Diagnostic system and diagnostic method

Abstract

A diagnostic system and a diagnostic method are provided. An example diagnostic system determines a state of a target device including a working device. The diagnostic system includes an acquisition unit configured to acquire first data generated in response to the working device operating the target device at a first pressure and second data generated in response to the working device operating the target device at a second pressure, a calculation unit configured to calculate a characteristic quantity indicating a relationship between the first data and the second data, and a determination unit configured to determine the state of the target device based on the characteristic quantity.

Description

Technical Field

[0001] One aspect of this disclosure relates to a diagnostic system and a diagnostic method. Background Technology

[0002] Japanese Patent Application No. 5638102B describes a spot welding system having a spot welding gun for welding workpieces while applying pressure to a workpiece between a movable electrode chip driven by a servo motor and a fixed electrode chip facing the movable electrode chip. Summary of the Invention

[0003] In one aspect of this disclosure, it is desirable to appropriately determine the state of the equipment used for processing workpieces.

[0004] A diagnostic system according to one aspect of this disclosure determines the state of a target device including a working device. The diagnostic system includes: an acquisition unit configured to acquire first data generated in response to operating the working device of the target device under a first pressure and second data generated in response to operating the working device under a second pressure; a calculation unit configured to calculate a characteristic quantity indicating the relationship between the first data and the second data; and a determination unit configured to determine the state of the target device based on the characteristic quantity.

[0005] A diagnostic method according to one aspect of this disclosure includes: acquiring first data generated in response to operating the working device of the target device under a first pressure and second data generated in response to operating the working device under a second pressure; calculating a characteristic quantity indicating the relationship between the first data and the second data; and determining the state of the target device based on the characteristic quantity.

[0006] A diagnostic procedure includes instructions that, when executed by a computer, cause the computer to: acquire first data generated in response to operation of a working device of a target device under a first pressure and second data generated in response to operation of the working device under a second pressure; calculate a characteristic quantity indicating the relationship between the first data and the second data; and determine the state of the target device based on the characteristic quantity.

[0007] According to one aspect of this disclosure, the state of the equipment used for processing workpieces can be appropriately determined. Attached Figure Description

[0008] Figure 1 This is a diagram illustrating the configuration of an example robot system.

[0009] Figure 2 This is a diagram illustrating an example of determining the hardware configuration of a device.

[0010] Figure 3 This is a diagram illustrating an example of defining the functional configuration of a device.

[0011] Figure 4 This is a diagram illustrating an example configuration for collecting response data.

[0012] Figure 5 This is a flowchart illustrating an example process for collecting response data.

[0013] Figure 6 This is a diagram showing an example of command and response signals that change over time.

[0014] Figure 7 This is a flowchart illustrating an example process for determining the state of a target device.

[0015] Figure 8 This is a graph showing an example of historical data showing how features change over time.

[0016] Figure 9 This is a graph showing example features in a multivariate normal distribution. Detailed Implementation

[0017] In the following description, with reference to the accompanying drawings, the same reference numerals are assigned to the same components or similar components having the same function, and repeated descriptions are omitted.

[0018] Robotic systems

[0019] refer to Figure 1 The example diagnostic system (also referred to herein as the "determining system") according to this disclosure can be applied to the determining device 4 of the robot system 1. The robot system 1 can automate various operations, such as processing and assembly, by causing the robot to perform actions instructed by an operator.

[0020] Figure 1 This is a diagram illustrating an example configuration of robot system 1. According to some examples, robot system 1 includes one or more robots 2, one or more robot controllers 3 corresponding to the one or more robots 2, and a determining device 4. Figure 1 The diagram illustrates a configuration where one robot 2 is connected to a robot controller 3. In other examples, multiple robots 2 may be connected to a single robot controller 3.

[0021] In some examples, robot 2 is a multi-axis serially connected vertical articulated robot, configured to perform various processes using a working device 5 held at its tip. Robot 2 is able to freely change the position and orientation of its tip within a predetermined range. Robot 2 can be a 6-axis vertical articulated robot or a 7-axis vertical articulated robot, with one redundant axis added to the 6 axes. In some examples, multiple robots 2 are arranged such that any one of them can perform the same process on the same workpiece arranged in the same position.

[0022] Robot 2 is an example of a device including a working device 5 and a motor 5a for actuating the working device 5. Such a device may also be referred to as the target device in this disclosure. The working device is a device for processing a workpiece, for example, a device that causes some physical change to the workpiece. The working device 5 can be various devices such as a welding torch, a press, etc. The motor 5a can be a component of the working device 5 or exist outside the working device 5. In some examples, the working device 5 is a welding torch mounted to robot 2 as an end executer. The welding torch uses electricity provided by a spot welding machine 6 to spot weld the workpiece.

[0023] Robot controller 3 is a device that controls robot 2 according to a pre-generated operating program. The operating program includes j data for controlling robot 2. Under the control of robot controller 3, robot 2 performs a series of processes. In this disclosure, this series of processes may also be referred to as a job. The smallest unit of processing that constitutes a job is called a task. Therefore, a job contains one or more tasks. Robot 2 can perform various tasks, such as "removing components", "placing components", "assembling components (to workpieces)", "adopting a waiting posture", etc. The operating program indicates, for example, the execution sequence of tasks and the path representing the trajectory of robot 2. The trajectory of robot 2 refers to the movement path of robot 2 or its components. For example, the trajectory of robot 2 may be the trajectory of a tip portion. In some examples, robot controller 3 calculates target joint angle values ​​(target angle values ​​for each joint of robot 2) to match the position and posture of the tip portion with the target values ​​indicated by the operating program and controls robot 2 according to the target angle values.

[0024] The determining device 4 is a computer that determines the state of the robot 2, which is the target device. In some examples, the determining device 4 determines the state of the robot 2 based on data obtained from the robot 2, the working device 5, or the motor 5a. The determining device 4 outputs a command to the motor 5a to start the working device 5. This command is an instruction for controlling the working device 5. The determining device 4 collects responses from the motor 5a, the working device 5, or the robot 2 that are started according to the command. The response is an output corresponding to the command and, for example, indicates the movement or state of an object actuated according to the command. The object can be, for example, at least one of the working device 5, the motor 5a, and the robot 2. In this disclosure, data indicating a set of collected responses may also be referred to as response data. The determining device 4 then determines the state of the robot 2 based on the response data.

[0025] In this disclosure, the state of the equipment can refer to the condition or status of the equipment itself or its components, for example, indicating whether the equipment or components are normal or abnormal. The state of the equipment itself can, for example, correspond to the state of robot 2. The state of the components can, for example, be the state of working device 5 or motor 5a. In the case where working device 5 is a welding torch, examples of welding torch abnormalities include wear of electrodes (electrode chips), axial misalignment between a pair of electrodes, deterioration of mechanical properties (such as gaps), and abnormalities of circuit elements such as converter ICs. The cause of equipment or component abnormalities may be aging, deterioration, or defective products.

[0026] Determine equipment

[0027] Figure 2 This is a diagram illustrating an example of the hardware configuration of determining device 4. Determining device 4 includes a main body 10, a monitor 20, and an input device 30.

[0028] The main body 10 may include at least one computer. The main body 10 houses circuitry 160, and circuitry 160 may include at least one of a processor 161, a memory 162, a storage device 163, and an input / output port 164. The storage device 163 stores programs in the form of processor-executable instructions, for example, for configuring various functional modules of the main body 10. The storage device 163 is a computer-readable recording medium, such as a hard disk, non-volatile semiconductor memory, magnetic disk, optical disk, etc. The memory 162 temporarily stores programs loaded from the storage device 163, calculation results from the processor 161, etc. The processor 161 provides each functional module by cooperating with the memory 162 to execute programs. The input / output port 164 inputs and outputs electrical signals between the monitor 20, the input device 30, and the robot controller 3 in response to commands from the processor 161.

[0029] Monitor 20 is a device for displaying information output from body 10. Monitor 20 may include any display device capable of displaying graphics, and examples include a liquid crystal panel. Input device 30 is a device for inputting information into body 10. Input device 30 may be any one or more devices capable of inputting desired information, and may include, for example, a keyboard and a mouse.

[0030] The monitor 20 and input device 30 can be integrated into a touch panel (e.g., a touchscreen). For example, similar to a tablet computer, the main body 10, monitor 20, and input device 30 can be integrated into a single device.

[0031] Figure 3An example of the functional configuration of the determining device 4 is shown. In some examples, the determining device 4 includes an instruction unit 101, an acquisition unit 102, a database 103, an acquisition unit 104, a calculation unit 105, and a determining unit 106 as functional modules. The instruction unit 101 is a functional module that outputs a command to the motor 5a to start the working device 5. The term "outputting a command to the motor" means that the output command is ultimately transmitted to the motor. The acquisition unit 102 is a functional module that acquires the response to the command. The database 103 is a functional module that stores the response obtained by the acquisition unit 102 as response data. The database 103 can be implemented on a computer separate from the determining device 4 or on a computer system separate from the robot system 1, so that the determining device 4 can access the database 103. The acquisition unit 104 is a functional module that acquires response data from the database 103. The calculation unit 105 is a functional module that calculates feature quantities based on the response data. In this disclosure, a feature quantity refers to a numerical value representing a characteristic of the motion of the target device, such as a numerical value representing a characteristic of the motion of at least one of the robot 2, the working device 5, and the motor 5a. The determination unit 106 is a functional module that determines the state of robot 2 based on feature quantities.

[0032] exist Figure 3 In the example, instruction unit 101 sends instructions to motor 5a via robot controller 3, and acquisition unit 102 acquires responses via robot controller 3. For example, robot controller 3 controls motor 5a based on commands, and motor 5a starts working device 5 under control. Robot controller 3 then receives responses to commands from at least one of robot 2, working device 5, and motor 5a, and sends the responses to determining device 4. Determining device 4 can send commands and acquire responses in other ways. For example, instruction unit 101 can send commands directly to motor 5a. Acquisition unit 102 can acquire responses directly from at least one of robot 2, working device 5, and motor 5a.

[0033] Figure 4 This diagram illustrates an example of collecting response data from a robot 2. In this example, instruction unit 101 repeats a combination of actual operations (also referred to herein as "work operations") and inspection operations, and acquisition unit 102 stores response records in database 103, indicating the response at each inspection operation (i.e., each inspection time point). As a result, database 103 stores multiple response records corresponding to multiple inspection time points as response data indicating a response history that changes over time.

[0034] In this disclosure, actual operation or work operation refers to the operation by which the target equipment processes one or more workpieces. In association with actual operation, the task corresponding to a process for a single workpiece is referred to as actual operation (also referred to herein as "work task"). For example, in actual operation, robot 2 uses a welding torch to perform one or more spot welding operations on a workpiece. In this actual operation, the welding torch is powered by a power supply provided by spot welding machine 6, and the welding torch performs spot welding through the heat generated by the current.

[0035] In this disclosure, an inspection operation refers to the operation of starting the target equipment to obtain a response record. An inspection operation is an operation used to start the target equipment without processing a workpiece. In the case of an inspection operation, performing the motion corresponding to the actual operation without processing a workpiece is called a no-work operation. When the work unit 5 performs a no-work operation, no workpiece is provided to the work unit 5. Through the inspection operation (no-work operation), a response record reflecting the state that interference factors (external factors) related to the workpiece have been eliminated or significantly reduced is obtained. Therefore, this response record indicates the state of the target equipment itself or the state of the components of the target equipment itself.

[0036] In some examples, instruction unit 101 causes robot 2 to perform multiple actual tasks in one actual operation, and then causes robot 2 to perform an empty task in an inspection operation. When the working device 5 is a welding torch, in each inspection task, instruction unit 101 actuates robot 2 with the welding torch de-energized. For example, instruction unit 101 actuates robot 2 with the welding torch de-energized and the electrodes of the welding torch in contact but not pressed against each other. Acquisition unit 102 stores the response records obtained from the empty task in database 103. Figure 4 As shown, by repeatedly determining that device 4 performs an inspection operation after an actual operation, multiple response records (e.g., a large number of response records) for a robot 2 are accumulated in database 103.

[0037] Diagnostic methods

[0038] As an example of a diagnostic method (also referred to herein as a "determining method") according to this disclosure, reference will be made to... Figure 5 The description determines the acquisition of response data performed by device 4. Figure 5 This is a flowchart illustrating the acquisition process as processing flow S1. For example, it is determined that device 4 executes processing flow S1. In some examples, it is determined that device 4 executes processing flow S1 in each inspection operation for each robot 2. Therefore, robot 2 and working device 5 do not process workpieces in processing flow S1.

[0039] In step S11, the instruction unit 101 outputs a first command to the motor 5a to actuate the working device 5 with a first pressure. The first command is, for example, a torque command. In some examples, the first pressure is the pressure required to process the workpiece using the working device 5. For example, the first pressure is the pressure applied to a pair of electrodes on the welding torch. In some examples, the instruction unit 101 sends the first command to the robot controller 3, and the robot controller 3 controls the motor 5a based on the first command. The motor 5a actuates the welding torch with the first pressure according to the control, and the first pressure is applied between the pair of electrodes. During this process, the instruction unit 101 outputs the first command when the welding torch is not energized. For example, the instruction unit 101 outputs the first command when the welding torch is not energized and the electrodes of the welding torch are in contact but not pressing against each other.

[0040] In step S12, the acquisition unit 102 acquires a first response to the first command. The first response is information obtained in response to the output of the first command to the motor 5a, and is typically acquired when the working device 5 and the motor 5a are in a stable state. The first response may be information obtained from at least one of the robot 2, the working device 5, and the motor 5a. In some examples, the first response indicates the position of the motor 5a along the gun axis of the welding torch, and this position may reflect the state of the welding torch. The gun axis is represented by a straight line extending between the tips of a pair of electrodes.

[0041] In step S13, the instruction unit 101 outputs a second command to the motor 5a to actuate the working device 5 with a second pressure. For example, the second command is a torque command. Like the first pressure, the second pressure is also the pressure required to process the workpiece using the working device 5, such as the pressure applied between a pair of electrodes of the welding torch. The second pressure is different from and, for example, higher than the first pressure. In some examples, the instruction unit 101 sends the second command to the robot controller 3, and the robot controller 3 controls the motor 5a based on the second command. The motor 5a operates the welding torch with the second pressure according to the control, and the second pressure is applied between a pair of electrodes. During this process, the instruction unit 101 outputs the second command when the welding torch is not energized. For example, the instruction unit 101 outputs the second command when the welding torch is not energized and the electrodes of the welding torch are in contact but not pressing against each other.

[0042] In step S14, the acquisition unit 102 acquires a second response to the second command. The second response is information obtained in response to the output of the second command to the motor 5a, and is typically acquired when the working device 5 and the motor 5a are in a stable state. Similar to the first response, the second response can be information obtained from at least one of the robot 2, the working device 5, and the motor 5a. For example, the second response indicates the position of the motor 5a along the axis of the welding torch.

[0043] In step S15, the acquisition unit 102 stores the response record indicating the first response and the second response in the database 103. In some examples, the acquisition unit 102 sets a record identifier to uniquely identify a pair of first and second responses, generates a response record indicating the combination of the record identifier, the first response, and the second response, and stores the response record in the database 103. The data item set as the identifier can be at least one of the following identifiers: serial number, inspection date and time, robot identifier 2. In some examples, the response record may include a data item indicating the determination result based on the response indicated by the record. When storing the response record in step S15, the determination result is empty.

[0044] Figure 6 This is a graph illustrating an example of commands and responses in an inspection operation (i.e., in one execution of process flow S1). The horizontal axis of the graph represents elapsed time, and the vertical axis represents the pressure applied to the welding torch (working device 5) and the position of the motor 5a along the torch axis. In this example, command unit 101 outputs a first command to motor 5a to actuate the welding torch with a first pressure P1. Acquisition unit 102 acquires the position L1 of motor 5a in a steady state as a first response to the first command. Command unit 101 then outputs a second command to motor 5a to actuate the welding torch with a second pressure P2. Acquisition unit 102 acquires the position L2 of motor 5a in a steady state as a second response to the second command. As shown in this example, command unit 101 can output a second command to motor 5a, causing the pressure of working device 5 to increase from the first pressure to the second pressure without decreasing the pressure of working device 5. By controlling the pressure in this way, both pressure and position change gradually. In some examples, the first pressure P1 is 3000 N and the second pressure P2 is 4000 N. The instruction unit 101 can output the first command and the second command and collect the first response and the second response without changing the attitude of the target device. Figure 6 This illustrates a scenario where the pressure is further increased from the second pressure P2, and the position of the motor 5a changes further in response to the pressure change. However, depending on the example, this further pressure increase is not necessary.

[0045] One execution of process flow Sl corresponds to one empty job. Each empty job is executed after an actual operation (or work operation). Figure 4 and Figure 5 As can be seen, the instruction unit 101 repeatedly performs the actual operation of the robot 2 processing one or more workpieces and then outputs a first command and a second command to the motor 5a after the actual operation. As a result of this iteration, the database 103 stores multiple response records for one robot 2.

[0046] As an example of a diagnostic method (or determination method) according to this disclosure, reference will be made to... Figure 7 The description refers to the determination of the state of the target device performed by the determining device 4. Figure 7 This is a flowchart illustrating the determination process as processing flow S2. That is, the determination device 4 executes processing flow S2. In some examples, the determination device 4 executes processing flow S2 for each robot 2. For example, the determination device 4 executes processing flow S2 for robot 2 after sufficient or necessary response data has been accumulated in database 103.

[0047] In step S21, the acquisition unit 104 sets one or more determination parameters. The determination parameters are data used to determine the state of a target device, such as robot 2. For example, the determination parameters may include a threshold for determination or the size of a reference window, described later. The acquisition unit 104 may set the determination parameters based on user input, or it may use determination parameters pre-stored in memory 163.

[0048] In step S22, the acquisition unit 104 retrieves response data corresponding to a robot 2 from the database 103. The acquisition unit 104 can read all response data or a portion of the response data corresponding to a time width for the robot 2. In any case, the response data includes first data indicating a first response and second data indicating a second response. That is, the acquisition unit 104 acquires first data in response to outputting a first command to motor 5a to start the working device 5 at a first pressure, and second data in response to outputting a second command to motor 5a to start the working device 5 at a second pressure. The response data displays the history of the response along a time sequence. Therefore, the acquisition unit 104 acquires time-series data of the working device 5 operating under the first pressure as first data, and acquires time-series data of the working device 5 operating under the second pressure as second data. (Refer to the above...) Figure 4 and Figure 5 The first and second data described herein are obtained through a combination of repeated actual operations and inspection operations.

[0049] In step S23, the calculation unit 105 calculates a feature quantity that indicates the relationship between the first data and the second data, and correspondingly indicates the relationship between the first response and the second response, based on the response data. For each of the plurality of inspection operations (i.e., the plurality of inspection time points), the calculation unit 105 calculates at least one of the difference between the first response and the second response (i.e., the difference between the first data and the second data) and the ratio of the first response to the second response (i.e., the ratio of the first data and the second data) as a feature quantity. The calculation unit 105 generates feature history data representing the change of the feature quantity over time. For the calculation of the ratio, the calculation unit 105 may calculate the ratio of the second data to the first data or the ratio of the first data to the second data.

[0050] In step S24, the determining unit 106 sets a reference window for the feature history data. In this disclosure, the reference window refers to a range set for the feature history data to determine the state of the target device at a given time point. The size of the reference window is defined, for example, by a determining parameter. In this disclosure, the time point for determining the state of the target device can also be called the target time point. The determining unit 106 sets a reference window for the feature history data that includes multiple time points prior to the target time point. The reference window does not include the target time point.

[0051] Figure 8 An example of feature history data and reference windows is shown. In this example, graph 210 indicates feature history data 211, which shows the change of the difference between the first and second data over time. Graph 220 is feature history data 221, which shows the change of the ratio between the first and second data over time. In this example, the determining unit 106 sets reference window 212 for feature history data 211 and reference window 222 for feature history data 221. The determining unit 106 sets the position and width of reference windows 212 and 222 to be the same. That is, when setting reference windows for both the difference and the ratio, the determining unit 106 matches the position and width of the two reference windows. Figure 8 In the example, since the trend of the ratio changed after 700 seconds, there is a high probability that the anomaly started from that point in time.

[0052] Back Figure 7 In step S25, determining unit 106 calculates the statistical values ​​of the feature quantities within the reference window. For example, when two corresponding feature quantities are calculated, such as a combination of difference and ratio, determining unit 106 calculates the mean (mean vector) and covariance (covariance matrix) of these two feature quantities as statistical values. The mean and covariance represent a two-dimensional multivariate normal distribution. Alternatively, determining unit 106 may calculate statistical values ​​such as mean, variance, etc., for each of the difference and ratio.

[0053] In step S26, the determining unit 106 determines the state of robot 2 (target device) at the target time point based on the feature quantity at the target time point and the calculated statistical value. In some examples, the determining unit 106 compares the feature quantity at the target time point with a given threshold set for the statistical value to determine the state of robot 2. This threshold is defined, for example, by a determining parameter. For example, the determining unit 106 calculates the Mahalanobis distance between the feature quantity at the target time point and the distribution of the feature quantity within a reference window based on the feature quantity and the statistical value at the target time point, and determines the state of robot 2 based on the Mahalanobis distance. If the Mahalanobis distance is equal to or less than the threshold, the determining unit 106 determines that robot 2 is normal; if the Mahalanobis distance exceeds the threshold, the determining unit 106 determines that robot 2 is abnormal. The determining unit 106 may perform the determination based on the Mahalanobis distance of two variables (two-dimensional) corresponding to the combination of difference and ratio, or based on the Mahalanobis distance of a single variable (one-dimensional) corresponding to each of the difference and ratio. In the case of determining based on each of the difference and ratio, if both Mahalanobis distances are less than or equal to the threshold, the determining unit 106 can determine that robot 2 is normal, and if at least one Mahalanobis distance exceeds the threshold, the determining unit 106 determines that robot 2 is abnormal.

[0054] Figure 9 This is a diagram illustrating an example of determining the state of robot 2, and the multivariate normal distribution is represented by the mean and covariance of the combination of differences and ratios using contours. Point 301 represents the individual feature quantities within the reference window, and box Th represents the threshold for Mahalanobis distance. In this example, determination unit 106 calculates the Mahalanobis distance, which is the distance between the feature quantity at the target time point and the distribution of the feature quantity within the reference window, and compares the Mahalanobis distance with the threshold (box Th). The threshold for Mahalanobis distance can be defined by a multiple of the standard deviation σ, such as 2σ, 3σ, or 4σ. The standard deviation σ represents the variation in the data distribution observed in multiple inspection operations (empty jobs) under the assumption that the distribution is normally distributed. If the calculated Mahalanobis distance is equal to or less than the threshold, determination unit 106 determines that robot 2 is normal; if the calculated Mahalanobis distance exceeds the threshold, robot 2 is determined to be abnormal. In the case where the feature quantity at the target time point is represented by point 311, determination unit 106 determines that robot 2 is normal because the Mahalanobis distance is less than or equal to the threshold. When the feature quantity at the target time point is represented by point 312, the determination unit 106 determines that robot 2 is abnormal because the Mahalanobis distance exceeds the threshold. Figure 9 An example is shown that the state of a target device is determined based on both the difference and the ratio.

[0055] Return to reference Figure 7In step S27, the determination unit 106 outputs the determination result. For example, the determination unit 106 may store the determination result in a recording medium such as storage device 163. Alternatively, the determination unit 106 may display the determination result on monitor 20 in the form of text, moving images, or still images using computer graphics (CG). In some examples, the determination unit 106 writes the determination result into a response record corresponding to the target time point. For example, the determination unit 106 may write a flag indicating whether robot 2 is normal or abnormal into the response record.

[0056] As shown in step S28, determining device 4 repeats steps S24 to S27 until all target time points have been processed. In step S24, determining unit 106 moves the position of the reference window by a predetermined amount along the time flow to reset the reference window. For example, determining unit 106 shifts the reference window by one time point. Figure 8 In the diagram, arrows extending from the right edges of reference windows 212 and 222 indicate the direction of reference window movement. In step S25, the determining unit 106 calculates the statistical values ​​of the feature quantities within the reset reference window. In step S26, the determining unit 106 determines the state of robot 2 at the next target time point based on the feature quantities and the calculated statistical values. In step S27, the determining unit 106 outputs the determination result. In other words, the determining unit 106 can determine the state of robot 2 at each of multiple target time points while simultaneously moving the reference window along the time axis.

[0057] In cases where the determination is repeated while simply moving the reference window along the time axis, the reference window may include characteristic quantities of the target time point at which robot 2 is determined to be anomalous. In some examples, when the determination unit 106 sets the reference window to determine the state of robot 2 at a second target time point after multiple first target time points, the determination unit 106 sets the reference window such that the first target time point at which robot 2 is determined to be anomalous is excluded from the reference window.

[0058] This section describes an example of setting up such a reference window. In the example, it is assumed that the feature history data is represented with one hundred time points T1 to T2. 100 The corresponding feature quantities are defined, and the width of the reference window is ten time points. Assume robot 2 operates from time point T1 to T... 10 If normal, then unit 106 will set the reference window to time points T1 to T2. 10 Within the range, to determine the robot 2 at time point T 11 The state of robot 2 at time point T. 11 Under normal circumstances, the determination unit 106 will shift the reference window to time points T2 to T3. 11 The range is used to determine the robot 2 at time point T. 12The state of robot 2 at time T. In this example, assume robot 2 is in this state at time T. 50 Previously determined to be normal and at time point T 51 The first time it was identified as an anomaly. In this case, the determining unit 106 sets the reference window to time point T. 41 To T 50 The range, not the point in time T. 42 To T 51 The range, in order to determine the position of robot 2 at time point T. 52 The state of the robot. Therefore, when the reference window is set to determine the robot 2 at the second target time point T. 52 When the state is such that the reference window is set such that the first target time point T is T 51 Excluded from the reference window. This is to determine robot 2 at time point T. 52 Under normal circumstances, the determination unit 106 sets the reference window to time point T. 42 To T 50 and T 52 The range, in order to determine the position of robot 2 at time point T. 53 The state of robot 2 at time T. 53 If this is also determined to be normal, then the determining unit 106 will set the reference window to time point T. 43 To T 50 T 52 and T 53 The range, in order to determine the position of robot 2 at time point T. 54 The state of robot 2 at time point T. 54 In abnormal situations, the determination unit 106 sets the reference window to time point T. 43 To T 50 T 52 and T 53 The range, in order to determine the position of robot 2 at time point T. 55 The state. As described above, the determining unit 106 can set a reference window while excluding the first target time point for determining the anomaly of the robot 2, so that the size of the reference window, i.e., the number of feature quantities included in the reference window, remains constant.

[0059] program

[0060] The functional modules of device 4 are determined by loading a diagnostic program (also referred to herein as the "determining program") onto processor 161 or memory 162 and having processor 161 execute the program. The diagnostic program includes code (e.g., processor-executable instructions) for implementing the various functional modules of device 4. Processor 161 operates input / output port 164 according to the diagnostic program and reads and writes data in memory 162 or storage device 163. The various functional modules of device 4 are determined through this process.

[0061] Diagnostic procedures can be provided after being permanently recorded on a non-transitory recording medium such as a CD-ROM, DVD-ROM, or semiconductor memory. In other examples, diagnostic procedures can be provided via a communication network as data signals superimposed on the carrier.

[0062] As described above, a determination system or diagnostic system according to some examples of this disclosure determines the state of a target device including a working device and a motor for operating the working device. The determination system includes: an acquisition unit configured to acquire first data generated in response to operating the working device of the target device under a first pressure, and second data generated in response to operating the working device under a second pressure; a calculation unit configured to calculate a characteristic quantity indicating the relationship between the first data and the second data; and a determination unit configured to determine the state of the target device based on the characteristic quantity.

[0063] The method for determining aspects of this disclosure includes: acquiring first data generated in response to operating the working device of the target device under a first pressure; acquiring second data generated in response to operating the working device under a second pressure; calculating a characteristic quantity indicating the relationship between the first data and the second data; and determining the state of the target device based on the characteristic quantity.

[0064] The determination procedure includes instructions that, when executed by a computer, cause the computer to: acquire first data generated in response to the operation of the working device of the target device under a first pressure; acquire second data generated in response to the operation of the working device under a second pressure; calculate a characteristic quantity indicating the relationship between the first data and the second data; and determine the state of the target device based on the characteristic quantity.

[0065] Based on this example, the state of the equipment is determined according to the relationship between two types of data obtained by actuating the working device under each of two pressures. Because considering the relationship between the two types of data eliminates or significantly reduces interfering factors, the state of the equipment can be determined more appropriately. When the state of the equipment is determined by a single type of data, i.e., by absolute values, it may be more difficult to accurately determine the state of the equipment, because absolute values ​​can vary greatly due to interfering factors (e.g., the environment surrounding the equipment) at the measurement time point. Since the characteristic quantity indicating the relationship between the first and second data is a relative value, the influence of interfering factors is essentially removed. Therefore, by using characteristic quantities, the state of the equipment can be determined in a more stable and / or more accurate manner. In some examples, it can be determined automatically without relying on the operator's knowledge and experience, and early detection of equipment anomalies and identification of defective individual components can be performed more effectively.

[0066] In some examples, the features may include the difference between the first and second data points, as well as the ratio between the first and second data points, thereby determining the device's state based on the difference and ratio. Therefore, by considering two types of features, such as difference and ratio, the device's state can be determined more reliably, thus enabling more reliable detection of any device anomalies (abnormal states).

[0067] In some examples, first time-series data of the working device operating under a first pressure can be acquired as first data, and second time-series data of the working device operating under a second pressure can be acquired as second data. The calculation unit can be further configured to generate feature history data based on the first and second time-series data. The feature history data can indicate the changes of feature quantities over time. The determination unit can also be configured to: set a reference window in the feature history data that includes multiple time points prior to the target time point; calculate statistical values ​​of feature quantities associated with the multiple time points within the reference window; and determine the state of the target device based on the feature quantities at the target time point and based on the statistical values. As the first and second data change over time, the feature history data changes accordingly. By calculating statistical values ​​of feature quantities over a predetermined time width in the past, and considering these statistical values ​​and the feature quantities at the target time point, the state of the device at the target time point can be determined more accurately, even if the data changes.

[0068] In some examples, the calculated statistics may include the mean and covariance of features within a reference window. By considering the mean and covariance, the state of the device can be determined more accurately.

[0069] In some examples, the determination unit is also configured to: calculate the Mahalanobis distance between the feature quantity at the target time point and the distribution of the feature quantity within the reference window based on the feature quantity at the target time point and based on statistical values; determine that the target device is normal when the Mahalanobis distance is equal to or less than a given threshold; and determine that the device is abnormal when the Mahalanobis distance exceeds the threshold.

[0070] In some examples, the determining unit is also configured to: determine the state of the target device at each of the multiple target time points by shifting a reference window along the time axis of the feature history data for each of the multiple target time points; and for each of the multiple target time points, calculate a statistical value by excluding any feature quantities in the corresponding reference window that are associated with an abnormal state of the target device. For example, the determining unit can be configured to: determine the state of the target device at each of the multiple first target time points while moving the reference window along the time axis; and set a reference window to determine the state of the device at a second target time point after the first target time point, such that any first target time point where the device is determined to be abnormal is excluded from the reference window. Since the reference window is set by data determined to be normal, and not by data determined to be abnormal, the statistical values ​​obtained from the reference window also continue to represent the normal movement of the device. Therefore, the state of the device can be determined more accurately along the time axis.

[0071] In some examples, the determination system may also include an instruction unit configured to output a first command to cause the motor of the target device to operate the working device at a first pressure and a second command to cause the motor to operate the working device at a second pressure. This configuration enables the determination of the device state and allows the instructions used for determination to be executed more comprehensively.

[0072] In some examples, first and second commands can be output to the motor without changing the orientation of the target device.

[0073] In some examples, a second command can be output to the motor to increase the pressure of the operating device from a first pressure to a second pressure, without decreasing the pressure below the first pressure. Because such pressure control avoids repeatedly applying the first pressure during the change from the first pressure to the second pressure, the characteristic quantity representing the relationship between the first and second data can be calculated more accurately.

[0074] In some examples, the working device can be a welding torch, and first and second commands can be output to the motor even when the welding torch is not powered. This process essentially eliminates interference factors associated with the welding torch, such as the weld nugget on the welded portion of the workpiece, thereby allowing for a more appropriate determination of the equipment's status, for example, in a more stable or accurate manner.

[0075] In some examples, while the first and second commands are being output to the motor, the pair of electrodes of the welding torch can be brought into contact with each other without any contact pressure being applied. During the operation of the welding torch, performing the process of not energizing the torch and bringing the electrodes into contact without applying pressure is to determine the change in distance between the pair of electrodes (in other words, to measure electrode wear). In this case, taking advantage of this necessary process of the welding torch to collect the first and second data improves the efficiency of the welding torch operation (e.g., inspection).

[0076] In some examples, the device can be a robot, and the welding torch can be mounted on the robot as an end effector. The welding torch mounted on the robot can perform various postures corresponding to the robot's movements. In this case, the state of the device can be more appropriately determined by essentially eliminating the interference factors caused by posture changes.

[0077] In some examples, the instruction unit can also be configured to repeat a sequence of operations multiple times, the sequence including performing an actual operation to cause the target device to process one or more workpieces, and subsequently performing an inspection operation, the inspection operation including outputting a first instruction and a second instruction to a motor, wherein the resulting first data and second data are stored in a storage device, so that the first data and second data can be retrieved from the storage device. For example, the instruction unit can be configured to repeat multiple times a combination of an actual operation (or working operation) of processing one or more workpieces by the device and an inspection operation of outputting a first command and a second command to the motor after the actual operation, acquiring first data and second data obtained through repetition. By acquiring the first data and second data during actual operation, it is more appropriate to acquire changes in the device state due to continuous use of the device as the first data and second data.

[0078] Additional examples

[0079] It should be understood that not all aspects, advantages, and features described herein must be implemented by or included in any particular example. In fact, various examples have been described and illustrated herein, and it is obvious that other examples can be modified in their arrangement and details can be omitted.

[0080] For example, the functional configuration of the diagnostic system (or determination system) is not limited to the examples above. A different functional configuration than the examples above can be used to perform the diagnostic method (or determination method) according to this disclosure.

[0081] The hardware configuration of the diagnostic system is not limited to the example of implementing each functional module by executing a program. In some examples, at least a portion of the aforementioned functional modules may be configured by logic circuitry dedicated to performing that function, or by application-specific integrated circuits (ASICs) in which that logic circuitry is integrated.

[0082] The process of a method executed by at least one processor is not limited to the examples above. For example, some of the steps or processes described above may be omitted or executed in a different order. Furthermore, two or more of the steps described above may be combined, or some steps may be modified or deleted. Alternatively, other steps may be performed in addition to those described above.

[0083] When comparing the size relationship between two values ​​in a computer system or computer, either the "greater than or equal to" or "greater than" standard can be used, and either the "less than or equal to" or "less than" standard can be used.

[0084] All modifications and variations falling within the spirit and scope of the subject matter claimed herein are protected.

[0085] The following appendix provides further illustrative information regarding the examples above.

[0086] (Appendix 1) A determining system for determining the state of a device including a working device and a motor for actuating the working device, said determining system comprising:

[0087] The acquisition unit is configured to acquire first data in response to a first command to output to a motor actuating the working device with a first pressure, and second data in response to a second command to output to a motor actuating the working device with a second pressure.

[0088] The calculation unit is configured to calculate a feature quantity indicating the relationship between the first data and the second data; and

[0089] The determination unit is configured to determine the state of the device based on characteristic quantities.

[0090] (Appendix 2) Determination system according to Appendix 1

[0091] The calculation unit is further configured to calculate the difference between the first data and the second data, and the ratio of the first data to the second data, as feature quantities; and

[0092] The determining unit is also configured to determine the state of the device based on both the difference and the ratio.

[0093] (Appendix 3) Determination system according to Appendix 1 or 2

[0094] The acquisition unit is also configured as follows:

[0095] The time-series data of the working device operating under the first pressure is obtained as the first data;

[0096] as well as

[0097] The time-series data of the working device operating under the second pressure are acquired as the second data.

[0098] The computing unit is also configured to generate historical feature data indicating how the feature quantity changes over time, and

[0099] The determining unit is further configured as follows:

[0100] To determine the status of the device at the target time point, a reference window is set for the feature historical data, including multiple time points prior to the target time point.

[0101] Calculate the statistical values ​​of the features within the reference window; and

[0102] The status of the equipment is determined based on the characteristic quantities and statistical values ​​at the target time point.

[0103] (Appendix 4) The determination system according to Appendix 3, wherein the determination unit is further configured to calculate the mean and covariance of the feature quantities within the reference window as statistical values.

[0104] (Appendix 5) The determining system according to Appendix 3 or 4, wherein the determining unit is further configured as follows:

[0105] While moving the reference window along the time axis, determine the state of the device at each of multiple first target time points; and

[0106] When setting a reference window to determine the status of the device at a second target time point after the first target time point, the reference window is set so that the first target time point at which the device is determined to be abnormal is excluded from the reference window.

[0107] (Appendix 6) The determining system according to any one of Appendices 1 to 5 further includes an instruction unit configured to output a first command and a second command to the motor.

[0108] (Appendix 7) The determining system according to Appendix 6, wherein the instruction unit is further configured to output a second command to the motor to increase the pressure of the working device from the first pressure to the second pressure without decreasing the pressure of the working device from the first pressure.

[0109] (Appendix 8) Determination system according to Appendix 6 or 7

[0110] The working device is a welding torch; and

[0111] The instruction unit is also configured to output a first command and a second command to the motor when the welding torch is not powered on.

[0112] (Appendix 9) According to the determination system described in Appendix 8, the instruction unit is further configured to output a first command and a second command to the motor when the welding torch is not energized and the electrodes of the welding torch are in contact but not pressing against each other.

[0113] (Appendix 10) The determination system according to Appendix 8 or 9

[0114] The equipment includes robots, and

[0115] The welding torch is installed on the robot as an end effector.

[0116] (Appendix 11) The determination system according to any one of Appendices 6 to 10,

[0117] The instruction unit is also configured to repeatedly perform a combination of actual operations involving the processing of one or more workpieces by the equipment and a check operation involving outputting a first and a second command to the motor after the actual operations.

[0118] The acquisition unit is also configured to acquire first and second data obtained through repeated acquisition.

[0119] (Appendix 12) A determination method performed by a determination system for determining the state of a device including a working device and a motor for actuating the working device, the method comprising:

[0120] Acquire first data in response to a first command to output to a motor actuating the working device with a first pressure, and second data in response to a second command to output to a motor actuating the working device with a second pressure;

[0121] Calculate the characteristic quantity indicating the relationship between the first data and the second data; and

[0122] The state of the device is determined based on characteristic quantities.

[0123] (Appendix 13) A determining procedure that enables a computer, as a determining system, to determine the state of a device including a working device and a motor for actuating the working device, the procedure causing the computer to execute:

[0124] Acquire first data in response to a first command to output to a motor actuating the working device with a first pressure, and second data in response to a second command to output to a motor actuating the working device with a second pressure;

[0125] Calculate the characteristic quantity indicating the relationship between the first data and the second data; and

[0126] The state of the device is determined based on characteristic quantities.

Claims

1. A diagnostic system, comprising: The acquisition unit is configured to acquire first data generated in response to the operation of the target device under a first pressure and second data generated in response to the operation of the operation device under a second pressure; The calculation unit is configured to calculate a feature quantity that indicates the relationship between the first data and the second data; as well as The determining unit is configured to determine the state of the target device based on the feature quantity. Specifically, the first time-series data of the working device operating under the first pressure is acquired and used as the first data. Specifically, the second time-series data of the working device operating under the second pressure is acquired as the second data. The computing unit is further configured to generate feature history data based on the first time series data and the second time series data, wherein the feature history data indicates the change of the feature quantity over time, and The determining unit is further configured as follows: A reference window is set in the historical feature data to include multiple time points prior to the target time point; Calculate the statistical values ​​of the feature quantities associated with the plurality of time points within the reference window; and The state of the target device is determined based on the characteristic quantities at the target time point and the statistical values.

2. The diagnostic system according to claim 1, in, The feature quantities include the difference between the first data and the second data, and the ratio between the first data and the second data. The state of the target device is determined based on the difference and the ratio.

3. The diagnostic system according to claim 1 or 2, wherein, The calculated statistical values ​​include the mean and covariance of the features within the reference window.

4. The diagnostic system according to claim 1 or 2, wherein, The determining unit is further configured to: The state of the target device at each of the multiple target time points is determined by shifting the reference window along the time axis of the historical data of the features for each of the multiple target time points. as well as For each of the plurality of target time points, the statistical value is calculated by excluding any feature quantities in the corresponding reference window that are associated with the abnormal state of the target device.

5. The diagnostic system according to claim 1 or 2 further includes an instruction unit configured to output a first command to cause the motor of the target device to operate the working device under the first pressure and a second command to cause the motor to operate the working device under the second pressure.

6. The diagnostic system according to claim 5, wherein, The second command is output to the motor to increase the pressure operating the working device from the first pressure to the second pressure, without reducing the pressure below the first pressure.

7. The diagnostic system according to claim 5, in, The working device is a welding torch, and The first and second commands are output to the motor when the welding torch is not powered on.

8. The diagnostic system according to claim 7, wherein, When the first and second commands are output to the motor, the pair of electrodes of the welding torch are in contact with each other, but no contact pressure is applied to the electrodes.

9. The diagnostic system according to claim 7, in, The target device is a robot, and The welding torch is mounted on the robot as an end effector.

10. The diagnostic system according to claim 5, in, The instruction unit is also configured to: Repeat a series of operations multiple times, the series of operations including: Perform actual operations to enable the target equipment to process one or more workpieces; and After performing the actual operation, a check operation is performed, which includes outputting the first command and the second command to the motor, wherein the first data generated in response to the first command and the second data generated in response to the second command are stored in a storage device, and The first data and the second data are obtained from the storage device.

11. A diagnostic method, comprising: Acquire first data generated in response to the working device operating the target device under the first pressure; Acquire second data generated in response to operating the working device under a second pressure; Calculate the feature quantity that indicates the relationship between the first data and the second data; as well as The state of the target device is determined based on the aforementioned characteristic quantities. Specifically, the first time-series data obtained under the first pressure is used as the first data. Specifically, the second time-series data obtained under the second pressure is used as the second data. The calculation of the feature quantity includes: generating historical feature data based on the first time series data and the second time series data, wherein the historical feature data indicates the change of the feature quantity over time, and Determining the state of the target device includes: A reference window is set in the historical feature data to include multiple time points prior to the target time point; Calculate the statistical values ​​of the feature quantities associated with the plurality of time points within the reference window; and The state of the target device is determined based on the characteristic quantities at the target time point and the statistical values.

12. A computer-readable storage medium storing processor-executable instructions, the instructions being configured to: Acquire first data generated in response to the working device operating the target device under the first pressure; Acquire second data generated in response to operating the working device under a second pressure; Calculate the feature quantity that indicates the relationship between the first data and the second data; as well as The state of the target device is determined based on the aforementioned characteristic quantities. Specifically, the first time-series data obtained under the first pressure is used as the first data. Specifically, the second time-series data obtained under the second pressure is used as the second data. The calculation of the feature quantity includes: generating feature history data based on the first time series data and the second time series data, wherein the feature history data indicates the change of the feature quantity over time, and Determining the state of the target device includes: A reference window is set in the historical feature data to include multiple time points prior to the target time point; Calculate the statistical values ​​of the feature quantities associated with the plurality of time points within the reference window; and The state of the target device is determined based on the characteristic quantities at the target time point and the statistical values.

Images