for sensor information acquisition and pulse command generation producing the location order of the human on board type robot knuckle driver, and the control device and method using the same are provided that the rotation angle and angular velocity of AC motor in a plurality of joint driving parts are controlled. A device for sensor information acquisition and pulse command generation producing the location order of the human on board type robot knuckle driver comprises a sensor information acquisition part receiving the detection signal of the sensor(400) through the CAN(Controller Area Network) communication, a PCI(Programmable Communication Interface) interface unit(120) delivering the detected signal to a master controller(200) of robot, a digital signal processor for controlling the pulse number generated for a certain time, a motor connector(160) for transmitting the motor control signal with the motor controller(300).

Description

Device of pulse command generation and position reference and sensor data acquistion of human riding robot's joint actuator and its control device and method approximately}

1 is a diagram showing a human riding biped robot (HUBO FX-1).

2 is a conceptual diagram of a whole control system of a human riding biped robot.

3 is a detailed system configuration diagram of a human riding biped robot.

4 is a view for explaining the structure of the RTX used in the main controller of the robot.

5 is a system configuration diagram of a motion controller of a human riding biped robot.

6 is an internal configuration diagram of a motion controller of a human riding biped robot.

7 is a circuit diagram for data exchange between a PCI interface chip, a DSP chip, and an AC motor chip in a motion controller.

8 is a data configuration diagram for data exchange between a PCI interface chip (PLX) and a DSP.

9 is a signal flow diagram between a PCI interface chip (PLX) and a DSP.

10 is a signal timeline between the PCI interface chip (PLX) and the DSP.

11 is a view showing a situation in which the height of the robot is changed by the initial position difference.

12 is a view showing the position of the AC motor shaft of the robot.

Fig. 13 is a time diagram when finding the Z-phase of an AC motor.

14 is a view showing the number of pulses generated in the X-axis and the Y-axis by the Digital Differential Analyzer (DDA) for a certain time.

15 is a diagram showing a process of generating a pulse signal for controlling an AC motor by a DDA algorithm for a predetermined time.

16 is a configuration diagram showing a connection relationship between a PCI interface chip (PLX) and a CAN controller.

<Explanation of symbols for the main parts of the drawings>

100: motion controller interface device

121: PCI (PC 104+) bus interface chip 123: dip switch 1-3

127: OSC 2 102a, 102b: EPLD1, EPLD2

111: CAN controller 112: photo coupler 1,2

113: CAN transceiver 114: CAN connector

130: buffer 5-8 140: DSP

141: OSC 1 150: Buffer 1-3

160: AC motor connector

The present invention relates to a motion controller interface device and method for a human-ride biped walking robot, and in particular, receives an angle command of each joint from the main controller of the robot and generates an angle command of each joint in a pulse form through a DSP, and generates a predetermined time. Pulse command generation and sensor information acquisition device for controlling position and angular velocity of the AC motor of the robot joint drive unit by adjusting the number of pulses generated during An apparatus and method are provided.

Since the first humanoid robots were developed, the appearance of Asimo in 2000 opened a new era of robots in the world. Of course, there were many robots before Asimo, but Asimo was a human-sized robot that had existed in the imagination in the past, showing quite natural movements. Many humanoid robots emerge in this environment. In addition to Asimo in Japan, there are HRP, KHR in Korea, and BHR in China.

Among the biped walking robots, a robot in which a person rides a biped on a robot is called a human riding biped robot. The joint drive of such a robot uses a high-capacity-lightweight driver.

This has opened up the possibility of being used in various industrial sites as well as the emergence of new vehicles after the development of automobiles.

Conventional human-ride biped robots include I-Foot of Toyota, Japan, and WL-15 and WL-16 (Waseda Leg-No. 15, No. 16) of Waseda University, and our HUBO FX-1. .

Toyota's I-Foot has a height of 2.36m, a weight of 200Kg, 12 degrees of freedom, and a walking speed of 1.35Km / h. It can carry and move a 60Kg person, and the robot's lower knee joint is like a bird's leg. It is characterized by bending in the opposite direction.

Waseda University's human-mounted robot has a leg of a robot with a payload of up to 50Kg in the form of a parallel mechanism using 12 linear drives based on a linear drive using a motor.

The HUBO FX-1 has a height of 1.98m, a weight of 185Kg, 12 degrees of freedom, a walking speed of 1.25Km / h, and a payload of 100Kg. Unlike Toyota's I-Foot, it walks like a human, and each joint is driven by an AC motor and a harmonic drive reducer.

Multifunctional robots with multi-jointed joints, such as biped walking humanoid robots (HUBO), generally use a PC for information processing because they require a lot of information to be processed in the operation and fast calculation in real time.

PCs, such as human bipedal walking robots, are commonly used for their integrated control of multi-joint, complex sensor systems due to the advantages of high computational speed, versatility of storage and input-output devices, and ease of remote control. Control hardware. To make full use of the PC's peripheral functions (easiness of remote control and monitoring using the Internet, general use of storage devices, availability of driver software for other hardware, and ease of development of control software), use Windows, a general purpose OS for PC. do.

The human riding humanoid robot should be able to fully drive even if there is a payload of about 100 kg for the human riding function. In addition, the human riding humanoid robot should have a space for people to ride. Therefore, the human riding humanoid robot uses an AC servo motor used in an industrial field unlike the conventional humanoid robot to have such a force.

Conventionally, no interface device for controlling the AC motor of the robot joint drive in the position control mode using a PC as the main controller has not been proposed.

In addition, when controlling the AC motor in the position control mode, a predetermined number of pulses have to be generated for a predetermined time, and an appropriate method of adjusting the pulse interval has not been proposed.

The present invention has been proposed to solve the problems of the prior art, and is interfaced with the main controller (PC) of the human-type biped walking robot, and receives the angle command of each joint from the main controller (PC) for each joint through the DSP Pulse for generating position command of the human riding robot joint driver for controlling the rotation angle and angular velocity of the AC motor of the joint drive of the robot by generating the angular command of the pulse type and adjusting the number of pulses generated for a predetermined time. An object of the present invention is to provide a command generation and sensor information acquisition device and a control device and method using the same.

In order to achieve the object of the present invention, the present invention is a pulse command generation and sensor information acquisition device for generating a position command of the human-mounted robot joint driver, sensor information for receiving signals detected from a plurality of sensors in a CAN communication method Acquisition unit; A PCI interface unit for transmitting signals detected from the plurality of sensors received by the sensor information acquisition unit to a main controller of the robot, and for transmitting an angle control command of each joint of the robot received from the main controller; Digital signal processing unit for generating a pulse signal as a motor control signal for controlling the motor driving each joint by receiving the angle control command of each joint of the robot from the PCI interface unit and adjusting the number of pulses generated for a predetermined time ; And a motor connector for transmitting a motor control signal output from the digital signal processor to each motor controller.

The motor for driving each joint of the robot is characterized in that the AC motor.

The digital signal processing unit receives a angular command of each joint from the main controller of the robot through the PCI interface unit and uses a DDA algorithm to distribute a time interval between pulses to be generated for a predetermined time.

The input port of the digital signal processing unit includes a port for receiving a command from the PCI interface unit and a port for receiving a signal from the motor controller, and the output port of the digital signal processing unit generates a pulse command signal regarding a rotation angle of the motor. And a direction command signal generating port indicating a rotation direction of the motor.

The angle control command of each joint of the robot consists of a 4-bit motor ID for designating the motor, a 1-bit sign bit for determining the rotation direction of the motor, and a 11-bit data bit for indicating the rotation angle of the motor. It is characterized by.

The plurality of sensors may include: a force / torque sensor measuring a repulsive force against the ground and torque applied to each ankle at each foot of the robot; An inertial sensor system for measuring the angle and the angular velocity of the robot with respect to the gravity direction; And an inclination sensor for measuring an inclination of the ground in contact with the foot of the robot.

The force / torque sensor is installed on each foot of the robot, measures the force in one direction and the moment in two directions, and auto-balancing and auto nulling point control algorithms. It characterized in that to use.

The inertial sensor system is characterized in that it comprises an accelerometer for measuring the degree of inclination of the robot relative to the direction of gravity and a rate gyro for measuring the angular velocity at which the robot is inclined.

A pulse command generation and sensor information acquisition device for generating position commands of a human-mounted robotic joint driver according to the present invention includes: a first buffer unit for lowering a drive voltage of the PCI interface unit to a drive voltage of the digital signal processor; And a second buffer unit for raising the driving voltage of the digital signal processor to a driving voltage of a motor controller that controls the motors driving the respective joints.

In order to achieve another object of the present invention, the present invention provides a joint drive control device for a human-mounted robot, comprising: a plurality of sensors for measuring the position of the robot and the state of the ground; A main controller of the robot for receiving sensor data measured from the plurality of sensors and providing an angle control command for each joint of the robot; And a motion controller for receiving an angle control command of each joint of the robot from the main controller of the robot and generating a pulse signal required to control the joint driving unit of each joint of the robot.

Each joint driving unit is characterized by comprising an AC motor and an AC motor controller.

The motion controller receives an angle command of each joint of the robot from the main controller of the robot and distributes the time interval between pulses that should be generated for a predetermined time from a pulse signal controlling an AC motor of a corresponding joint drive unit. It is characterized by using the DDA algorithm.

In order to achieve another object of the present invention, the present invention provides a method for controlling a joint driver of a human-mounted robot, comprising: (a) receiving sensor data measured from a plurality of sensors in a CAN communication method; (b) a main controller (PC) of the robot generating an angle control command of each joint of the robot based on the sensor data; (c) The motion controller receives the angle control command of each joint of the robot and adjusts the interval of pulses generated for a predetermined time by the DDA algorithm to provide the motor control signal necessary to control the AC motor driving each joint of the robot. Generating; And (d) transmitting the motor control signal to an AC motor controller of an AC motor that drives each joint of the robot.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. Prior to this, terms or words used in the specification and claims should not be construed as having a conventional or dictionary meaning, and the inventors should properly explain the concept of terms in order to best explain their own invention. Based on the principle that can be defined, it should be interpreted as meaning and concept corresponding to the technical idea of the present invention. Therefore, the embodiments described in the specification and the drawings shown in the drawings are only the most preferred embodiment of the present invention and do not represent all of the technical idea of the present invention, various modifications that can be replaced at the time of the present application It should be understood that there may be equivalents and variations.

1 is a view showing a human riding biped robot (HUBO FX-1).

The HUBO FX-1 developed by the applicant is a biped walking robot capable of human riding. Height is about 139cm (199cm with chair) and weighs about 130Kg (150kg with chair). It has 12 degrees of freedom. It can carry about 100Kg payload to allow people to ride. The joint drive is driven by an AC motor and a harmonic drive. The main specifications of the HUBO FX-1 are as follows.

    Height     1.393 m (1.988 with chair)     Weight     130Kg (150Kg with chair)  Walking speed               1.25 Km / h    Actuator AC Servo motor + Harmonic Reduction Gear + Drive Unit   Control unit Main controller + sub-controller and AC servo controller Sensor  Foot  3-Axis Force-Torque Sensor and Inclinometer  Torso     Rate-Gyro and Inclinometer      Power     External AC power (220V)    Operation Windows XP and RTX with Wireless network and joy stick       DOF                 12 DOF

2 is a conceptual diagram of the entire control system of the human riding biped robot, Figure 3 is a detailed configuration.

The control system of the human riding biped robot has a main controller 200 at the upper side, a motion controller interface device 100 at the middle position, and an AC servo motor controller 300 at the lower portion. And a plurality of sensors 400.

The human riding biped robot generally uses a PC as the main controller 200. The main controller 200 may use a general desktop, but it is preferable to use an industrial PC in consideration of power consumption and size. The master controller 200 communicates with the motion controller interface device 100 using the PCI bus. The master controller 200 uses Microsoft's Windows as an OS to familiarize the development environment, and to enable various applications.

In the present embodiment, the main controller 200 uses a computer having a performance of pentium 3 933Mhz, and used Microsoft's Windows XP as an OS. In addition, RTX software was used for real-time control of the robot.

4 is a view for explaining the structure of the RTX used in the main controller of the robot.

RTX enables real-time execution using the RTX scheduler in the HAL Extension. Since data exchange between Windows API and RTX is done through RTX shared memory, various information can be checked in real time through Windows application.

The main controller 200 is in charge of controlling the overall robot. The motion generated by the robot's motion plan is generated by inverse kinematics analysis to generate a reference for each axis, and perform real-time control of the robot through various sensor data .

The motion controller interface device 100 is connected to the plurality of AC motors 310 and the AC motor controller 300, and is connected to the sensors with a CAN cable.

The AC motor controller 300 is a device for driving the AC motor 310. When the AC motor controller 310 transmits a pulse and a direction corresponding to the position of the motor, the AC motor controller 310 controls the position of the motor according to the corresponding information.

The AC motor 310 is a drive unit used to drive each joint of the robot. In the case of the riding robot, an AC motor was used because it requires a large power, and in this embodiment, two types of motors, 400W and 800W, were used.

The plurality of sensors include a force / torque sensor for measuring the repulsive force with the ground and the torque applied to each ankle at each foot of the robot; An inertial sensor system for measuring the angle and the angular velocity of the robot with respect to the gravity direction; And an inclination sensor for measuring the inclination of the ground in contact with the foot of the robot.

Force / Torque sensor (410) is a device for measuring this information, because it needs to know the repulsive force and torque with the ground to stabilize the attitude of the robot.

An inertial sensor system 420 measures the degree to which the robot is inclined to stabilize the robot's posture. To this end, the inertial sensor system includes an accelerometer (tilt sensor) for measuring the degree of tilt of the robot relative to the direction of gravity, and a rate gyro for measuring the angular velocity at which the robot is tilted.

5 is a system configuration diagram of a motion controller of a human riding biped robot.

The motion controller interface device 100 includes a PCI interface unit for interfacing with the main controller 200 through a PCI bus, a digital signal processor (DSP) for generating a position command for the AC servo motor, and a plurality of sensors ( It is composed of a CAN communication unit for acquiring the data of 400) and transferring it to the main controller 200.

Each joint drive of the robot uses an AC motor, which uses a commercially available AC motor controller to drive-control it. The control mode of AC motor controller includes position control mode, speed control mode and torque control mode. To facilitate robot posture control and joint angle control, the control mode of AC motor controller uses position control mode. In the position control mode of the commercial AC motor controller, the angle of the motor is moved as much as the number of input pulses. For example, in the case of a motor that makes one rotation at 1000 pulses, one thousandth rotation is performed at one pulse. Rotate by). Therefore, the rotation angle and the angular velocity of the AC motor can be controlled by adjusting the number of pulses input for a certain time.

At this time, the constant time is the discrete sum of the derivative time, and if the number of differential time which makes up the constant time is not divided by the number of pulses to be generated within this constant time, the pulse is You cannot achieve equidistant intervals within a certain time. In this case, the pulse is generated as close to equidistant interval as possible, so that the angular velocity between commands is constant so that the angular velocity as well as the angle can be effectively controlled.

FIG. 6 is a diagram illustrating an internal configuration of a motion controller of the human-ride biped walking robot (HUBO FX-1), and FIG. 7 is a circuit diagram for exchanging data between a PCI interface chip, a DSP chip, and an AC motor in the motion controller. to be.

According to the present embodiment, the motion controller interface device 100 of the human-ride biped walking robot includes a PCI bus interface chip (PLX chip) 121, an oscillator 127 for a PCI interface chip, and a PCI interface chip ( 121) and a bus buffer (Buffer 5, 6, 7, 8) 130 connecting the DSP 140, the DSP 140 for pulse generation, the oscillator for DSP 141, the DSP 140 and the AC motor connector. Buffers for connecting a plurality of AC motor controllers through the 160 (Buffer 1,2,3,4) (150), AC motor connector 160 for connecting the motion controller interface device and the AC motor controller, CAN controller (111) ), Two photo couplers 112, a CAN transceiver 113, and a CAN connector 114.

/ WR is a write strobe signal of the PCI interface chip 121, / RD is a read strobe signal of the PCI interface chip 121, / ADS is an address strobe signal of the PCI interface chip 121, ALE is a PCI interface chip 121 Address latch enable signal, / CAN_CS is the CAN controller selection signal, CLK is the clock signal supplied to the CAN controller, RESET is the CAN controller reset signal, INT11 is the interrupt signal that the CAN controller places on the PCI bus, and AD8 ~ 15 is the CAN controller. Data line between PCI bus, RX is the receiving line of CAN communication, TX is the sending line of CAN communication, CAN_H is the high signal of CAN communication, CAN_L is the low signal of CAN communication.

The main controller (PC) 200 of the robot is connected to the PC104 + bus connector 122 of the motion controller interface device 100. Since the PC 104+ is an industrial standard PCI bus, the PC 104+ bus can be interfaced using the PCI interface chip 121.

The PCI interface chip 121 is a device that connects the devices to the PCI bus so that the main controller (PC) 200 of the robot and other devices can communicate.

The PCI interface chip 121 receives an angle control command for each motor from the robot main controller (PC) 200 at a cycle of, for example, 10 msec, and reads sensor data read from the CAN controller 111. To pass).

The DSP 140 generates a pulse according to the angle control command of each joint of the robot received from the PCI interface chip 121 and transmits the pulse to the AC motor controller.

The CAN controller 111 acquires data of a force / torque sensor, an inertial sensor, and an inclination sensor, and transmits the data to the main controller 200 through the PCI interface chip 121.

EPLD (102a, 102b) is an abbreviation of Electrically Programmable Logic Device can implement a logic circuit desired by the user with a simple device only. In this embodiment, the EPLD generates device selection signals (eg, / CAN_CS: CAN controller selection signal).

The DSP 140 uses TMSF2811 and generates pulses required to control an AC motor for driving each joint of the robot by using a digital differential analyzer (DDA) algorithm. The PCI interface chip 121 uses the DSP 140 because it is difficult to perform interrupts in a fast cycle. The DDA algorithm is described in detail later.

The input port of the DSP is a port A0 to 15 that receives a command from the PCI interface chip 121 and an input port where a signal MZ0 to 11 representing the Z phase of the motor from the AC motor controller 300 has passed through a buffer ( E2, F12 ~ 14, B12 ~ 15, D0 ~ 5). The output port of the DSP includes pulse command signal generation ports B0 to 11 for generating the rotation angle of the motor, direction command signal generation ports F0 to 11 for indicating the rotation direction, and command data from the PCI interface chip 121. It is composed of the transfer completion signal port E1.

The DSP 140 is driven at a relatively low voltage (e.g., 3.3V) to reduce heat generated due to fast calculations, while other components (PCI interface chip, EPLD, AC motor controller) are driven by the DSP 140. Because of operating at a higher drive voltage (eg 5V), all input-output signals of the DSP 140 pass through a buffer.

Buffer1-3 uses 3.3V to 5V buffer to adjust the voltage level between the two because the DSP output is 3.3V and AC motor control level is 5V. The signals B0 to B11 and F0 to F11 are converted into 5V signals by this buffer.

Buffer 4-8 is a 5V to 3.3V buffer for adjusting the voltage level between the EPLDs 102a and 102b and the DSP 140. In addition, the voltage is adjusted to the 3.3V level before delivering the Z phase of the AC motor to the DSP 140. The MZ0 to MZ11 and DATA0 to DATA15 values are converted to 3.3V by this buffer and transferred to the DSP 104.

Oscillator 1 (141) is a 30Mhz oscillator for driving the DSP 140, oscillator 2 (127) is a 20Mhz oscillator for driving the PCI interface chip 121.

The AC motor connector 160 is a connector for connecting signals (pulse0 to pulse11, sign0 to sign11) and Z phases (MZ0 to MZ11) necessary for controlling twelve AC motors.

Dip switches 1-3 (123) can stack multiple devices in the case of PCI104 / PCI104 +, to set a unique number for each device.

The CAN controller 111 is a device that enables controller area network (CAN) communication. The CAN transceiver 113 is a device that processes CAN_H and CAN_L transmitted and received from the CAN connector 114. The CAN connector 114 provides a function of connecting to each sensor and a connector (CAN connector).

8 illustrates a data configuration diagram for data exchange between the PCI interface chip 121 and the digital signal processor 140.

Since the PCI interface chip (PLX) 121 is driven at 100 Hz and the DSP is driven at 50 KHz, a protocol is required for data exchange between the two. In this embodiment, a total of 16 bits of data are exchanged, and the data format is as follows. In other words, 16bit data of each port for reading and writing data between PCI interface chip (PLX) and DSP is 4 bits using the motor ID to designate the motor and 1 bit to determine the rotation direction of the motor. And a data field for transmitting the rotation angle of the motor to the digital signal processor 104 using 11 bits. Such data must be delivered, for example, by the number of AC motors within 10 msec. In addition, a start bit and an acknowledge bit are required.

9 is a signal flow diagram between a PCI interface chip (PLX) and a DSP.

Referring to FIG. 9, start bits include RQ (request) and Find_Z (find z-phase), and 16 bits of Data [0:15] and an acknowledge bit of RT (return). The start bit sent from the PCI interface chip 121 to the DSP 140 is latched using an EPLD as shown in the figure to maintain for a desired time.

There are two types of start bits because two types of control methods for moving the motor to the position where the general motor position control and the Z phase of the motor are detected are required.

10 is a signal time diagram between a PCI interface chip (PLX) and a DSP.

A signal generation timing diagram for generating motor rotational pulses from the specified angle input is shown in FIG.

The RQ signal is generated by the PCI interface chip (PLX) 121, which turns on READY and latches in the EPLDs 102a and 102b to prepare to send data. DATA transmits the joint angle data to DSP 140 while the RQ signal is set, and when the data reading is completed in DSP 140, generates an RT signal to terminate the data signal and resets the latch of the RQ signal. The DSP 140 calculates the rotation direction and the rotation angle of the motor based on the read data to generate the direction signal and the pulse.

In case of transmitting position command, RQ is used as start bit. When RQ is used as a start bit, when the RQ signal is set high by the PCI interface chip (PLX) 121 to inform the DSP 140 that DATA is valid, the DSP 140 reads 16-bit data. After reading the data, a series of processes are completed by informing the PCI interface chip (PLX) 121 to set the RT signal to high, indicating that the data has been received normally.

This process is a series of processes to transmit data to one motor. Since the control cycle is 10msec, the above process should be performed 12 times for 10msec to control the 12-axis motor.

The Find_Z signal is a command signal that finds the encoder Z-phase inside the motor to find the recursive origin of the AC motor. When the Find_Z signal is set, the DSP 140 issues an angle command for rotating the motor at a constant speed. When the motor rotates at constant speed, the Z-phase signal of the encoder is transmitted to the DSP 140 through the MZ0 to 11 pins, and the DSP 140 generates the constant-speed rotation command of the motor until it detects the Z phase.

To find the Z-phase is to control the absolute position of the motor. In order to measure the absolute position, an absolute type encoder may be mounted. However, when the encoder signals are output in parallel, the wire becomes complicated, and when the signals are output in series, an additional device is required. When using an incremental encoder, an additional potential meter can be used to determine the absolute position, but in this case the design is complicated and requires an A / D converter device. For this reason, the absolute position control of the motor is possible through the incremental encoder through z-phase detection.

Finding a desired absolute position is to keep the initial starting state of the robot always the same. This process is very important when driving an actual robot or performing an experiment because the initial state must be the same to increase the repeatability of the experiment.

를 0에 가까운 상태를 유지하면, λ는 0이 된다. 항상 로봇이 이런 상태를 유지하기 위해서는

이 0이 되는 모터의 절대적 위치를 알고 있어야 한다.11 is a view showing a situation in which the height of the robot is changed by the initial position difference. With the mechanical construction of the initial robot finished

If is kept near 0, λ becomes 0. In order for the robot to stay in this state at all times,

You must know the absolute position of this zero motor.

이다. 오차를 0으로 만들기 위해 3가지 과정이 필요하다. 먼저, 현재 위치에서 z-상이 검출되는 위치까지 α만큼 이동한다. 이후, δ만큼 이동하여, 바람직한 절대 위치(desired absolute position)로 모터 축을 위치시킨다. 하지만 이 경우에도 N회전 만큼의 오차가 있기 때문에, 모터를 N회 회전하여, 원하는 절대 위치로 이동하면, ε 값이 0이 되어 원하는 초기 위치를 구성할 수 있다.12 shows the position of the motor shaft. The error in the previous figure

to be. Three steps are required to bring the error to zero. First, it moves by α from the current position to the position where the z-phase is detected. Then move by δ to position the motor axis in the desired absolute position. However, even in this case, since there is an error as much as N rotations, when the motor is rotated N times and moved to a desired absolute position, the value of epsilon becomes 0, so that a desired initial position can be configured.

Fig. 13 is a time diagram when finding the Z-phase of an AC motor.

When the signal named Find_Z is set high, the digital signal processor receives data as well. From the received data, it is checked which motor detects the Z-phase, and the motor is rotated at constant speed until the Z-phase is detected.

FIG. 14 is a diagram illustrating the number of pulses generated on the x-axis and the y-axis by the Digital Differential Analyzer (DDA) for a predetermined time.

In position control mode, the AC motor is driven by pulses according to direction and displacement. Thus, for example, how to generate a pulse for 10 msec is an important problem. If there are 10 pulses for 10 msec, one pulse can be generated for 1 msec, but if 9 pulses are required for 10 msec, it becomes ambiguous at what time to generate the pulse. In the case of controlling a plurality of motors, the intermediate path may show a great difference according to such a case.

In order to move to the (m, n) coordinate, m pulses are required in the X direction, and n pulses are required in the Y direction. Fig. 14 shows that pulses in the X direction and the Y direction are appropriately generated in order to move to the coordinates of (m, n). In order to generate such a pulse, the present invention generates a pulse by using a digital differential analyzer (DDA) algorithm.

FIG. 15 is a diagram illustrating a process of generating a pulse signal for controlling an AC motor by a digital differential analyzer (DDA) algorithm for a predetermined time.

C value is (DSP Interrupt Period) / (Period to receive command from PLX). In the present embodiment, since the interrupt period of the digital signal processing unit is 50 kHz, and the period of receiving a command from the PCI interface chip (PLX) 121 is 100 Hz, the C value is 500. The P value represents the amount of displacement the motor must move for 10 msec. The Q value is an internal variable for generating the Y value, and the Qmax value is the maximum number of pulses that can be generated for 10 msec. Each time the Y value increases by one, one pulse is generated. The Y value is from 0 to P value for a time of 10 msec. The C value is equal to Qmax as the maximum number of executions of a loop that generates a pulse between commands transmitted from the main controller (PC) 200 of the robot. Therefore, a loop is executed by C every time a command is transmitted from the main controller (PC) 200 of the robot. P is a value transmitted from the main controller (PC) 200 of the robot and is the number of pulses to be transmitted to the motor before the next command is transmitted. Qmax divided by P is the number of loops corresponding to the pulse time interval. If this value is not an integer, pulses cannot be generated at equal intervals. In this case, to make the pulse closest to the isochronous interval, generate the pulse as soon as the Count value becomes larger than the Qmax divided by P, and consider the rest of the Qloop divided by P in the next loop. If you pass the calculation to the next loop without ignoring the rest, you can solve the accumulated rest by executing one more loop as soon as the sum of the rest is equal to or greater than P. As a result, the difference in pulse intervals becomes a maximum of one loop to achieve the time interval closest to the equi-time interval.

In the human-ride biped robot (HUBO FX-1) with 12 DOF, two signals per position and direction are required. Therefore, a total of 24 outputs are required. In the case of the DSP (TMSF2811) used in this study, since there are 56 GPIOs in total, 12-axis motor control is possible.

16 is a diagram illustrating a connection relationship between a PCI interface chip (PLX) and a CAN controller.

The EPLD 102a contains logic circuitry that performs address decoding. When data is sent from the PCI interface chip (PLX) 121 to an address allocated for CAN communication, the EPLD 102a interprets the address and the chip select signal (/ CAN_CS) of the CAN controller 111 is generated. When this signal is generated, the CAN controller 111 assumes the data as valid data and transmits the corresponding CAN signal. When the CAN controller 111 receives the CAN signal from the outside, the interrupt signal is generated to the PCI interface chip 121 through the INT11 connection, and in response, the PCI interface chip 121 generates the PCI interrupt signal. In the interrupt processing routine of the main controller (PC) 200 of the robot, a CAN message is received by accessing an address allocated for CAN communication and reading data.

As described above, although described with reference to a preferred embodiment of the present invention, those skilled in the art various modifications of the present invention without departing from the spirit and scope of the invention described in the claims below Or it may be modified.

As described above, the present invention is interfaced with the main controller of the human-type bipedal walking robot using the AC motor, and using the motion controller interface device for transmitting sensor data to the main controller through CAN communication, It receives the angle command and generates the angle command of each joint in the form of pulse and controls the rotation angle and angular velocity of AC motors of the joint drive of many robots by adjusting the number of pulses generated for a certain time.

In addition, since the motion controller interface device according to the present invention can control a plurality of AC motors and perform a sensor interface, the motion controller interface device is suitable for a system requiring a plurality of multi-axis sensors such as a human-ride biped walking robot. When controlling the AC motor, the angle command of the robot joint generates pulses input from the command from the PC to the AC motor controller at the closest equidistant interval so that the angular velocity between the commands is constant. Take control effectively.

Claims (13)

A pulse command generation and sensor information acquisition device for generating a position command of a human riding robot joint driver, A sensor information acquisition unit configured to receive signals detected from a plurality of sensors in a CAN communication method; A PCI interface unit for transmitting signals detected from the plurality of sensors received by the sensor information acquisition unit to a main controller of the robot, and for transmitting an angle control command of each joint of the robot received from the main controller; Digital signal processing unit for generating a pulse signal as a motor control signal for controlling the motor driving each joint by receiving the angle control command of each joint of the robot from the PCI interface unit and adjusting the number of pulses generated for a predetermined time ; A motor connector for transmitting a motor control signal output from the digital signal processor to each motor controller; Pulse command generation and sensor information acquisition device for generating a position command of the human-type robot joint driver including a. The method of claim 1, The motor for driving each joint of the robot is an AC motor, pulse command generation and sensor information acquisition device for generating a position command of the human-type robot joint driver. The method of claim 1, The digital signal processing unit, In order to distribute the time interval between the pulses that should be generated for a predetermined time by receiving the angle command of each joint from the main controller of the robot through the PCI interface unit of the human riding robot joint driver, characterized in that Pulse command generation and sensor information acquisition device for position command generation. The method of claim 1, The input port of the digital signal processing unit includes a port for receiving a command from the PCI interface unit, a port for receiving a signal from the motor controller, The output port of the digital signal processor includes a pulse command signal generation port for a rotation angle of the motor, and a direction command signal generation port for indicating a rotation direction of the motor. Pulse command generation and sensor information acquisition device. The method of claim 1, The angle control command of each joint of the robot Human-type robot joint comprising 4 bits of motor ID to designate the motor, 1 bits of sign bit to determine the rotation direction of the motor, and 11 bits of data bit to indicate the rotation angle of the motor. Pulse command generation and sensor information acquisition device for generating position command of the driver. The method of claim 1, The plurality of sensors A force / torque sensor for measuring the repulsive force against the ground and the torque applied to each ankle at each foot of the robot; An inertial sensor system for measuring the angle and the angular velocity of the robot with respect to the gravity direction; And Pulse command generation and sensor information acquisition device for generating a position command of the human-type robotic joint drive, characterized in that it comprises an inclination sensor for measuring the inclination of the ground in contact with the foot of the robot. The method of claim 6, The force / torque sensor, It is installed on each foot of the robot and measures the force in one direction and the moment in two directions, and uses auto-balancing and auto nulling point control algorithms. Pulse command generation and sensor information acquisition device for generating position command of the human-type robotic joint driver. The method of claim 6, The inertial sensor system, Pulse command generation and sensor information for generating position command of the human-mounted robot joint driver, comprising an accelerometer for measuring the degree of tilt of the robot relative to the direction of gravity and a rate gyro for measuring the angular velocity of tilting of the robot Acquisition device. The method of claim 1, A first buffer unit for lowering a driving voltage of the PCI interface unit to a driving voltage of the digital signal processing unit; And A second buffer unit for raising the driving voltage of the digital signal processing unit to a driving voltage of a motor controller controlling the motors driving the respective joints; Pulse command generation and sensor information acquisition device for generating a position command of the human riding robot joint driver further comprising. As a joint driver control device of a human riding robot, A plurality of sensors for measuring the position of the robot and the state of the ground; A main controller of the robot for receiving sensor data measured from the plurality of sensors and providing an angle control command for each joint of the robot; And A motion controller configured to receive an angle control command of each joint of the robot from the main controller of the robot and generate a pulse signal required to control the joint driving unit of each joint of the robot; Joint driver control device for a human riding robot comprising a. The method of claim 10, The joint drive control apparatus of the human riding robot, characterized in that each of the joint drive unit has an AC motor and an AC motor controller. The method of claim 10, The motion controller, Receiving the angle command of each joint of the robot from the main controller of the robot and using the DDA algorithm to distribute the time interval between the pulses to be generated for a predetermined time from the pulse signal for controlling the AC motor of the corresponding joint drive unit Joint driver control apparatus for a human riding robot, characterized in that. As a method of controlling a joint driver of a human riding robot, (a) receiving sensor data measured from a plurality of sensors in a CAN communication scheme; (b) a main controller (PC) of the robot generating an angle control command of each joint of the robot based on the sensor data; (c) The motion controller receives the angle control command of each joint of the robot and adjusts the interval of pulses generated for a predetermined time by the DDA algorithm to provide the motor control signal necessary to control the AC motor driving each joint of the robot. Generating; And (d) transmitting the motor control signal to an AC motor controller of an AC motor that drives each joint of the robot; Joint driver control method of a human riding robot comprising a.

Images