JP2642122B2 - Machine tool abnormality monitoring device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial application field) The present invention uses a predetermined membership function which detects a processing load at the time of processing by a machine tool and compares the detected value with a predetermined set value. The present invention relates to a machine tool abnormality monitoring device that identifies a machining state by using the machining load of a machine tool and determines the degree of machining or tool abnormality.

(Prior art) Conventionally, an abnormality monitoring device of this type of machine tool has been
The detected processing load signal is compared with a predetermined set value, and a binary determination such as normal or abnormal is made. However, actually, there are cases where it is determined that the processing load is abnormal even if the processing load is lower than the set value, and that when the processing load is higher than the set value, it is determined that the processing load is normal and not abnormal. That is, there is a possibility that an erroneous judgment is made in the judgment based on the unique set value. For this reason, a monitoring device that uses a plurality of sensors to improve the determination accuracy has been considered. However, since the respective detection values of the plurality of sensors also perform a binary determination to be compared with each of the predetermined set values, the determination accuracy near the set value has not improved much.

Here, such a prior art is shown in FIG. 3 and its operation will be described.

The machining load signals LS1 and LS2 separately detected by the two machining load sensors 1 and 2 are input to comparators 7 and 8 as actual machining loads via amplifiers / filters 3 and 4, respectively. Correspondingly predetermined setting values 5 and 6 are compared by comparators 7 and 8. 2
The comparison results RS1 and RS2 of the two comparators 7 and 8 are input to a control signal generator 9 that generates a control signal CS of the machine tool. The control signal generator 9 outputs the actual machining load obtained from the machining load sensors 1 and 2. When both are larger than the set values 5 and 6, an abnormal signal AN is generated. In this example, two processing load sensors 1
And 2 are used to increase the generation accuracy of the abnormal signal AN.

(Problems to be Solved by the Invention) In the conventional abnormality monitoring device, the determination based on the comparison between the actual processing load and the set value is based on the binary determination of “1” or “0” shown in FIG. is there. In other words, when the actual processing load exceeds the set value, it is determined that the actual processing load is abnormal (degree of abnormality = “1”), and when the actual processing load is lower than the set value, it is normal (degree of abnormality = “0”). The judgment result has changed stepwise with the set value as a boundary. However, the degree of the abnormal state with respect to the actual actual machining load changes smoothly as shown in FIG. 4 (B). Therefore, in the case of the binary decision method as shown in FIG. 4A, the shaded portion in FIG. 4B is erroneously decided. Therefore, as in the abnormality monitoring device shown in FIG. 3, a method is used in which two machining load sensors 1 and 2 are used to determine that "there is an abnormality when the result of both sides is abnormal". According to this method, the possibility of erroneously determining an abnormality when normal is small. However, the method of judging “the result is abnormal when the result of the other party is abnormal” is to make a judgment that “the result is normal when either result is normal”. The likelihood of judging is not diminished, but rather increased.

The present invention has been made in view of the above problems, and an object of the present invention is not to judge a difference between a detected actual machining load and a set value in a binary manner, but to express a degree of abnormality by a numerical value. By using a membership function that correlates with the degree of actual machining load and cutting time detected by multiple sensors, the degree of abnormality is determined, and abnormality monitoring of machine tools that enables more accurate abnormality determination It is to provide a device.

(Means for Solving the Problems) The present invention relates to a machine tool abnormality monitoring device for detecting a machining load by detecting a machining load on a machine tool, and an object of the present invention is to detect the machining load. A first memory 11 for storing a processing load sensor 30 and a membership function indicating a temperature of an abnormal load corresponding to the magnitude of the processing load;
And a second memory 13 for storing a membership function indicating a degree of the machining abnormality corresponding to the magnitude of the abnormal load, and a membership stored in the first and second memories 11, 13. This is achieved by providing first calculation means 12 and 14 for obtaining the degree of the processing abnormality from the processing load based on a function. Further, the degree of the processing abnormality is determined by a cutting time meter 17 for accumulating the cutting time of the used tool, a third memory 18 for storing a membership function indicating the degree of tool life according to the length of the cutting time, A fourth memory for storing a membership function indicating a degree of machining abnormality caused by a degree of tool life; and a use function based on the membership functions stored in the third and fourth memories. Second calculating means 19 and 21 for calculating the degree of the processing abnormality from the cutting time of the tool, and the degree of processing abnormality determined by the first calculating means 12 and 14 and the second calculating means 19 and 21. A third calculating means 22 for synthesizing the degree of machining abnormality and a fourth calculating means 23 for calculating and calculating a center of gravity of the degree of combined processing abnormality are provided. It can also be determined.

(Operation) The present invention introduces a membership function for associating the amount of machining load exceeding a set value and the degree of an abnormal state, instead of making a binary judgment of the normal / abnormal state of a machine tool, and making a judgment based on the degree of abnormality. Thus, when a plurality of sensors are used, the degree of abnormality can be calculated based on the relationship between the sensors. For this reason, a means for detecting the processing load of the machine tool and a means for detecting the cutting time are provided, and a membership function indicating a degree of abnormal load corresponding to the magnitude of the processing load and a degree of tool life are provided. A membership function indicating the degree of the life of the tool from the cutting time, and a function indicating the degree of machining abnormality caused by the abnormality of the tool from the degree of abnormal load; By having a function representing the degree of machining abnormality that occurs with the tool life from the degree of life, and by obtaining the degree of machining abnormality that occurs in relation to a more reliable tool from the degree of each tool abnormality Implements an abnormality monitoring device that links multiple ambiguous inputs (machining load, cutting time) and detects machining errors or tool errors that occur in relation to tools with higher accuracy. It can be.

(Embodiment) FIG. 1 shows an embodiment of the abnormality monitoring apparatus of the present invention. This embodiment shows a machine tool abnormality monitoring device that detects a machining abnormality based on the detection of the machining load by the machining load sensor 30 and the cutting time by the cutting time meter 17, and outputs an alarm AN.

The membership function 11 is a function representing the degree of the abnormal load state and is input to the membership value calculator 12, and the membership function 13 is a function of the degree of the machining abnormality occurring from the degree of the abnormal load due to the tool abnormality at that degree. Is input to the MIN calculator 14. The membership function 15 is a function indicating the degree of the tool life state and is input to the membership value calculator 16, and the membership function 18 is a function indicating the degree of the tool life according to the magnitude of the cutting time. The membership function 20 is input to the MIN calculator 21 and is a function representing the degree of machining abnormality generated due to the tool abnormality. MIN calculator 14 and
The degrees AR1 and AR2 obtained in 21 are MAX arithmetic units 22 respectively.
The calculated value MAR is input to the barycenter calculator 23, and the calculated barycenter value WS is input to the comparator 25. The set value 32 is input to the arithmetic unit 10 and the set value 24 is input to the comparator 25.

In such a configuration, the processing load signal LS detected by the processing load sensor 30 is input to the subtracter 10 as an actual processing load via the amplifier / filter 31 and the difference between the processing load signal LS and the set value 32 predetermined by the subtracter 10 is obtained. ER is required. Based on the magnitude of the difference ER between the actual machining load and the set value 32, the current value D1 of the abnormal load is obtained by the membership value calculator 12 based on the membership function 11 representing the degree of the abnormal load. In FIG. 2 (A), Z ₀ is the difference ER between the actual machining load and the set value 32, and α indicates the normalized abnormal load. Next, based on the membership function 13 indicating the degree of machining abnormality caused by the tool abnormality in the degree of abnormal load and the current degree of abnormal load D1, the degree of machining abnormality AR1 caused by the tool abnormality is calculated by MIN. It is obtained by the arithmetic unit 14. The shaded portion in FIG. 2 (B) is a set indicating the degree AR1 of the processing abnormality caused by the tool abnormality. Further, based on the magnitude of the difference ER between the actual machining load obtained by the subtractor 10 and the set value 32, a membership value calculator 16 (membership value calculation) is performed based on a membership function 15 indicating the degree of the tool life state. Table 12
The same operation as is performed) to obtain the current tool life degree D2. Difference ER between x ₀ is the set value 32 and the actual machining load of FIG. 2 (C)
And β represents the normalized tool life. Also, apart from the tool life degree D2 obtained from the machining load as described above, the cutting time CT from the cutting time meter 17 that measures the cutting time of the tool in use, and the degree of tool life depending on the size of the cutting time And the membership value calculator 19 obtains the tool life degree D3 from the membership function 18 representing In FIG. 2 (D), y ₀ is the cutting time CT, and γ indicates the degree of the normalized tool life.

Next, a membership function 20 representing the degree of machining abnormality that occurs with the tool life, the degree D2 of tool life determined from the machining load (β in FIG. 2 (C)), and the tool determined from the cutting time The degree of life D3 (γ in Fig. 2 (D)) and MIN
A machining error occurring with the tool life by the MIN computing unit 21 which is input to the computing unit 21 and obtains the degree of machining abnormality occurring with the tool life based on the small value of the tool life and the membership function 20 The degree AR2 of the machining abnormality that occurs with this is obtained. The MIN calculator 14 receives an input α (α
≤1) and "1" to determine the degree of machining abnormality caused by a tool abnormality.
1 and 14 perform the same operation. The shaded portion in FIG. 2 (E) is a set indicating the degree AR2 of machining abnormality that occurs with the tool life. From the degree of processing abnormality AR1 generated due to the tool abnormality obtained by the MIN calculators 14 and 21 and the processing abnormality AR2 generated according to the tool life, the MAX calculator 22
Set sum of degrees of machining abnormalities caused by tools due to
Ask for MAR. The shaded portion in FIG. 2 (F) indicates this degree. Further, the center-of-gravity calculator 23 calculates a weighted average center-of-gravity value WS from a set sum MAR of machining abnormalities occurring in relation to the tool.
(The center of gravity of the hatched portion in FIG. 2 (F)) was obtained, and the obtained value was obtained.
WS (z _{0 in} FIG. 2 (F)) is compared with a predetermined set value 24 by a comparator 25, and if it is larger than the set value 24, an alarm signal
Outputs AN. With the above, the alarm signal with higher accuracy in the case of machining abnormalities related to the tool
AN can be output.

Further, in this example, the MIN computing units 14 and 21 provide an α value (α ≦
Although a smaller value such as 1) or β value (β <γ) is adopted, the function does not necessarily have to adopt the smaller value. The input / output characteristics of the membership functions 11, 13, 15, 18, and 20 are set based on the experience of the worker. In the input / output characteristic setting operation of the membership function, first, input values—output values are sequentially input via data input means (for example, a keyboard, a numeric key switch, etc.) not shown, for example, when the processing load is 50% —the membership value. Data input work is performed as discrete value data, such as 0.6.

Thus, the membership value between the input discrete value data is generated by linearly interpolating the left and right discrete value input data.

When the above-mentioned data input operation is completed, each membership function sets input variables and output variables corresponding to data display means (for example, a color CRT, a liquid crystal display, etc.), not shown, to the horizontal axis or the vertical axis, respectively. In addition to the two-dimensional graph display, the individual membership values are displayed in a different manner in the display luminance of the data display means (when the degree of abnormality is a large value, the screen display luminance is made brighter) and in the display colors (in the case of abnormalities). When the degree is large, the screen display color is redened) or displayed on the screen in correspondence with special characters, so that the input / output characteristics of each membership function can be understood.

(Effects of the Invention) As described above, according to the abnormality monitoring device of the present invention, a plurality of processing loads and cutting times having ambiguity are supplied to a monitoring system, and a more accurate abnormality of a machine tool is detected. Can be.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of an abnormality monitoring device according to the present invention, FIG. 2 is a diagram for explaining the operation thereof, and FIG.
FIG. 1 is a block diagram showing an example of a conventional monitoring device, and FIG. 4 is a diagram for explaining its operation. 1,2,30 …… Processing load sensor, 5,6,24,32 …… Set value, 7,
8,25… Comparator, 9… Control signal generator, 10… Calculator, 11,13,15,18,20… Membership function memory, 12,1
6,17 ... Membership value calculator, 14,21,22,23 ... Calculator.

Claims (5)

(57) [Claims] 1. A machine tool abnormality monitoring apparatus for detecting a machining load by detecting a machining load on a machine tool, comprising: a machining load sensor for detecting the machining load; and an abnormality corresponding to the magnitude of the machining load. A first memory 11 for storing a membership function indicating the degree of the load, a second memory 13 for storing a membership function indicating the degree of the machining abnormality corresponding to the magnitude of the abnormal load, A first calculating means for obtaining the degree of the processing abnormality from the processing load based on the membership functions stored in the first and second memories 11 and 13
An abnormality monitoring device for a machine tool, comprising: (12) and (14). 2. A cutting time meter for accumulating the cutting time of a tool to be used.
17, a third memory 18 for storing a membership function indicating the degree of life of the tool according to the length of the cutting time, and a membership function indicating the degree of abnormal processing caused by the degree of life of the tool. A second memory means 20 for obtaining the degree of machining abnormality from the cutting time of the tool to be used based on the membership functions stored in the third and fourth memories 18 and 20 A third calculating means for combining the degree of machining abnormality obtained by the first calculating means 12 and 14 and the degree of processing abnormality obtained by the second calculating means
2. The machine tool abnormality monitoring device according to claim 1, further comprising: a fourth arithmetic means for calculating and calculating a center of gravity of the combined degree of processing abnormality, wherein the center of gravity is used as a final degree of processing abnormality. . 3. A comparison means 25 for comparing the value of the final degree of machining abnormality with a threshold value set in advance in a fifth memory 24, and outputs an alarm or a control signal of the machine tool. The machine tool abnormality monitoring device according to claim 1. 4. A data input means for defining the availability characteristic of each membership function and a data display means for the data, wherein the input / output characteristics of the membership function are discrete via the data input means. 4. The machine tool abnormality monitoring device according to claim 1, wherein data is input as value data, and values between these data are generated by linear interpolation. 5. An input / output characteristic of each of the membership functions,
Input data and output data are displayed two-dimensionally on the data display means in correspondence with the horizontal axis and the vertical axis, respectively, and the individual membership values are displayed on the data display means with different display brightness and different display colors. The abnormality monitoring device for a machine tool according to claim 4, wherein the abnormality can be determined by displaying the characters in correspondence with special characters.