Failure symptom detection method

Abstract

To provide a failure symptom detection method with which it is possible to detect an abnormality symptom that conventionally cannot be detected.SOLUTION: The failure symptom detection method includes: measuring a current value of a servo motor every prescribed time (S201); computing a primary finite difference of the current value (S205); computing the absolute value of a secondary finite difference from the primary finite difference (S206); integrating the sum of secondary finite differences during a first cycle (S210); counting the sum total of the number of times the current changed to decrease from increasing during one cycle of control of the servo motor and the number of times the current changed to increase from decreasing (S212); and comparing the sum total of the number of times a change occurred with a prescribed upper-limit threshold and comparing the integrated sum of secondary finite differences with a prescribed upper-limit threshold, thereby determining a symptom of equipment failure.SELECTED DRAWING: Figure 2

Description

[Technical field]

[0001] The present invention relates to a method for detecting a symptom of a failure. [Background technology]

[0002] In manufacturing facilities, ball screws that transmit torque for reducers of industrial robots and servo motors can be damaged, causing the equipment to stop operation and disrupting production activities. Patent Document 1 describes a method and device for diagnosing abnormalities in reducers. In order to avoid such unexpected damage, it is necessary to replace parts before a failure occurs by detecting signs of a drive system failure. In addition, because these parts are expensive and have a complex structure, it is difficult to replace them frequently. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2008-032477 A Summary of the Invention [Problem to be solved by the invention]

[0004] Therefore, there is a need to detect signs of failure and replace parts at the appropriate time just before the equipment is shut down. Since parts cannot be replaced in time if signs are detected only a few days in advance, it is necessary to detect signs of failure several tens of days in advance, but there is no established method for detecting clear signs.

[0005] An object of the present disclosure is to provide a method for detecting a failure symptom that can accurately diagnose an abnormality by introducing a quantitative judgment criterion. [Means for solving the problem]

[0006] In one embodiment, a method for detecting signs of failure measures a current value of a servo motor at predetermined time intervals, counts the total number of times the current changes from increasing to decreasing and the total number of times the current changes from decreasing to increasing during one cycle of control of the servo motor, calculates a first-order difference of the current value, calculates an absolute value of a second-order difference from the first-order difference, accumulates the sum of the second-order differences during the one cycle, compares the sum of the number of changes with a predetermined upper limit threshold, and compares the accumulated sum of the second-order differences with a predetermined upper limit threshold to determine signs of equipment failure. Effect of the Invention

[0007] According to the method for detecting a failure symptom of one embodiment, it is possible to capture an abnormality symptom that could not be detected by conventional methods. [Brief description of the drawings]

[0008] [Figure 1] 1 is a block diagram showing an example of an apparatus used in a failure symptom detection method according to a first embodiment; [Diagram 2] 4 is a flowchart showing an example of a method for detecting a failure symptom according to the first embodiment. [Diagram 3] 4 is a graph showing a change in a motor current value. [Figure 4] 13 is a graph showing the sum of absolute values ​​of second differences of motor current values. [Diagram 5] 11 is a graph showing the number of increases and decreases in the first-order difference of a motor current value. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] First embodiment Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Fig. 1 is a block diagram showing an example of an apparatus used in a failure symptom detection method according to the first embodiment. In Fig. 1, the apparatus 100 includes a controller 101, a servo amplifier 102, a servo motor 103, an ammeter 104, a trigger signal input unit 105, an analog current input unit 106, a general-purpose PLC 107, a determination device 108, and a display 109.

[0010] The controller 101 is a circuit that controls a servo motor 103 via a servo amplifier 102 .

[0011] Based on an instruction from the controller 101, the servo amplifier 102 supplies the servo motor 103 with an output (electric power) required for the servo motor 103 to move according to a target value.

[0012] The servo motor 103 is a motor that operates using power from the servo amplifier 102. In addition, the servo motor 103 returns a feedback of its operation to the servo amplifier 102.

[0013] The ammeter 104 measures the current flowing between the servo amplifier 102 and the servo motor 103. Then, the ammeter 104 outputs the measured current value to the analog current input unit 106.

[0014] The trigger signal input unit 105 receives a trigger bit signal for starting one cycle operation from the controller 101. The trigger signal input unit 105 then outputs the trigger bit signal to the data division unit 171.

[0015] The analog current input unit 106 receives an input of an analog current value from the ammeter 104. Then, the analog current input unit 106 converts the current value from an analog value to a digital value. Then, the analog current input unit 106 outputs the converted current value to the data division unit 171.

[0016] The general-purpose PLC 107 includes a data division unit 171 , a first storage unit 172 , a buffer calculation unit 173 , a second storage unit 174 , an averaging unit 175 , and a feature output unit 176 .

[0017] The data division unit 171 divides the current value in units of sampling periods and outputs the divided current value to the first storage unit 172. The first storage unit 172 is a storage circuit that stores a current value. The buffer calculation unit 173 calculates the sum of the absolute values ​​of the first and second differences of the current values.

[0018] The second storage unit 174 stores the sum of the absolute values ​​of the first and second differences of the current values. The averaging unit 175 performs a moving average process on the accumulated values. The feature output unit 176 outputs the total number of times the current changes from increasing to decreasing and the total number of times the current changes from decreasing to increasing during one cycle of servo motor control, and the sum of the accumulated second-order differences to a threshold determination unit 181 of the determination device 108.

[0019] The determination device 108 includes a threshold determination unit 181 , a threshold D / B 182 , and a result output unit 183 .

[0020] The threshold determination unit 181 compares the total number of times the current changes from increasing to decreasing and the total number of times the current changes from decreasing to increasing during one cycle of the servo motor control with a predetermined upper threshold. The threshold determination unit 181 also compares the total sum of the integrated secondary differences with a predetermined upper threshold.

[0021] If these values ​​exceed the upper thresholds, the threshold determination unit 181 determines that there is a failure symptom. A specific method of determination will be described later. The threshold D / B 182 is a database that stores an upper limit threshold. The result output unit 183 outputs the determination result to the display 109 based on the determination by the threshold determination unit 181 .

[0022] The display 109 displays the result of the determination. Specifically, the display 109 displays the presence or absence of a failure symptom. The display 109 is preferably a liquid crystal display or the like.

[0023] Next, a process of the method for detecting a fault symptom will be described with reference to a flowchart of FIG 2, which is a flow chart showing an example of the method for detecting a fault symptom according to the first embodiment.

[0024] First, in step S201, the signal of the ammeter is stored in a register in the PLC, and the process proceeds to steps S202 and S203.

[0025] In step S202, scaling is performed to convert the unit to amperes [A]. Then, the process proceeds to step S204.

[0026] In step S203, a trigger bit signal for starting one cycle operation is received from the equipment control PLC, and the process proceeds to step S204.

[0027] In step S204, it is determined whether or not one cycle is to be started. Specifically, calculation is started at the timing when the trigger signal is turned ON. If one cycle is to be started, the process proceeds to step S205. If one cycle is not to be started, the process returns to S201.

[0028] In step S205, the first-order difference of the current values ​​is calculated, and the process proceeds to steps S206 and S211.

[0029] In steps S206 to S210, the sum of |second differences| is calculated. In step S206, the absolute value of the second differences of the current values ​​is calculated. Then, the process proceeds to step S207.

[0030] In step S207, the calculated value is temporarily stored in the array register, and the process proceeds to step S208.

[0031] In step S208, it is determined whether one cycle has been completed. If one cycle has not been completed, the process proceeds to step S209. If one cycle has been completed, the process proceeds to step S210.

[0032] In step S209, a wait is performed to keep the sampling period constant, and the process returns to step S205.

[0033] In step S210, an accumulation process is performed on the absolute values ​​of the second differences for one cycle, and then the process proceeds to step S215.

[0034] On the other hand, the total number of increases and decreases is calculated in steps S211 to S214. In step S211, noise is removed by neutral current determination. Then, the process proceeds to step S212.

[0035] In step S212, it is determined whether the sign of the first difference value changes. That is, it is determined whether it changes from + to - or from - to +. If the sign of the first difference value changes, proceed to step S213. If the sign of the first difference value does not change, return to step S205.

[0036] In step S213, when the code changes, a count is added to obtain the number of increases or decreases, and the process proceeds to step S214.

[0037] In step S214, it is determined whether one cycle has been completed. If one cycle has not been completed, the process proceeds to step S209. If one cycle has been completed, the process proceeds to step S215.

[0038] In step S215, the sum of the absolute values ​​of the second differences integrated during one cycle is stored in the buffer register. Also, the sum of the number of increases and decreases determined at equal intervals during one cycle is stored in the buffer register. Then, the process proceeds to step S216.

[0039] In step S216, a moving average process is performed on the values ​​stored in the buffer register. The value obtained in step S216 is output as the feature "sum of |second differences|", and the value obtained in step S213 is output as the feature "sum of number of increases and decreases", and the process ends.

[0040] Next, the sign of the first difference used to count the number of times the current value increases and decreases will be explained. FIG. 3 is a graph showing changes in the motor current value. In FIG. 3, the vertical axis represents the motor current value, and the horizontal axis represents time. As shown in the enlarged view of FIG. 3, when the current value increases, the first difference is positive (+). When the current value decreases, the first difference is negative (-). The times when the sign of this first difference changes are counted. Specifically, the number of times it switches from + to - or from - to + is counted.

[0041] Next, the sum of the absolute values ​​of the second differences will be described. Fig. 4 is a graph showing the sum of the absolute values ​​of the second differences of the motor current values. In Fig. 4, the vertical axis shows the sum of the absolute values ​​of the second differences of the motor current values ​​in one cycle of servo motor control, and the horizontal axis shows the date and time.

[0042] Next, the total number of times the first-order difference of the motor current value increases and decreases will be described. Here, the number of times the first-order difference increases and decreases is the number of steps at which the sign of the first-order difference changes, as described above. FIG. 5 is a graph showing the number of times the first-order difference of the motor current value increases and decreases. In FIG. 5, the vertical axis shows the number of times the first-order difference of the motor current value increases and decreases in one cycle of servo motor control, and the horizontal axis shows the date and time.

[0043] When rattle begins to occur in a drive system such as a servo motor, a ripple current begins to occur in an attempt to correct the encoder misalignment caused by the shock, and the current fluctuation increases (for example, see the enlarged view in Figure 3). As a result, the amount of change in the servo motor's current fluctuation also increases. Since the SN ratio is proportional to the amount of shock caused by the rattle, the SN ratio increases according to the amount of rattle, but the beginning of the rattle cannot be detected.

[0044] The number of increases and decreases in the first-order difference can be detected at the stage when the ripple current begins to occur, so that the rattle can be detected at an earlier stage compared to the sum of the absolute values ​​of the second-order differences.

[0045] On the other hand, even if the backlash increases, the signal-to-noise ratio is unlikely to increase, so the accuracy of using it for definitive diagnosis is reduced.

[0046] These two feature amounts are not easily affected by variations in the magnitude of the current value, and therefore can capture signs that could not be detected by conventional methods. Therefore, the two feature amounts are input to the determination device 108, and a threshold value is set for each of them.

[0047] In this way, according to the method for detecting signs of failure in the first embodiment, early detection is performed using the "total number of increases and decreases," and abnormal signs are definitively judged using the "sum of |second differences|," making it possible to prevent abnormalities before they occur. That is, by collecting the current values ​​of the servo motor, calculating two feature amounts that are not easily affected by disturbances, and comparing them with a predetermined upper limit threshold value, it is possible to capture signs of equipment failure and prevent abnormalities before they occur.

[0048] Furthermore, the method of detecting a failure symptom according to the first embodiment can be implemented at a lower cost than a dedicated determination machine because the method can be implemented using a general-purpose PLC.

[0049] Furthermore, the method for detecting a failure symptom according to the first embodiment can be used by any RB manufacturer or servo device manufacturer.

[0050] The present invention is not limited to the above-described embodiment, and can be modified as appropriate without departing from the spirit of the present invention. For example, a torque value may be used instead of a current value. A dedicated microcomputer or PC may be used instead of a PLC. A current value may be obtained from a controller. [Explanation of symbols]

[0051] 100 device, 101 controller, 102 servo amplifier, 103 servo motor 104 ammeter, 105 trigger signal input section, 106 analog current input section 107 General-purpose PLC, 108 Judgment device, 109 Display, 171 Data division unit 172 first memory unit, 173 buffer calculation unit, 174 second memory unit 175 Averaging processing unit, 176 Feature output unit 181 threshold determination unit, 182 threshold D / B, 183 result output unit

Claims

[Claim 1] Measure the servo motor current value at specified intervals, Calculate the first-order difference of the current value, Calculating the absolute value of the second difference from the first difference; Integrating the sum of the second-order differences during the one cycle; Counting the total number of times the current changes from an increase to a decrease and the total number of times the current changes from a decrease to an increase during one cycle of control of the servo motor; A method for detecting signs of failure that determines signs of equipment failure by comparing the total number of changes with a predetermined upper threshold value, and by comparing the sum of the integrated second-order differences with a predetermined upper threshold value.

Images