KR101234204B1 - Real-time monitoring method for distributed robot control with high reliability - Google Patents

Abstract

The present invention relates to a real-time monitoring method for high reliability distributed real-time control using a plurality of multi-core processor of the robot, the technical problem to be solved is to check and record the operating state of the robot, and to detect, analyze, cope with and record the transient abnormality It is to provide a distributed real-time monitoring method of the robot that performs the operation, and records the detailed event for post-mortem analysis in case of permanent abnormality.
To this end, the distributed real-time monitoring method of the robot according to the present invention includes an operation state confirmation recording step of confirming and recording the operation state of the robot periodically and aperiodically, and analyzing and coping with it when transient abnormality is detected in the robot. Transient abnormality detection, analysis, countermeasure and recording steps for restoring the robot to the normal state, and permanent abnormality recording step for comprehensively recording the occurrence time, contents, internal state, etc. of the detailed events of the sequential event record for the permanent abnormality of the robot. Characterized in that it comprises a.

Description

REAL-TIME MONITORING METHOD FOR DISTRIBUTED ROBOT CONTROL WITH HIGH RELIABILITY}

The present invention relates to a distributed real-time monitoring method of a robot, for the high reliability and distributed control of a robot using a plurality of multi-core processors, operation status, transient fault determination, analysis, response and recording, permanent fault (permanent) A distributed real-time monitoring method of a robot that presents a method of recording a fault).

Intelligent robots actively provide a variety of services to users while interacting with changing surroundings, and for this purpose, complex processes such as image processing, autonomous driving, voice recognition, motor and sensor control, dialogue with users, and task execution are performed. All features are included.

Control software tasks required for intelligent robots to perform these functions include tasks that interface high-performance hardware sensors to detect the environment, such as distance-based sensors, vision sensors, and voice sensors, and sensor interfaces. Tasks for recognizing and knowledge of people, objects or voices by extracting features from imported environmental data, and tasks for intelligently planning work of robots through inference for human interaction based on the recognized knowledge The moving unit, the operation unit, which implements physical actions by receiving commands from the task control tasks and the task control tasks such as reliably moving to the target position or safely grabbing or manipulating the target object while avoiding obstacles according to the work plan. Of robots like head pan / tilt There are tasks which the synchronous interface.

On the other hand, the inventor of the related patent publication "10-2010-0048857" associated with the intelligent robot is composed of a robot application management unit, life cycle management unit, robot application component, sensor and actuator unit, data port, for data transfer between tasks, The event port, method port, etc. are defined.If necessary, the task can be operated when the state of the finite state machine is changed by the input of the event.The task can be called back when an error occurs and the error is recovered. Although a robot software task management technique has been proposed that may include a callback function, there is a problem in that the present invention lacks a specific specification of an error determination method, a recovery method, and the like.

In addition, the registered patent number "10-0753054" is a network connection for fault tolerant support of module-based personal robots to support fault tolerant such as detection of unforeseen network errors, error reporting, and network reconnection. The invention relates to a maintenance system, a method of maintaining a network connection, and a recording medium recording the same.A fault tolerant method of connection between modules that enables error detection, error reporting, module isolation, and module reuse of a network connection is presented. However, there is a problem in that the recovery method is insufficient because only a passive method of restoring an error after the error detection depends on the reconnection status report message upon reconnection is present.

In addition, the registered patent number "10-0637056" is equipped with an abnormality detection device for the humanoid walking robot to determine the abnormality of the internal state quantity or the internal sensor value. Although a method of storing the state quantity of the memory and the like in the internal memory and the memory of the remote control computer has been proposed, there is a problem in that no recovery countermeasure is provided in the present invention, and the registered patent number "10-0877715" discloses a single processor. Although the structure of the adopted real-time robot software is presented, there is a problem in that the structure for multiple processors is lacking.

The present invention has been invented to solve the above problems, for the high reliability robot control using a plurality of multi-core processor, to check the periodic / non-periodic robot operation status, recovery for detection and coping with transient abnormalities It is an object of the present invention to provide a distributed real-time monitoring method of a robot that provides a method and a method of comprehensively recording the contents, occurrence time, internal state, and the like of permanent events.

Another object of the present invention is to extend the normal operating period of the robot, to prove the reliability and certainty of the robot operation, and to improve the hardware and software through post-mortem analysis by accident recording. It is to provide a distributed real-time monitoring method.

In order to achieve the above object, the distributed real-time monitoring method of the robot according to the present invention includes an operation state checking and recording step of confirming an operation state of the robot, and detecting and analyzing the transient abnormality when a transient abnormality occurs in the robot. A transient abnormality detection, analysis, coping and recording step of restoring the robot to its normal state through coping, and a permanent abnormality recording step of recording a series of situations for detection, analysis, and coping of permanent abnormality. .

In addition, the operation status check recording step may be performed by a task manager in each processor to record aperiodic robot operation for driving, exception occurrence and stop of each task of the software for operating the robot.

The task manager may be divided into a user level task manager and an operating system level task manager.

In addition, the task manager may exchange data with the task manager of another processor through a network.

In addition, the task manager may record the details of the details of the event, the date and time of occurrence of the task and the internal state information of the task in order to record the detailed events such as the start of the task, an exception occurrence event, and a stop.

In addition, the operation status check recording step is performed by the task manager to perform the periodic robot operation recording for the detection and recording of the voltage and current of the power source for driving the robot, the measurement of the temperature of the processor and the drive motor, the detection and recording Can be.

In addition, the task manager may record the cycle and time required for the interrupt service routine for asynchronous external signal processing on a recording medium.

In addition, the task manager may record the statistical data including the real time task period, the execution time, the number of success times and the number of failures in a certain time interval.

In addition, the transient abnormality detecting, analyzing, coping, and recording step transmits an alarm message to the user when the periodically measured power supply voltage is less than or equal to the preset first reference voltage, and the preset power supply voltage is less than the first reference voltage. When the voltage is less than or equal to the second reference voltage, the power may be charged.

In addition, the transient abnormality detecting, analyzing, coping, and recording steps may be performed when the temperature of the processor periodically measured is greater than or equal to a preset first reference temperature, and the cooling fan of the processor is operated at a maximum speed, and the temperature of the processor is set to the first temperature. If the reference temperature is higher than the second reference temperature higher than the reference temperature can stop the task of the robot.

In addition, the transient abnormality detecting, analyzing, coping, and recording steps may reduce the maximum allowable speed of the driving motor when the temperature of the driving motor measured periodically is equal to or greater than a first reference temperature, and the temperature of the driving motor may be set to the first. When the temperature is equal to or greater than the second reference temperature higher than the first reference temperature, the driving motor may be stopped.

In addition, the transient anomaly detection, analysis, coping and recording step is to record the detailed event of the sensor when the measured value of the simple sensor is outside the minimum value or the maximum value range, the processor timer ( Restart the timer, initialize or restart the sensor driver, reset or restart the sensor driver when the sensor detection time exceeds the preset reference time, and if the watchdog timer of the sensor triggers an alarm, The driver can be initialized or restarted.

In addition, the transient anomaly detection, analysis, coping and recording step is to perform the detailed event recording of the sensor when the measured value of the intelligent sensor is outside the minimum value or the maximum value range, the microcontroller ( restarts the timer of the microcontroller, restarts the sensor interrupt service routine, initializes or restarts the sensor driver when the sensor detection execution time exceeds the preset reference time, and the watchdog of the sensor If a watchdog timer generates an alarm, the sensor driver can be initialized or restarted.

In addition, the transient abnormality detecting, analyzing, coping, and recording may be initiated or stopped and restarted when the controller abnormality is detected.

In addition, the transient abnormality detecting, analyzing, coping, and recording steps may include initializing or software reset the driving controller when the deviation between the driving command and the driving result of the actuator is greater than or equal to a preset reference deviation. If is greater than or equal to the reference deviation, the drive controller is hardware reset or the power is turned off and then turned on.In case of periodic reporting error, the drive controller timer is restarted or hardware reset is performed. It can be turned on after the power is turned off.

In addition, the transient anomaly detection, analysis, coping, and recording steps may reset the main processor system when the watchdog timer of the main processor generates an alarm and reload and start the entire program of the main processor.

In addition, the transient anomaly detection, analysis, countermeasure and recording step may be performed by resetting the secondary processor system in the event that the watchdog timer of the main processor generates an alarm or a periodic reporting abnormality or an aperiodic abnormality. Can be reloaded and started.

In addition, the permanent abnormal recording step may record the occurrence of abnormal events, abnormal execution stop, etc. for each of the tasks for which the permanent abnormality occurred.

In addition, in the real-time monitoring method, a series of situations regarding the operation status, transient abnormality, and permanent abnormality may be sequentially recorded in files by date and type on a recording medium such as an internal memory, a USB flash memory, or an internal hard disk of an external computer. have.

As described above, according to the distributed real-time monitoring method of the robot according to the present invention, comprehensive situation data such as contents, date and time, and internal state of periodic / non-periodic detailed events can be obtained, and transient abnormality detection, determination, and response can be obtained. And it is possible to record the robot can continuously perform the operation with the reliability and certainty as possible, there is an effect that can be utilized to improve the hardware and software through the post-analysis of the robot through the accident record of the permanent abnormality.

1 is a state transition diagram of the distributed real-time monitoring method of the robot according to an embodiment of the present invention.
2 is a schematic diagram of software for implementing a distributed real-time control and monitoring method of a robot according to an embodiment of the present invention.
3 is a structural diagram of software for implementing a distributed real-time control and monitoring method of a robot according to an embodiment of the present invention.
4 is a detailed structural diagram of processor 2 (CPU 2) of the plurality of processors shown in FIG. 3;
FIG. 5 is a detailed structural diagram of processor 1 (CPU 1) of the plurality of processors shown in FIG. 3; FIG.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. First, it should be noted that the same components or parts among the drawings denote the same reference numerals whenever possible. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted so as not to obscure the subject matter of the present invention.

1 is a state transition diagram of a distributed real-time monitoring method of the robot according to an embodiment of the present invention, Figure 2 is a schematic diagram of software for implementing a distributed real-time control and monitoring method of the robot according to an embodiment of the present invention 3 is a schematic diagram of software for implementing a distributed real-time control and monitoring method of a robot according to an embodiment of the present invention.

The distributed real-time control method of the robot according to an embodiment of the present invention, as shown in Figure 1, the operation status check recording step (S10), transient abnormality detection, analysis, coping and recording step (S20) and permanent abnormality recording Step S30 is included.

Specifically, when the robot is normally controlled, the monitoring system is in the operation status check recording step (S10), when the transient abnormality (1) is detected, the transient abnormality detection, analysis, coping and recording step (S20). When the robot is restored (2) by the above measures, the operation returns to the operation state confirmation recording step (S10) again. In addition, if a permanent abnormality 3 is detected by the robot, the process proceeds to the permanent abnormality recording step S30. In this step, the robot can be normalized again through ex post analysis through replacement of control software and hardware repair.

At this time, the software tasks for implementing a distributed real-time control and monitoring method of the robot according to an embodiment of the present invention, as shown in Figure 2, a plurality of functions having a multi-core (multi-core) function of the robot Are distributed to multi-processors 100, 200,. The user interface, knowledge-based high-level planning tasks, and the like are distributed and distributed at the user level 90, and the tasks related to sensors and driving hardware are distributed and distributed at the operating system level 70, and the robot is provided through the appropriate hardware interface 20. It is connected to the sensor, the driver, the mechanism (10). At this time, each processor can send and receive information through an effective appropriate network means.

Specifically, as shown in FIG. 3, the distributed real-time control software may be performed by four dual core processors 100, 200, 300, and 400 sharing the functions of movement, manipulation, voice processing, and image processing, respectively. have.

In other words, the software in each processor core is responsible for the task of planning the movement, manipulation, and the like of a planner layer 60 that performs the highest level robot-user interface, mission planner, and the like. In addition to the sequencing layer 50 included therein, it can be composed of various reactive layers 30 and 40 including sensors, actuators, and judgment / control.

The planning layer 60 and the sequential processing layer 50 are typically implemented in user level tasks, and the immediate response layers 30, 40 are typically implemented in operating system kernel level tasks, such as Xenomai. In a real-time operating system, the real-time operating system level 30 and the real-time operating system level 30 may be implemented. Various controller tasks are arranged at the real time user level 40, and sensor drivers and driver drivers are arranged at the operating system level 30.

In this case, the controller tasks may manage hardware 10 such as a sensor, a driver, and a mechanical part of a robot through various drivers 30 and hardware interfaces 20. In order to connect mechanical parts such as a robot's mobile device, an operation arm, a head, a face, and various images, audio, and environmental sensors with the plurality of processors, an appropriate interface board is required for the hardware interface 10.

4 is a detailed structural diagram of the processor 2 200 of the plurality of processors illustrated in FIG. 3.

In detail, as shown in FIG. 4, when the processor 2 200 controls the robot arm, the robot's manipulation arm, that is, the robot arm driver 10f may be formed of several joints, and each joint may be conventional. It can be driven by a motor, each motor may be equipped with a motor controller based on a microcontroller (microcontroller).

The motor controller may exchange data with the outside through a communication means such as RS485 or a controller area network (CAN), and further, an RS485 / CAN interface board for transmitting a command to the robot arm driver 10f and receiving a report. 20f can be mounted. At this time, the software for driving the interface board 20f is the RS485 / CAN driver 30f.

On the other hand, the data obtained from the various sensors (10e, 10g) of the processor 2 (200) is stored in the form of a shared memory in the sensor resource (sensor resource, 39b) can be used later in the real-time operating system level and user level.

The shared memory may be directly read by the real time operating system level task, but the real time user level task may acquire sensor data through a standardized interface called an input device interface (IDIF) 47b or 47c. 47b, 47c) may be used to obtain data of user level tasks in various languages.

In order for the processor 2 200 to control the robot arm, the execution order of the given task of the robot is determined by the mission planner task 60e of the planning layer 60, and the corresponding robot arm is driven. The order is determined by the sequential management planner 50e, and this driving order is controlled by the arm control 40e of the immediate response stratum 30, 40 by the AIF (Action Interface 49b). Is delivered to.

When the arm controller 40e commands an arm driver 35f to know the current state of the robot arm, the arm driver 35f is converted to a protocol for valid and proper robot arm communication. The command is transmitted to the RS485 / CAN driver 30f.

The RS485 / CAN driver 30f transmits a proper physical signal to the robot arm driver 10f through the RS485 / CAN interface 20f, receives a position value of each joint, and transmits it to the sensor resource 39b. Save the value.

Meanwhile, when force control of the robot arm is required, the force / torque signal generated from the force / torque sensor 10e is transmitted through the force / torque interface 20e. Is the F / T driver task 30e.

At this time, the obtained force / torque sensor value is similarly stored in the sensor resource 39b, and the robot arm controller 40e is from the driving order command value and sensor resource 39b received through the AIF 49b from the upper side. Based on the current position and the force / torque value read out, an effective algorithm is used to calculate the drive value. The drive value is based on the arm driver (35f), RS485 / CAN driver (30f), and RS485 / CAN interface (20f). It is transmitted to the robot arm driver 10f through it to move the robot arm to the desired position.

On the other hand, when grasping that the robot grabs an object is required during the robot's mission, the grasp planer 50g determines the order of grasping motions, and this command determines the output device interface 48a (ODIF). It can be transmitted to the hand driver (hand driver, 35h), in this case, is transmitted to the robot hand (10h) through another RS485 / CAN driver 35h and RS485 / CAN interface 30h is driven.

When the contact force value of the finger is required in the grip planner 50g, the magnitude of the contact force obtained through the touch sensor 10g may be the touch interface hardware 20g, the touch driver software 30g, the sensor resource 39b, or the IDIF ( 47c) may be delivered to the planner 50g.

As described above, the operation state confirmation recording step S10 is a step of confirming and recording an operation state of software tasks in the plurality of multi-core processors.

In order to operate the robot, a number of tasks such as sensors, judgment, control, and driving are started when the robot starts up software, and when the robot is stopped, various tasks are stopped on the contrary. After is loaded first, the upper task must be loaded for normal operation.

In other words, in order to smoothly control the robot, a plurality of tasks must be started and stopped in a predetermined order, in which order the tasks are started, when they are stopped, and what exceptions occur in the middle. Robot operation records are important to demonstrate the normal operation of the robot.

FIG. 5 is a detailed structural diagram of processor 1 (CPU 1, 100) of the plurality of processors shown in FIG.

In the operation status confirmation recording step S10, aperiodic robot operation recording of driving, exception occurrence, and stop of a software task operating the robot may be performed by a task manager in each processor.

In detail, as shown in FIG. 5, the task managers 68a and 48a of the processor 1 100 may perform the aperiodic robot operation recording, wherein the task managers 68a and 48a may be at a user level. The task manager 68a and the operating system level task manager 48a may be divided and operated. This is because the system call method is different at the user level and the operating system level.

In addition, the task managers 68a and 48a may exchange data with a task manager of another processor through a network 20n.

The task managers 68a and 48a may be typically executed in the first core in one processor, and may be in charge of instructions and operation records for tasks performed in the first core as well as tasks performed in the other core. In this case, information may be stored using a shared memory.

In detail, the task managers 68a and 48a provide detailed event contents, date and time of occurrence, and internal state information of the corresponding task for recording the detailed event by the event, exception, stop, etc. of starting the task and changing the internal state. Can be recorded on the recording medium. In addition, if necessary, the size of the recording file can be reduced by adding an optional recording function that enables selective recording in accordance with each event.

In this case, the recording medium may be formed of an internal memory such as a hard disk, a removable memory such as a USB serial memory (USB Flash memory), and externally developed / connected when connected to a development / operation computer through a wireless network or the like. It may be an internal hard disk of the operating computer.

Accordingly, the aperiodic robot operation record is stored in the internal memory, the USB flash memory, and the external hard disk in the same manner, thereby minimizing the loss of data. It can be done easily.

Meanwhile, in addition to the aperiodic robot operation recording, the operation state confirming recording step (S10) includes sensing and recording the voltage and current of the power source for driving the robot, and measuring, sensing and recording the temperature of the processor and the driving motor. Periodic robotic operation records may be performed by the task managers 68a and 48a.

Specifically, the task managers 68a and 48a may measure the remaining capacity by periodically sensing and recording the voltage and current of the power supply, instead of directly measuring the remaining capacity of the battery which is a normal operating element of the robot. By periodically measuring, sensing, and recording the temperatures of the processor and the driving motor during robot operation, it is possible to measure whether the processor and the driving motor operate normally.

In addition, the task managers 68a and 48a may determine whether each task operates smoothly periodically, and a period of a real time task that is a driving time interval of each task, an execution time of each task, and a periodic task period in a predetermined time interval. And the average value, the minimum value, the maximum value of the execution time may be determined and recorded, and the number of successful executions, the number of failed executions, the number of executed errors, and the like may be recorded.

In addition, the task managers 68a and 48a may detect the period and the time required for the interrupt service routine for the asynchronous external signal processing associated with each task and record the same in the recording medium.

As described above, the task managers 68a and 48a may have information exchange and cooperative work 81a functions of the user task manager 68a and the real-time task manager 48a, and each task performed on several cores. It can have load, initialize, start, stop, and write functions 82a and 83a, and communicates with task managers of other processes via the network interface 20n 84a. )

In addition, the task managers 68a and 48a may have a hardware reset function for each hardware sensor, a driver, and a power on / off control function 85a to 85e through a power distribution module, and a cooling fan start function and a timer restart. It can have a running function, a software reset function for the processor, and a periodic / aperiodic robot operation recording function and an accident recording function.

The transient abnormality detecting, analyzing, coping, and recording step (S20) is a step of recovering the robot to its normal state through the transient abnormality detecting, analyzing, and coping when the abnormality occurs in the robot, and recording the same.

The normal operation period of the robot may be extended when the appropriate abnormality is determined and coped with in response to the transient abnormality generated during operation of the robot in the step S20 of detecting, analyzing, coping, and recording the transient abnormality.

In the transient abnormality detecting, analyzing, coping, and recording step (S20), if an abnormality occurs in the power supply, that is, if the power supply voltage is less than or equal to the preset first reference voltage after periodically measuring the power supply voltage, energy is insufficient. If it is determined, an alarm message can be transmitted to the user, and if the power supply voltage is less than or equal to the predetermined second reference voltage less than the first reference voltage, the operation in operation is stopped (for example, the cup held by the robot hand is put down). ), It can be moved to a charging station to charge the power.

In addition, in the abnormal abnormality detecting, analyzing, coping, and recording step (S20), when an abnormality occurs in the temperature of the processor, that is, after periodically measuring the temperature of the processor, the temperature of the processor is equal to or greater than a preset first reference temperature. If the temperature is determined to be excessive, the cooling fan of the processor may be driven at the maximum speed, and if the temperature of the processor is higher than the second reference temperature higher than the first reference temperature, it is determined that the computer is unreasonable to perform the task of the robot. Can freeze and stop the entire computer, including the processor

In addition, in the step S20 of detecting, analyzing, coping, and recording the transient abnormality, when an abnormality occurs in the temperature of the driving motor, that is, after periodically measuring the temperature of the driving motor, the temperature of the driving motor is preset. When the temperature is higher than the reference temperature, the maximum allowable speed of the driving motor may be reduced by determining that the temperature of the driving motor is excessive. When the temperature of the driving motor is higher than the second reference temperature higher than the first reference temperature, the motor cannot be driven. In this case, the motor is stopped immediately, and in this case, the brake can be driven simultaneously to stop.

On the other hand, in general, a simple sensor is a sensor that outputs an analog quantity in a case where a microcontroller does not exist in the sensor and hardware that simply amplifies a signal exists. Etc., the sensor value can be converted to digital.

In contrast, an intelligent sensor is a sensor in which a microcontroller is attached to the sensor to perform signal conversion and calculation, and outputs the amount of digital data using an appropriate communication method. The sensor value can be accepted via.

In the transient abnormality detecting, analyzing, coping, and recording step (S20), when an abnormality occurs in the simple sensor, specifically, when the measured value of the simple sensor is outside the minimum value or the maximum value range, it is determined as a sensor failure. Detailed event recording of sensor abnormalities can be performed.

In addition, when the sensor detection cycle is abnormal, the processor timer may be restarted and the sensor driver may be initialized or restarted by determining that the processor timer or the sensor driver is abnormal. When the sensor detection execution time is longer than the preset reference time, the sensor driver is abnormal. In this case, the sensor driver may be initialized or re-driven. When the watchdog timer of the sensor generates an alarm, the sensor driver may be determined to be an error and may be initialized or re-driven.

On the contrary, when an abnormality occurs in the intelligent sensor in the step S20 of detecting, analyzing, coping, and recording the transient abnormality, specifically, when the measured value of the intelligent sensor is outside the minimum value or the maximum value range, the sensor malfunctions. By judging, detailed event recording of sensor abnormality can be performed.

In addition, when the sensor detection cycle is abnormal, the controller may restart the microcontroller and restart the sensor interrupt service routine by determining that the microcontroller has a timer or sensor interrupt. If it is more than the reference time, the sensor driver may be initialized or restarted by determining that the sensor driver is abnormal. If the watchdog timer of the sensor generates an alarm, the sensor driver may be determined to be abnormal and may be initialized or restarted.

In this case, when the above abnormality is repeated twice, the microcontroller may be initialized in hardware or may be turned on again after turning off the power of the microcontroller.

On the other hand, in the transient abnormality detecting, analyzing, coping and recording step (S20), when an abnormality occurs in the controller, that is, when a period abnormality of the controller is detected, it is determined as a controller error to initialize or stop the controller task. Can be driven again

In addition, in the case of the abnormal abnormal detection, analysis, coping and recording step (S20), when an abnormality occurs in the actuator (actuator), it can effectively cope with the following effectively.

In general, an actuator typically includes a digital signal processor (DSP) to perform a driving function. The actuator may be configured to receive a driving result periodically from the driver after transmitting a driving command to the driver.

When an abnormality occurs in such an actuator, that is, when the deviation between the driving command and the driving result of the driver is equal to or more than the preset reference deviation, it is determined that the DSP is abnormal and initializes the DSP or, if possible, performs a software reset. You can do this by hardware reset of the DSP or by turning the power off and then on again.

In addition, if the deviation is more than the reference deviation for a predetermined time or more, it can be judged as DSP error and reset the hardware or turn the power off and then on again.In case of periodic reporting error, restart the DSP timer or DSP Can be turned on again after a hardware reset or power off.

In addition, the transient abnormality detection, analysis, response and recording step (S20) in the case of an abnormality in the main processor, that is, when the watchdog timer of the main processor is an alarm is determined as an error of the overall software to determine the main processor system It can reset and reload and start the entire program of the main processor.

On the other hand, the transient abnormality detection, analysis, coping and recording step (S20) when the abnormality occurs in the sub-processor set to report the status to the main processor periodically and also to report aperiodic confirmation, that is, In the event of a watchdog timer alarm, periodic reporting fault, or non-periodic check fault, all can be determined to be an error in the overall software to reset the subprocessor system and reload and start the entire program of the subprocessor.

The permanent abnormal recording step (S30) is a step for detecting and recording the permanent abnormality.

Specifically, in the permanent abnormal recording step (S30), the occurrence of abnormal events, abnormal execution stops, and the like for each of the tasks for which the permanent abnormality occurred may be recorded on the recording medium.

In this case, the recording medium may be formed of an internal memory, a USB flash memory, an internal hard disk of an external computer, or the like, and a series of situations of the transient abnormality are recorded on the recording medium by date and type to facilitate later retrieval. can do.

In general, it is very important for the post-accident analysis that the hardware, such as sensors and drivers, and the sequence of the driver software for driving them and the controller software tasks for judgment and control have failed in the event of a robot accident. It can be used to distinguish between causes and dependent failures and to find ways to deal with major causes.

As described above with reference to the drawings illustrating a distributed real-time monitoring method of the robot according to the present invention, the present invention is not limited by the embodiments and drawings disclosed herein, but within the technical scope of the present invention Of course, various modifications may be made by those skilled in the art.

S10: Operational Status Check Recording Step
S20: Transient Anomaly Detection, Analysis, Response and Recording Steps
S30: Permanent Abnormal Recording Step
10: Hardware
20: Interfaces
30: Kernel Real-time
40: User-real-time
50: Sequencing
60: Deliverative
100, 200, 300, 400, 900: Processor (CPU)

Claims (19)

For smooth operation of distributed control of a robot using a plurality of multi-core processors,
An operational status confirmation recording step of performing an operation status confirmation and recording;
A transient abnormality detecting, analyzing, coping, and recording process of restoring the robot to a normal state by detecting, analyzing, coping, and recording the transient abnormality when the robot has a transient abnormality; And
Permanent abnormality recording step of recording a series of sequential situations when the permanent abnormality occurs in the robot,
The operation status check recording step,
Aperiodic robot operation recording of the driving, exceptioning event and stopping of each task of the software running the robot is performed by the task manager in each processor,
The task manager is divided into a user level task manager and an operating system level task manager is operated in a distributed real-time monitoring method of the robot.
delete delete The method of claim 1,
The task manager is a distributed real-time monitoring method of the robot, characterized in that for exchanging data with the task manager of the other processor through the network.
The method of claim 1,
The task manager records the date and time of occurrence, the details of the detailed event, and the internal status information of the task on a recording medium for recording the start of the task, the exception event, and the detailed event of the stop. Way.
The method of claim 1,
The operation status check recording step,
A distributed real-time monitoring method of a robot, characterized in that the task manager performs the recording and recording of the voltage and current of the power source for driving the robot, and the measurement and recording of the temperature of the processor and the driving motor. .
The method according to claim 6,
The task manager is a distributed real-time monitoring method for a robot, characterized in that for recording the cycle and the time required for the interrupt service routine (interrupt service routine) for asynchronous external signal processing on a recording medium.
The method according to claim 6,
The task manager is a distributed real-time monitoring method of the robot, characterized in that for recording the statistical data including the cycle, execution time, the number of success times and the number of failures in a certain time interval of the real-time tasks.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
If the periodically measured power supply voltage is lower than or equal to the preset first reference voltage, an alarm message is sent to the user.
If the power supply voltage is less than the second reference voltage less than the first reference voltage is less than the first reference voltage distributed real-time monitoring method of the robot, characterized in that for switching to the task of charging.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
If the periodically measured temperature of the processor is greater than or equal to the first predetermined reference temperature, the cooling fan of the processor is driven at the maximum speed,
And discontinuing the task of the robot when the temperature of the processor is greater than or equal to the preset second reference temperature higher than the first reference temperature.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
When the temperature of the drive motor measured periodically is equal to or greater than the first preset reference temperature, the maximum allowable speed of the drive motor is decreased,
And disabling the driving motor when the temperature of the driving motor is greater than or equal to a second predetermined reference temperature higher than the first reference temperature.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
If the measured value of the simple sensor is outside the minimum or maximum value range, record the detailed event of the sensor.
If it is longer than the sensor detection period, restart the processor timer, reset or restart the sensor driver,
If the sensor detection execution time is longer than the preset reference time, the sensor driver is initialized or restarted.
A distributed real-time monitoring method of a robot, wherein the sensor driver initializes or restarts a sensor driver when a watchdog timer of the sensor generates an alarm.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
If the measured value of the intelligent sensor is outside the minimum or maximum value range, record the detailed event of the sensor.
If the sensor detection cycle is over, restart the timer of the microcontroller, restart the sensor interrupt service routine,
If the sensor detection execution time is longer than the preset reference time, the sensor driver is initialized or restarted.
A distributed real-time monitoring method of a robot, wherein the sensor driver initializes or restarts a sensor driver when a watchdog timer of the sensor generates an alarm.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
If a period of a controller is detected, the robot real-time distributed real-time monitoring method characterized in that the controller task is initialized or stopped and restarted.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
If the deviation between the driving command and the driving result of the actuator is more than the preset reference deviation, initialize the driving controller or reset the software.
If the deviation is more than the reference deviation for a predetermined time, the drive controller is hardware reset or the power is turned off (on), then turned on (on),
In the event of periodic reporting abnormality, a method for distributed real-time monitoring of a robot comprising restarting a drive controller timer, hardware reset of the drive controller, or turning on and off the power.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
A method of distributed real-time monitoring of a robot, characterized in that if the watchdog timer of the main processor generates an alarm, it resets the main processor system and reloads and starts the entire program of the main processor.
The method of claim 1,
The transient anomaly detection, analysis, response and recording step,
The distributed real-time of the robot, characterized by a reset of the secondary processor system and reloading and starting the entire program of the main processor in the event that the watchdog timer of the main processor generates an alarm, or in the case of periodic reporting anomalies or anomalies, Surveillance Method.
The method of claim 1,
The permanent abnormal recording step,
Distributing real-time monitoring method of the robot, characterized in that to record the occurrence of abnormal events, abnormal execution stop for each of the tasks in which the transient abnormality occurred.
The method of claim 1,
The real time monitoring method,
Distributed real-time monitoring of the robot, characterized in that the operation status, transient abnormality detection, permanent abnormal series is sequentially recorded in files by date and by type in the internal memory, USB flash memory, the internal hard disk recording medium of the external computer Way.

Images