Optical component soundness estimation device, laser processing system, and optical component soundness estimation method

Abstract

An optical component soundness estimation device (20) is provided with: a parameter estimation unit (21) that determines, on the basis of either a laser light oscillation command to a laser oscillator (40) or a laser light measurement value of the laser oscillator (40), and a measurement value of the temperature of a member heated by generation or emission of laser light, an estimation parameter being an estimation value of a parameter indicating a relationship between states of the laser light and temperatures; and a soundness estimation unit (22) that estimates soundness of an optical component on the basis of the estimation parameter and outputs a soundness estimation result indicating a degree of soundness.

Description

Optical component soundness estimation device, laser processing system, and optical component soundness estimation method

[0001] The present disclosure relates to an optical component health assessment device, a laser processing system, and a method for assessing the health of an optical component.

[0002] In a laser oscillator in a laser processing device, abnormal heating may occur due to a defect in an optical component in the path of the laser light, resulting in damage to the optical component. To prevent damage to the optical component, a technique has been proposed in the past for detecting signs of abnormality by measuring the temperature of the optical component and comparing it with a threshold value. Furthermore, Patent Document 1 listed below proposes a technique for reducing the risk of damage to the laser oscillator due to abnormal heating at a location other than the installation point of the temperature sensor.

[0003] Japanese Patent Application Laid-Open No. 2020-47816

[0004] However, in the conventional technology, when the state of the laser light used changes, there was a problem in that it was not possible to estimate with high accuracy the health of the optical components based on the measurement results of the state of the laser light or the temperature of the optical components.

[0005] The present disclosure has been made in consideration of the above, and aims to provide an optical component healthiness estimation device that can estimate the healthiness of optical components with high accuracy even when the state of the laser light used changes.

[0006] To solve the above-mentioned problems and achieve the object, an optical component health estimation device according to the present disclosure includes a parameter estimation unit and a health estimation unit. The parameter estimation unit determines an estimation parameter, which is an estimate of a parameter representing the relationship between the state of the laser beam and the temperature, based on either a laser beam oscillation command for a laser oscillator or a measurement value of the laser beam in the laser oscillator, and a measurement value of the temperature of a member heated by the generation or emission of the laser beam. The health estimation unit estimates the health of the optical component based on the estimation parameter, and outputs a health estimation result representing the degree of health.

[0007] The optical component healthiness estimation device according to the present disclosure has the advantage of being able to estimate the healthiness of optical components with high accuracy even when the state of the laser light being used changes.

[0008] FIG. 1 is a diagram showing an example of the configuration of a laser processing system according to embodiment 1; FIG. 2 is a diagram showing an example of the configuration of the laser module shown in FIG. 1; FIG. 3 is a block diagram explaining the function of an optical component healthiness estimation device according to embodiment 1; FIG. 4 is a block diagram showing an example of a hardware configuration that realizes the function of an optical component healthiness estimation device according to embodiment 1; FIG. 5 is a block diagram explaining the function of a healthiness estimation unit provided in an optical component healthiness estimation device according to embodiment 2; FIG. 6 is a diagram showing a simulation result when an optical component healthiness estimation device according to embodiment 2 is implemented; FIG. 2 is a functional block diagram showing the configuration of the parameter estimation unit provided in the optical component health estimation device according to embodiment 5; FIG. 3 is a diagram showing the simulation results when the optical component health estimation device according to embodiment 5 is implemented; FIG. 4 is a block diagram showing the function of the optical component health estimation device according to embodiment 6; and FIG. 5 is a diagram showing the simulation results when the optical component health estimation device according to embodiment 6 is implemented.

[0009] An optical component health assessment device, a laser processing system, and an optical component health assessment method according to embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings. Note that in the following description, multiple components of the same type will be indicated by reference numerals with subscripts, but when describing these components without distinguishing between them, the subscripts will be omitted as appropriate.

[0010] Embodiment 1. Fig. 1 is a diagram showing an example of the configuration of a laser processing system 1 according to embodiment 1. The laser processing system 1 shown in Fig. 1 is a system that irradiates a workpiece 72 with laser light L to laser-process the workpiece 72. As shown in Fig. 1, the laser processing system 1 includes a laser processing machine 2 that processes the workpiece 72 with the laser light L, a control device 10 that controls the laser processing machine 2, and an optical component soundness estimation device 20.

[0011] The optical component healthiness estimation device 20 is a device that estimates the healthiness of optical components. The optical component healthiness estimation device 20 may be a component of the laser processing system 1, or may be configured as a device separate from the laser processing system 1, i.e., as an external device configured to be able to communicate with the laser processing system 1. Furthermore, the optical component healthiness estimation device 20 may be a device configured inside the control device 10.

[0012] The laser processing machine 2 includes a laser oscillator 40, an optical fiber 50, a processing head 60, and a relative moving unit 65. The control device 10 includes an oscillator control unit 12, a calculation unit 14, and a processing machine control unit 16.

[0013] The laser oscillator 40 includes, as a light source device that generates laser light L, four drive power sources 43-1, 43-2, 43-3, and 43-4, four laser modules 41-1, 41-2, 41-3, and 41-4 that are connected to the four drive power sources 43, respectively, and a combiner unit 42 that receives the optical output from each of the four laser modules 41. Note that in Fig. 1, the number four of the laser modules 41 and drive power sources 43 is an example, and the number of laser modules 41 and drive power sources 43 may be one or more.

[0014] The calculation unit 14 calculates the driving conditions for each of the multiple laser modules 41 of the laser oscillator 40. The processing machine control unit 16 controls the processing head 60 and the relative movement unit 65 based on the processing conditions. The oscillator control unit 12 controls the driving power supply 43 of the laser oscillator 40 based on a laser light oscillation command given to the laser oscillator 40. In this document, the oscillation command will be described as a command value for the power output by the laser oscillator 40.

[0015] The optical fiber 50 guides the laser light L generated by the laser oscillator 40 to the machining head 60. The workpiece 72 is supported by the workpiece support part 70. The relative position of the workpiece support part 70 with respect to the machining head 60 is controlled by the relative movement part 65. The machining head 60 processes the workpiece 72 by irradiating the laser light L onto the workpiece 72 supported by the workpiece support part 70.

[0016] In the laser oscillator 40, each laser module 41 generates a laser beam La. Each laser module 41 generates laser beam La of the same wavelength. The combiner unit 42 collects the laser beam La output from each laser module 41 to generate laser beam L.

[0017] The optical component health estimation device 20 estimates the health of the optical components based on the laser light emission command issued by the oscillator control unit 12 to the laser oscillator 40 and the temperature detection value by a temperature sensor not shown in Figure 1, and calculates a health estimation result that indicates the degree of health.

[0018] The laser module 41, the combiner unit 42, and the optical fiber 50 are examples of optical components, but in this document, the term "optical components" also refers to optical components such as lenses, mirrors, optical filters, diffraction gratings, diodes, acousto-optic elements, and wavelength conversion elements, as well as adhesive materials for the optical fiber 50, which are heated by the generation or emission of laser light. In addition, the components of the laser module 41, which will be described later, are also optical components.

[0019] Fig. 2 is a diagram showing an example of the configuration of the laser module 41 shown in Fig. 1. As shown in Fig. 2, the laser module 41 includes a plurality of semiconductor lasers 411 that generate laser light Lb, and a pump combiner 412.

[0020] The semiconductor laser 411 generates laser light Lb. In the first embodiment, the semiconductor laser 411 is a semiconductor laser package. The semiconductor laser package is an integrated unit in which a semiconductor laser is mounted on a heat sink. The pump combiner 412 collects the multiple laser light beams Lb generated from the multiple semiconductor lasers 411 and outputs a single laser light beam La.

[0021] Although not shown in Figures 1 and 2, the laser processing system 1 is equipped with a cooling device for cooling the optical components, and a cooling water passage for cooling the optical components is provided inside the laser module 41.

[0022] FIG. 3 is a block diagram illustrating the function of the optical component health estimation device 20 according to the first embodiment. As shown in FIG. 3 , the optical component health estimation device 20 includes a parameter estimation unit 21 and a health estimation unit 22. Around the optical component health estimation device 20 are shown the oscillator control unit 12 provided in the laser processing system 1, an optical component 80 whose health is to be estimated, and a temperature sensor 82 for measuring the temperature of the optical component 80. In FIG. 3 , a laser module 41 and a combiner unit 42 are shown as representative examples of the optical component 80 whose health is to be evaluated. Other examples of the optical component 80 whose health is to be evaluated include the temperature of a semiconductor element in a drive circuit within the laser oscillator 40, the temperature of the cooling water or cold water plate in the cooling water channel, the output temperature of the combiner unit 42, the temperature of a cladding light removal component within the combiner unit 42, and the temperature of components within the processing head 60.

[0023] Next, we will explain the operation of the optical component healthiness estimation device 20. The temperature sensor 82 measures the temperature of the optical component 80. There are no particular restrictions on the location where the temperature sensor 82 is placed, but if possible, it is best to install it near the target component whose healthiness is to be evaluated.

[0024] The parameter estimation unit 21 receives the temperature measurement value y(k) obtained by the temperature sensor 82 and the oscillation command u(k). The parameter estimation unit 21 calculates an estimated parameter p based on the temperature measurement value y(k) and the oscillation command u(k). A first example of the estimated parameter p is given by the following equation (1).

[0025]

[0026] The estimated parameter p shown in the above equation (1) １In the above, y~ is the average value of the temperature measurement value y(k), and u~ is the average value of the oscillation command u(k). Hereinafter, "y~" will be referred to as the "average measured temperature value" and "u~" as the "average oscillation command value". These average measured temperature value y~ and average oscillation command value u~ can be found by arithmetic averaging. "~" is an alternative notation for the horizontal line (overbar) that should be drawn above the text. In this paper, horizontal lines (overbars) are replaced with "~" except for the mathematical formulas inserted as images.

[0027] In addition, in the above formula (1), k is an index representing the sampling point, and takes a value from 1 to N. The index k is synonymous with the time of measurement, and N is the number of sampling points. The unit of the oscillation command u(k) is [w] or [kW], and the unit of the measured temperature value y(k) is [°C].

[0028] The soundness estimation unit 22 estimates the estimated parameter p １ The soundness of the optical component 80 is estimated based on the threshold value TH １ For example, the health estimation unit 22 can use the estimation parameter p １ is the threshold value TH １ If the threshold value TH exceeds the threshold value TH, it can be determined that an abnormality has occurred in the optical component 80. １ is the estimated parameter p when the optical component 80 is in good condition. １ , and the estimated parameter p when the optical component 80 is unhealthy. １ The determination can be made by the user with reference to at least one of the following:

[0029] Estimated parameter p １ If the threshold value TH １ It is possible to estimate the health of the optical component 80 without changing the parameter p. For example, when the oscillation command average value u is 5 [kW], the measured temperature average value y of a healthy optical component 80 is 100 [°C]. １is 100 [°C] / 5 [kW] = 20 [m°C / W]. Also, if this optical component 80 deteriorates, the measured temperature average value y~ will be higher than when it is in good condition, even with the same oscillation command u(k). For example, when the oscillation command average value u~ is 5 [kW], the measured temperature average value y~ is 160 [°C]. In this case, the estimated parameter p １ is 160[°C] / 5[kW]=32[m°C / W]. For example, the threshold value TH １ If the temperature is 30 [m℃ / W], the estimated parameter p １ This state where the temperature is 32 [m°C / W] can be detected as an abnormality.

[0030] Furthermore, in the same laser processing system 1, when the oscillation command average value u is 1 [kW], the measured temperature average value y of the same healthy optical component 80 is 20 [°C]. In this case, the estimated parameter p １ is 20[°C] / 1[kW]=20[m°C / W]. Also, if this optical component 80 deteriorates, the average measured temperature value y~ will be higher than when it is in good condition, even with the same oscillation command u(k). For example, when the oscillation command u(k) is 1[kW], the average measured temperature value y~ is 40[°C]. In this case, the estimated parameter p １ is 40 [°C] / 1 [kW] = 40 [m°C / W]. The threshold value TH set when the oscillation command average value u~ is 5 [kW] １ = 30 [m ° C / W], the estimated parameter p １ This state where the estimated parameter p is 40 [m°C / W] can be detected as an abnormality. １ If the oscillation command average value u is used, the threshold value TH １ It is possible to estimate the soundness of the optical component 80 without changing the

[0031] In the above description, the parameter estimation unit 21 calculates the estimated parameter p based on the temperature measurement value y(k) and the oscillation command u(k), but this is not limiting. The oscillation command u(k) may be any information as long as it is information about a heat source that heats the optical component 80. For example, it may be information about a measurement value of the laser light in the laser oscillator 40. Alternatively, it may be information about a measurement value of the power of the laser light L, a current flowing through a laser diode (not shown), a voltage applied to the laser diode, or an input current, input voltage, output current, or output voltage in a drive circuit (not shown) in the laser oscillator 40.

[0032] In the prior art, the temperature of the optical component 80 is measured and compared with a threshold value to determine whether the optical component 80 is in an abnormal state. For example, when the oscillation command u(k) is 1 kW, the temperature of the optical component 80 when healthy is 20°C, and the temperature when deteriorated is 30°C. In this case, if the threshold value is set to 25°C, for example, it is possible to determine whether the optical component 80 is in an abnormal state. However, when the oscillation command u(k) is 5 kW, for example, in the above example, the average measured temperature y~ of the optical component 80 is 100°C, so the temperature of the optical component 80 will certainly rise to about 100°C. Therefore, if the state of the optical component 80 is determined using the same threshold value, it will be determined to be abnormal even if the optical component 80 is in a healthy state. On the other hand, in the method of the first embodiment, the threshold value TH １ The value to be compared with the estimated parameter p １ Therefore, even if the oscillation command average value u~ is different, the threshold value TH １ It is possible to estimate the soundness of the optical component 80 without changing the

[0033] The threshold value TH １ Regarding the threshold value TH, if the properties of the optical component 80, such as the function, performance, material, etc., of the optical component 80 to be judged are different, the threshold value TH １ It is desirable to set different threshold values ​​TH depending on the properties of the optical component 80. １By using the above, it is possible to further improve the accuracy of estimating the soundness of the optical component 80.

[0034] In the above, the estimated parameter p １ However, the temperature measurement value y(k) may be pre-processed to remove the influence of the ambient temperature, etc. The following equation (2) shows a second example of the estimation parameter p.

[0035]

[0036] The estimated parameter p ２ is an example of the estimated parameter p after preprocessing. ０ is the initial temperature. ０ is, for example, the temperature immediately before laser processing is performed. ０ The estimated parameter p may be the temperature at power-on, the temperature of the cooling water, the outside air temperature, etc. ２ Since is a parameter that takes into account only the temperature increase between before and after processing, it is possible to improve the accuracy of estimating the soundness.

[0037] In addition, the above estimated parameter p １ , p ２ Instead, the estimated parameter p ３ may also be used.

[0038]

[0039] In the above formula (3), s ｕｙ is the covariance of the oscillation command u(k) and the temperature measurement value y(k). ｕ is the standard deviation of the oscillation command u(k), and s ｙ is the standard deviation of the temperature measurement value y(k). Note that preprocessing using a filter such as a low-pass filter, a high-pass filter, or a band-pass filter may be applied to each of the oscillation command u(k) and the temperature measurement value y(k). In addition, the estimated parameter p ３In the calculation of (a), it is not necessary to use all the oscillation commands u(k) and all the temperature measurement values ​​y(k) acquired at a fixed sampling period. For example, based on conditions such as the presence or absence of laser light output, the magnitude of the measurement value, and the magnitude of the temperature change, those that satisfy these conditions can be extracted, and the extracted data can be used to calculate the estimated parameter p ３ may be calculated.

[0040] In addition, the threshold value TH set in the health estimation unit 22 １ is one estimated parameter p １ Instead, multiple estimation parameters p １ Specifically, the soundness estimation unit 22 determines the estimated parameter p when the optical component 80 is sound. １ is determined multiple times, and the determined multiple estimated parameters p １ The health estimation unit 22 calculates the average μ and standard deviation σ based on the estimated parameters p １ If is determined 100 times, the average μ and standard deviation σ at that time can be expressed by the following equations (4) and (5).

[0041]

[0042] At this time, the threshold value TH ２ is set, for example, as in the following equation (6).

[0043]

[0044] The soundness estimation unit 22 estimates the estimated parameter p １ is the threshold value TH ２ If the standard deviation σ is exceeded, the state of the optical component 80 is determined to be abnormal. In the above formula (6), X is an arbitrary positive real value assigned to the standard deviation σ, and is preferably set to a value of about 2 to 5.

[0045] The threshold value TH in the above equation (6) ２ is the estimated parameter p １ The upper limit of the estimated parameter p １ is the threshold that determines the upper limit of the estimated parameter p １ The lower limit of the estimated parameter p １The threshold value TH that determines the lower limit ３ The threshold value TH may be set. ３ is set, for example, as in the following equation (7).

[0046]

[0047] The soundness estimation unit 22 estimates the estimated parameter p １ is the threshold value TH ３ If the difference is smaller than , the state of the optical component 80 is determined to be abnormal.

[0048] These thresholds TH ２ or threshold value TH ３ In this case, the soundness estimation unit 22 may set both of the estimation parameters p １ is the threshold value TH ２ or when the threshold value TH ３ If the difference is smaller than , the state of the optical component 80 can be determined to be abnormal.

[0049] In the above description, the health estimation result for the optical component 80 is described as a binary determination result of whether the state of the optical component 80 is abnormal or not, i.e., whether the state of the optical component 80 is normal or abnormal, but this example is not limiting. Because the optical component 80 deteriorates with use, a health estimation result indicating the degree of deterioration of the optical component 80 may be output. As a health estimation result indicating the degree of deterioration, for example, the following information (a) to (c) may be output to a user or an administrator.

[0050] (a) The condition of the optical component 80 is abnormal or normal. (b) The condition of the optical component 80 is not abnormal, but is in a state of deterioration that is close to abnormal. (c) The remaining life of the optical component 80 is "approximately XX hours." The remaining life of the optical component 80 will be described in other embodiments below.

[0051] As described above, the optical component health estimation device according to the first embodiment includes a parameter estimation unit and a health estimation unit. The parameter estimation unit determines estimation parameters, which are estimates of parameters representing the relationship between the state of the laser beam and the temperature, based on either a laser beam oscillation command for the laser oscillator or a measurement value of the laser beam in the laser oscillator, and a measurement value of the temperature of a member heated by the generation or emission of the laser beam. The health estimation unit estimates the health of the optical component based on the estimation parameters and outputs a health estimation result representing the degree of health. In the optical component health estimation device according to the first embodiment, the health estimation unit estimates the health of the optical component based on the estimation parameters determined by the parameter estimation unit, making it possible to estimate the health of the optical component with high accuracy even when the state of the laser beam used changes.

[0052] Furthermore, the laser processing system according to the first embodiment includes a laser processing machine that processes a workpiece with laser light, a control device that controls the laser processing machine, and an optical component healthiness estimation device that estimates the healthiness of an optical component. The optical component healthiness estimation device includes a parameter estimation unit and a healthiness estimation unit. The parameter estimation unit determines an estimation parameter that is an estimate of a parameter that represents the relationship between the state of the laser light and the temperature based on either a laser light oscillation command for the laser oscillator or a measurement value of the laser light in the laser oscillator, and a measurement value of the temperature of a member heated by the emission of the laser light. The healthiness estimation unit estimates the healthiness of the optical component based on the estimation parameter and outputs a healthiness estimation result that represents the degree of healthiness. In the laser processing system according to the first embodiment, the healthiness estimation unit estimates the healthiness of the optical component based on the estimation parameter determined by the parameter estimation unit. Therefore, it is possible to estimate the healthiness of the optical component with high accuracy even when the state of the laser light used changes.

[0053] Furthermore, the optical component health estimation method according to the first embodiment is a method for estimating the health of an optical component provided in a laser processing system. This health estimation method includes the following first and second steps. The first step is a step of determining an estimation parameter, which is an estimate of a parameter representing the relationship between the state of laser light and temperature, based on either a laser light oscillation command issued to a laser oscillator or a measurement value of the laser light emitted from the laser oscillator, and a measurement value of the temperature of a component heated by the emission of the laser light. The second step is a step of estimating the health of the optical component based on the estimation parameter and outputting a health estimation result representing the degree of health. According to the optical component health estimation method according to the first embodiment, the estimation parameter is determined in the first step, and the health of the optical component is estimated in the second step based on the estimation parameter determined in the first step. Therefore, it is possible to estimate the health of the optical component with high accuracy even when the state of the laser light used changes.

[0054] At the end of the first embodiment, a description will be given of a hardware configuration for realizing the functions of the above-described optical component healthiness estimation apparatus 20. Fig. 4 is a block diagram showing an example of a hardware configuration for realizing the functions of the optical component healthiness estimation apparatus 20 according to the first embodiment.

[0055] To realize the functions of the optical component health estimation device 20 according to embodiment 1, the configuration can include a processor 400 that performs calculations, a memory 402 that stores programs read by the processor 400, an interface 404 that inputs and outputs signals, and a display 406 that displays the detection results, as shown in FIG. 4 .

[0056] The processor 400 is an example of a computing unit. The processor 400 may be a computing unit called a microprocessor, a microcomputer, a microcontroller, a central processing unit (CPU), or a digital signal processor (DSP). Examples of the memory 402 include non-volatile or volatile semiconductor memory such as random access memory (RAM), read-only memory (ROM), flash memory, erasable programmable read-only memory (EPROM), and electrically programmable read-only memory (EEPROM), as well as a magnetic disk, a flexible disk, an optical disk, a compact disk, a minidisk, and a digital versatile disk (DVD).

[0057] The memory 402 stores a program that executes the functions of the optical component health estimation apparatus 20 according to embodiment 1. The processor 400 exchanges necessary information via the interface 404, executes the program stored in the memory 402, and refers to the data stored in the memory 402, thereby performing the above-described processing. When the optical component health estimation apparatus 20 is configured as an external device that is configured to be able to communicate with the laser processing system 1, the interface 404 is configured to include a communication circuit.

[0058] The calculation results by the processor 400 can be stored in the memory 402. The processing results by the processor 400 can also be displayed on the display 406. The display 406 may be provided outside the optical component health estimation apparatus 20.

[0059] Furthermore, when realizing the functions of the optical component healthiness estimation apparatus 20 according to embodiment 1, the configuration shown in Fig. 5 may be used. Fig. 5 is a block diagram showing another example of a hardware configuration for realizing the functions of the optical component healthiness estimation apparatus 20 according to embodiment 1. In Fig. 5, the processor 400 and memory 402 shown in Fig. 4 are replaced with a processing circuit 403.

[0060] The processing circuit 403 may be a single circuit, a composite circuit, an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), or a combination thereof. Information input to and output from the processing circuit 403 can be exchanged via an interface 404.

[0061] It should be noted that some of the processing in the optical component health assessment device 20 may be performed by the processing circuit 403 , and the processing that is not performed by the processing circuit 403 may be performed by the processor 400 and the memory 402 .

[0062] Second Embodiment In a second embodiment, detailed functions of the parameter estimating unit 21 and the soundness estimating unit 22 described in the first embodiment will be described.

[0063] The parameter estimation unit 21 models the temperature characteristics of the optical component 80 in the laser processing machine 2. If the input to the system to be modeled is considered to be an "oscillation command u(k)" and the output of the system to be modeled is considered to be a "temperature measurement value y(k) by the temperature sensor 82," many general heat transfer problems can be approximated by a first-order lag model G(s) using Laplace transform as shown in the following equation (8).

[0064]

[0065] In the above equation (8), K is a “process gain,” τ is a “time constant,” and s is a “Laplace operator.” The process gain K and the time constant τ are constants determined by the properties of the hardware, and these process gain K and time constant τ are estimated parameters estimated by the parameter estimation unit 21.

[0066] The process gain K is a value that indicates the ratio of the amount of increase in temperature of a member heated by the generation or emission of laser light to the amount of heat generated in the member. The time constant τ is a value that indicates the speed or degree of speed of the temperature response.

[0067] Next, a method for calculating the process gain K and the time constant τ will be described. First, a linear model expressed by the relationship in equation (9) below will be considered.

[0068]

[0069] Regarding Y, Z, and Ψ shown in the above equation (9), Y is called the "output vector," Z is called the "input vector," and Ψ is called the "parameter coefficient vector." ε represents the error in the linear model. To express the first-order lag model G(s) as a linear model, the elements of the output vector Y, input vector Z, and parameter coefficient vector Ψ are expressed by the following equation (10).

[0070]

[0071] In the above formula (10), the superscript dot " ・ " indicates time differentiation. Also, the estimated values ​​are marked with a superscript " ＾ In addition, since equation (10) assumes a linear model in which the steady-state output is zero when the input is zero, the measured temperature value y(k) used for estimation is calculated by subtracting the initial temperature y from the measured temperature value y(k) as in the first equation of equation (10). ０ is removed. Furthermore, although the example shown in equation (10) is a scalar output, which is a one-dimensional output, it may be a vector output depending on the model. The parameter estimation unit 21 determines the process gain K and the time constant τ, which are estimation parameters, so as to reduce the error of the assumed linear model. As a method for determining the estimation parameters, known methods such as the least squares method, Kalman filter, black box optimization, etc. can be used.

[0072] 6 is a block diagram illustrating the function of a health estimation unit 22 provided in an optical component health estimation device 20 according to embodiment 2. As shown in FIG. 6, the health estimation unit 22 includes a temperature prediction unit 221 and an estimation result output unit 222.

[0073] The temperature prediction unit 221 estimates the set of estimated parameters Ψ ＾ and assumed laser beam state u ＾ (k), the predicted temperature value T ＾ ｉｎｆ The assumed laser beam state u ＾(k) is the state of the laser light given from the outside for two or more times k. In this paper, as an example, the temperature that will be reached when the rated power is continuously output to the laser oscillator 40 is predicted. That is, in this paper, the assumed laser light state u ＾ As an example of (k), the temperature to be reached is predicted assuming that the laser oscillator 40 is continuously operated at rated power. Hereinafter, continuous operation at rated power is referred to as "continuous rated output."

[0074] Temperature prediction value T ＾ ｉｎｆ To calculate this, the final value theorem in control engineering can be used. Specifically, the rated power W at the continuous rated output and the initial temperature y ０ For the temperature prediction value T ＾ ｉｎｆ can be expressed as the following equation (11) according to the final value theorem.

[0075]

[0076] In the calculation formula in the above formula (11), "U(s) = 1 / s" is used. Since the continuous rated output corresponds to a "step response" in control engineering, the first-order lag model G(s) is multiplied by "1 / s", which is the Laplace transform formula for the step response.

[0077] The above equation (11) includes the estimated process gain K ＾ is included, but the estimated time constant τ is ＾ Therefore, in the method using the final value theorem in control engineering, the estimated process gain K ＾ and the estimated time constant τ ＾ The estimated parameter set Ψ including ＾ Among these, the process gain estimate K ＾ Only the temperature prediction value T ＾ ｉｎｆ This has the advantage that it can estimate

[0078] The predicted temperature value T ＾ ｉｎｆcan also be obtained using numerical simulation as the system response when any laser light power is given, not just the continuous rated output. When using numerical simulation, the final or maximum value of the temperature response obtained in the numerical simulation can be used as a predicted temperature value to estimate the integrity. Known numerical simulation methods include the Euler method and the Runge-Kutta method.

[0079] The temperature predicted value T predicted by the temperature prediction unit 221 ＾ ｉｎｆ is input to the estimation result output unit 222. The estimation result output unit 222 outputs the temperature prediction value T ＾ ｉｎｆ is the threshold value TH ４ If the threshold value TH exceeds the threshold value TH, it is determined that an abnormality has occurred in the optical component 80, and the determination result is output as the soundness estimation result. ４ The determination by the temperature prediction value T ＾ ｉｎｆ The temperature prediction value T ＾ ｉｎｆ The determination result may be notified to an administrator who is not at the site via a cloud server (not shown).

[0080] The threshold value TH ４ is preferably set lower than the heat resistance temperature of the optical component 80 to be judged. The reason for setting it in this way is that when the laser processing machine 2 outputs the rated power, the predicted temperature value T ＾ ｉｎｆ If the temperature exceeds the heat resistance temperature, the optical component 80 cannot be said to be in good condition.

[0081] FIG. 7 shows simulation results when the optical component health estimation device 20 according to the second embodiment is implemented. The upper part of FIG. 7 shows raw power data based on the oscillation command u(k) as a thick solid line. The change in power is due to the change in the oscillation command u(k). FIG. 7 shows a situation in which a certain arbitrary processing pattern is repeated three times on the laser processing machine 2. The upper part of FIG. 7 also shows the change in power at continuous rated output as a dashed line. The power value at continuous rated output is the power value corresponding to the oscillation command u(k) when the laser oscillator 40 is assumed to be continuously operated at rated power without actually operating the laser oscillator 40. The left vertical axis indicates the power value of the raw data, and the right vertical axis indicates the power value at continuous rated output. As can be seen from a comparison between the values ​​on the left and right vertical axes, the arbitrary processing pattern shown in FIG. 7 is an example of a processing pattern for intermittent, low-power laser processing. In this document, for the sake of simplicity, this processing pattern will be referred to as the "first processing pattern."

[0082] In the lower part of FIG. 7, the raw data of the measurement values ​​measured by the temperature sensor 82 when processing was performed using the first processing pattern is shown by a thick solid line, the measurement data of the cooling water temperature at that time is shown by a thin solid line, and the estimated parameter set Ψ ＾ The thin dashed line shows the response waveform of the estimation model used to calculate , and the thick dashed line shows the response waveform at continuous rated output using the same estimation model. The vertical axis on the left shows the temperature value of the raw data, and the vertical axis on the right shows the temperature value from the response waveform.

[0083] The predicted temperature value T when processing is performed using the first processing pattern ＾ ｉｎｆ The result is shown in the following equation (12).

[0084]

[0085] As shown in the above formula (12), the predicted temperature value T ＾ ｉｎｆThe temperature range of the raw data for processing using the first processing pattern is 23.6 to 24.8°C, as shown by the thick dashed line. Therefore, if the technique of the second embodiment is used, the predicted temperature value T that is predicted to be reached when operating at a continuous rated output of 2.0 kW without actually performing processing can be calculated using only the raw temperature data when intermittent and low-power laser processing is performed. ＾ ｉｎｆ It can be seen that the above can be estimated.

[0086] As described above, in the method using the final value theorem, the process gain estimate K ＾ Only the temperature prediction value T ＾ ｉｎｆ On the other hand, the estimated process gain K ＾ In addition to this, the estimated time constant τ ＾ Using the temperature prediction value T ＾ ｉｎｆ In addition, the estimated parameter set Ψ ＾ The estimated process gain K ＾ and the estimated time constant τ ＾ When other estimation parameters are included, the soundness may be estimated by combining at least two estimation parameters from the plurality of estimation parameters including the other estimation parameters. For the soundness estimation, known methods such as the Mahalanobis-Taguchi System (MT) method, T2 statistics, Q statistics, one-class support vector machine, isolation forest, k-nearest neighbor method, local outlier factor, and self-organization mapping can be used.

[0087] As described above, the optical component health estimation device according to the second embodiment is the same as the optical component health estimation device according to the first embodiment, except that the estimation parameters determined by the parameter estimation unit include a process gain, which is a value indicating the ratio of the amount of increase in temperature of a component heated by the emission of laser light to the amount of heat generated in the component. Use of the process gain allows the model in the parameter estimation unit to be constructed as a highly versatile linear model. Therefore, by further applying the method of the second embodiment to the optical component health estimation device according to the first embodiment, it becomes possible to inexpensively configure the optical component health estimation device while enjoying the effects of the first embodiment.

[0088] Furthermore, according to the optical component health estimation device of the second embodiment, the health estimation unit includes a temperature prediction unit and an estimation result output unit. The temperature prediction unit calculates a predicted temperature value, which is a predicted value of the temperature, based on the estimation parameters and an assumed laser light state, which is a state of the laser light provided from outside at two or more times. The estimation result output unit determines and outputs a health estimation result based on the predicted temperature value and a preset threshold value. The optical component health estimation device of the second embodiment can estimate a predicted temperature value that is predicted to be reached when the laser processing machine is operated at a high-power continuous rated output without actually performing processing, using only raw temperature data obtained when low-power laser processing is performed. Therefore, the optical component health estimation device of the second embodiment can estimate a predicted temperature value without placing a load on the optical components of the laser processing machine.

[0089] Third Embodiment In a third embodiment, a function different from that in the second embodiment or a function added to that in the second embodiment in the health estimation unit 22 described in the first embodiment will be described.

[0090] 8 is a block diagram illustrating the function of the health estimation unit 22 provided in the optical component health estimation apparatus 20 according to embodiment 3. As shown in FIG. 8, the health estimation unit 22 includes a life model determination unit 223 and a life estimation unit 224.

[0091] The life model determination unit 223 receives the estimated parameters Ψ obtained at two or more points in time. ＾and the estimated parameter Ψ ＾ The lifetime model determination unit 223 receives the estimated parameter Ψ ＾ and the time information, the estimated parameter Ψ ＾ A lifespan model is determined to estimate the relationship between the time information and the time information.

[0092] FIG. 9 is a first diagram illustrating the operation of the soundness estimation unit 22 provided in the optical component soundness estimation device 20 according to the third embodiment. In FIG. 9, the filled dots represent estimated parameters obtained in a time series. In the third embodiment, the estimated parameters obtained in a time series are referred to as "p ４ (t)" and the estimated parameter p ４ The time when the estimated parameter p(t) is obtained is defined as "t(i)". ４ (t) is the predicted temperature T ＾ ｉｎｆ Therefore, in the description of FIG. 9, the estimated parameter p ４ (t) is the "temperature prediction value T ＾ ｉｎｆ The horizontal axis in FIG. 9 represents time, and the values ​​on the horizontal axis corresponding to the dots represent the time t(i). The vertical axis represents the estimated parameter p ４ In the example of FIG. 9, it is the temperature.

[0093] The life model determination unit 223 determines the estimated parameter p ４ A regression model expressing the relationship between (t) and time t(i) is determined. The regression model shown in Fig. 9 is an example of a lifespan model determined by the lifespan model determination unit 223. The regression model shown in Fig. 9 is an exponential function model expressed by the following equation (13).

[0094]

[0095] In the above equation (13), φ, α, and β are parameters of the regression model, and are estimated parameters p obtained before the current time A, i.e., in the past. ４ 9 shows the estimated parameter p４ If the regression model is accurately expressed, the estimated parameters p obtained after time A are ４ (t) will be distributed so as to fit on the curve of the regression model.

[0096] The lifespan estimation unit 224 regards time B, when the estimated parameter reaches a preset threshold value, as the lifespan, and determines the time difference between time A and time B as the remaining lifespan. Time A is generally the current time, which is the time when the user wishes to obtain the remaining lifespan, but it may be any other time. In this document, time A will be referred to as the "first time point" as appropriate, and time B will be referred to as the "second time point" as appropriate. The first time point is the time when the optical component 80 has not yet reached its lifespan, and the second time point is the time when the optical component 80 is predicted to reach the end of its lifespan.

[0097] The lifespan model determination unit 223 may determine a lifespan model that includes, as time t(i), a period during which the laser processing machine 2 is not in use. Alternatively, the lifespan model determination unit 223 may determine a lifespan model based on the accumulated time for which laser light is emitted by the laser processing machine 2, taking into account, as time t(i), only a period during which the laser processing machine 2 is in use. In this case, the lifespan estimation unit 224 can calculate, as the remaining lifespan, the time that the laser processing machine 2 can be used until the time when it is predicted that the laser processing machine 2 will reach the end of its lifespan.

[0098] Furthermore, when the remaining life falls below a preset remaining life judgment value, the optical component soundness estimation device 20 may display a warning or the like together with the remaining life information. Furthermore, this displayed information may be notified to a manager who is not on-site, a maintenance person for the laser processing machine 2, or the like, via a cloud server (not shown).

[0099] The following provides additional information about the lifespan model determined by the lifespan model determination unit 223. When the lifespan model is a regression model, the regression model may be a model other than an exponential function model, and any model such as a linear function model, a nonlinear function model, an artificial neural network model, a Gaussian process regression model, or a support vector machine regression model may be applied.

[0100] Furthermore, the lifespan model does not have to be a regression model. The lifespan model may be determined by a method for finding an analytical solution, such as the least squares method, or by a method for finding a solution through iterative calculation. Furthermore, the lifespan model may be a model that has not only the average characteristics of the model but also a confidence interval that includes the variability of the estimation results. In this case, the lifespan estimation unit 224 can estimate the remaining lifespan taking the confidence interval into consideration.

[0101] Fig. 10 is a second diagram illustrating the operation of the health estimation unit 22 provided in the optical component health estimation apparatus 20 according to embodiment 3. Fig. 10 shows the concept of the process for calculating the estimated life and estimated remaining life using a life model called a similar data reference model. The horizontal axis of Fig. 10 represents the operating time, and the vertical axis represents the value of the estimation parameter.

[0102] In Figure 10, the thin solid curves represent a collection of historical time-series data of estimated parameters obtained from a large number of devices (approximately 200 devices). The estimated lifetime calculated from all of this past data is represented by a dashed line drawn parallel to the vertical axis, which is approximately 207 hours in the example shown in Figure 10. Meanwhile, the thick solid curves represent a collection of estimated parameters obtained from only 10 devices selected from a large number of devices similar to the device whose health is to be estimated. The estimated lifetime calculated from only the data on similar devices is represented by a dashed line drawn parallel to the vertical axis, which is approximately 187 hours in the example shown in Figure 10. Also, in Figure 10, the thick dashed line represents data on estimated parameters for currently operating devices, and the time difference between the current operating time and the estimated lifetime calculated from the data on similar devices alone can be calculated as the estimated remaining lifetime.

[0103] As described above, the optical component health estimation device according to the third embodiment is the optical component health estimation device according to the first embodiment, except that the health estimation unit provided in the optical component health estimation device includes a life model determination unit and a life estimation unit. The life model determination unit determines a life model for estimating the relationship between the estimation parameters and the time information based on estimation parameters obtained at two or more times and time information, which is information on the times corresponding to the estimation parameters. The life estimation unit estimates a remaining life, which is the time difference between a first time point, at which the optical component has not yet reached its life, and a second time point, at which the optical component is predicted to reach its life, based on the life model determined by the life model determination unit, and outputs the remaining life as a health estimation result. The optical component health estimation device according to the third embodiment allows a user or administrator to periodically or appropriately know the remaining life of an optical component. Therefore, by further applying the method of the third embodiment to the optical component health estimation device according to the first embodiment, it is possible to reduce the burden of estimating component delivery dates and suppress component inventory while enjoying the effects of the first embodiment.

[0104] Fourth Embodiment In a fourth embodiment, the recursive least squares method is used to estimate the parameter Ψ ＾ The recursive least squares method determines the estimated parameter Ψ at time k. ＾ This is a method of sequentially calculating the estimated parameter Ψ ＾ The calculation formula can be expressed by the following formula (14).

[0105]

[0106] Ψ shown in the above formula (14) ＾ , Z, Y are the parameter coefficient vector, input vector, and output vector shown in the above formula (10). (k) means the measured data or calculated value at time k, and (k-1) means the measured data or calculated value at time k-1, which is one sampling period before time k. Furthermore, P(k) shown in formula (14) is the covariance matrix. In the recursive least squares method, the recursive calculation shown in formula (14) is repeated for each sampling period.

[0107] 11 is a functional block diagram of a case where the parameter estimation unit 21 provided in the optical component health estimation apparatus 20 according to embodiment 4 is realized using the recursive least squares method. As shown in FIG. 11 , the parameter estimation unit 21 includes a differentiator 210, a differential processing unit 211, a vector connection unit 212, a parameter coefficient vector update unit 213, a covariance matrix update unit 214, delay units 215 and 216, and a vector extraction unit 217.

[0108] The differentiator 210 generates an output vector Y(k). The differential processing unit 211 generates a differential value y~(k) of the temperature measurement value y(k) by, for example, pseudo-differential processing. The vector connection unit 212 connects the oscillation command u(k) and the differential value y~(k) to generate an input vector Z(k). The parameter coefficient vector update unit 213 updates the parameter coefficient vector Ψ ＾ (k) is updated. The parameter coefficient vector Ψ ＾ (k) is updated using the input vector Z(k), the output vector Y(k), the covariance matrix P(k-1) of the previous sampling time, and the parameter coefficient vector Ψ ＾ (k-1). In the covariance matrix update unit 214, the covariance matrix P(k) is updated. The covariance matrix P(k) is updated using the input vector Z(k) and the covariance matrix P(k-1) of one sampling time before. In the delay unit 215, the parameter coefficient vector Ψ ＾ The delay unit 216 generates the covariance matrix P(k-1) for the previous sampling time. The vector extraction unit 217 extracts the parameter coefficient vector Ψ ＾ From (k), the process gain estimate K ＾ (k) and the estimated time constant τ ＾ (k) are extracted and output. The generated process gain estimate K ＾ (k) and the estimated time constant τ ＾ (k) is input to the health estimation unit 22.

[0109] Alternatively, a forgetting recursive least squares method may be used, which introduces a forgetting factor λ, which is an element for gradually forgetting past states. The calculation formula for the forgetting recursive least squares method can be expressed by the following formula (15).

[0110]

[0111] The forgetting factor λ is a constant set in advance, and is often set to a value close to but less than 1, such as 0.99. The smaller the forgetting factor λ, the easier it is to forget data from older samples, and the quicker the response of the estimated value to true parameter changes becomes, but the more susceptible it is to noise. For this reason, when implementing the forgetting recursive least squares method, an estimated response frequency wid is introduced as a parameter set by the user to adjust the temporal speed of the estimation. The estimated response frequency wid can be expressed by the following equation (16):

[0112]

[0113] Sampling period T ｓ When the estimated response frequency wid is 0.1 rad / s when the time t is 100 ms, the forgetting factor λ is calculated as shown in the following equation (17).

[0114]

[0115] In order to prevent a decrease in the accuracy of the calculation results, the estimated response frequency wid should be set to a value within the range of approximately 2 to 5 times the reciprocal of the total time during which the data used for analysis was acquired.

[0116] Fig. 12 is a diagram showing the results of a simulation when the optical component health assessment device 20 according to embodiment 4 is implemented. The upper diagram of Fig. 12 is the same as the upper diagram of Fig. 7, and the central diagram of Fig. 12 is the same as the lower diagram of Fig. 7. Therefore, the simulation conditions are the same as those of embodiment 2.

[0117] 12 shows the process gain estimate K ＾ The waveform of (k) is shown by a thin solid line, and the estimated time constant value τ ＾As can be seen from the waveforms in the lower diagram, the process gain estimate K ＾ (k) and the estimated time constant τ ＾ Regarding (k), good results have been obtained.

[0118] Other considerations when using the recursive least squares method will be described below. As described above, the method of the fourth embodiment has two internal variables, the parameter coefficient vector Ψ and the covariance matrix P, in order to accumulate and store information on past measurement data. The initial value Ψ(0) of the parameter coefficient vector Ψ may basically be a zero vector, but if a rough estimate can be predicted in advance, the rough estimate may be set as the initial value Ψ(0). For example, if the process gain estimate K ＾ is 0.005 [°C / W], and the estimated time constant τ ＾ When it is assumed that the value of Ψ(0) is about 20 seconds, the initial value Ψ(0) can be set as shown in the following equation (18).

[0119]

[0120] Furthermore, the reason why a rough estimate can be set to the initial value Ψ(0) is because the recursive least squares method is a method for calculating an estimate in an updated manner. Therefore, by using the recursive least squares method, parameters can be estimated without storing time-series measurement data over a long period of time. This makes it possible to estimate the state of the optical component 80 with high accuracy without using a processor with high calculation power or a memory with a large storage capacity.

[0121] As described above, the optical component health estimation device according to embodiment 4 is the same as the optical component health estimation device according to embodiment 1, except that the parameter estimation unit provided in the optical component health estimation device determines estimation parameters using the recursive least squares method. The optical component health estimation device according to embodiment 4 makes it possible to accurately and adaptively estimate the state of optical components without using a processor with high calculation power or a memory with large storage capacity, while enjoying the effects of embodiment 1.

[0122] Fifth Embodiment In a fifth embodiment, an improved configuration of the optical component soundness estimation device according to the fourth embodiment will be described.

[0123] FIG. 13 shows simulation results used to explain points to consider when implementing the optical component health estimation device 20 according to embodiment 5. In the upper part of FIG. 13 , raw power data based on the oscillation command u(k) is shown by a thick solid line, and the change in power at continuous rated output is shown by a dashed line. The meanings of the raw data and continuous rated output are as explained in FIG. 7 . However, the processing pattern shown in FIG. 13 is different from the processing pattern shown in FIG. 7 . While FIG. 7 shows an example in which processing using a certain arbitrary processing pattern is repeated three times, FIG. 13 shows an example in which processing using a certain arbitrary processing pattern is performed only once. In this paper, for the sake of simplicity, the processing pattern shown in FIG. 13 will be referred to as the "second processing pattern."

[0124] In the center of FIG. 13, the raw data of the measurement values ​​measured by the temperature sensor 82 when processing was performed using the second processing pattern is shown by a thick solid line, the measurement data of the cooling water temperature at that time is shown by a thin solid line, and the estimated parameter set Ψ ＾ The response waveform of the estimation model for calculating is shown by the thin dashed line, and the response waveform at continuous rated output using the same estimation model is shown by the thick dashed line.

[0125] 13 shows the process gain estimate K extracted by the vector extraction unit 217 of the fourth embodiment shown in FIG. ＾ The waveform of the time constant estimated value τ extracted by the vector extraction unit 217 of the fourth embodiment is shown by a thin solid line. ＾ The waveform of the temperature is shown by a thick solid line. The characteristic feature of the results shown in FIG. 13 is that the predicted values ​​of the reached temperature diverge (see the waveforms of the thin and thick dashed lines in the center of the figure), and the estimated time constant τ ＾ The problem is that the value of is negative.

[0126] In FIG. 13, the estimated time constant τ ＾The reason why the value of is negative is thought to be that the machining time by the second machining pattern is short and the temperature change is small. In addition to the short machining time, the cooling water temperature is gradually decreasing, which may have combined to cause the data used in Figure 13 to be unable to be represented by the assumed linear model. Since the time constant τ cannot be a negative value due to physical characteristics, the estimated parameter Ψ obtained at this time ＾ The accuracy of the health estimation results obtained using this method may be poor.

[0127] In order to avoid such a phenomenon, in the fifth embodiment, the estimated parameter Ψ ＾ Specifically, we consider limiting the range of the estimated parameter Ψ during the sequential calculation process. ＾ A process is performed in which a lower limit is set for the following equation (19).

[0128]

[0129] In the above formula (19), K ｍｉｎ and τ ｍｉｎ are the respective lower limit values ​​that are set. If the value of the time constant τ becomes zero or less during the calculation, the calculation results may diverge, so the respective lower limit values ​​are set to positive values ​​close to zero. For example, K ｍｉｎ = 1 x 10 －４ [℃ / W] = 0.1 [m℃ / W], τ ｍｉｎ = 1.0 [s].

[0130] 14 is a functional block diagram showing the configuration of a parameter estimation unit 21 provided in an optical component health estimation device 20 according to embodiment 5. Components that are the same as or equivalent to those in FIG. 11 are given the same reference numerals. In the parameter estimation unit 21 shown in FIG. 14, an estimated parameter Ψ is inserted between a parameter coefficient vector update unit 213 and a delay unit 215. ＾ A limiter 218 is provided to limit the value of (k). Note that the differentiator 210, the differential processing unit 211, and the vector connection unit 212 are not shown in FIG.

[0131] Fig. 15 is a diagram showing the results of a simulation when the optical component health assessment device 20 according to embodiment 5 is implemented. The upper diagram in Fig. 15 is the same as the upper diagram in Fig. 13, and the simulation conditions are the same as those in Fig. 13. What is distinctive about the results in Fig. 15 is that, as shown by the waveforms of the thin dashed line and the thick dashed line in the central diagram, divergence of the predicted value of the reached temperature can be avoided. The reason for this is that the estimated parameter Ψ ＾ This is because the limiter 218 is provided to limit the range of the time constant estimate τ ＾ The lower limit of the temperature is limited to a set lower limit of 0.1 [m°C / W].

[0132] As described above, the optical component health estimation device according to embodiment 5 is the same as the optical component health estimation device according to embodiment 4, except that the parameter estimation unit provided in the optical component health estimation device is provided with a limiter that limits the value of the estimation parameter. This limiter acts to prevent the predicted value of the reached temperature from diverging, so that even when the number of measurement data is small, it is possible to suppress a decrease in estimation accuracy. Therefore, the optical component health estimation device according to embodiment 5 makes it possible to suppress a decrease in the accuracy of estimating the state of an optical component while enjoying the effects of embodiments 1 to 4.

[0133] Embodiment 6 Fig. 16 is a block diagram illustrating the function of an optical component healthiness estimation device 20 according to embodiment 6. As shown in Fig. 16, the optical component healthiness estimation device 20 includes an estimation validity determination unit 24 and an estimation validity determination unit 25.

[0134] The oscillation command u(k) is input to the estimation effectiveness determination unit 24. The estimation effectiveness determination unit 24 determines the cumulative value of the oscillation command u(k) as the estimated effectiveness and outputs it to the estimation effectiveness determination unit 25. Specifically, the estimation effectiveness determination unit 24 determines the relationship between the oscillation command u(k) and the sampling period T ｓ By integrating the product of and , the cumulative value E of the oscillation command u(k) is calculated.

[0135]

[0136] In the above equation (20), N is the number of data points, that is, the number of oscillation commands u(k) to be integrated. ｓ The product of these represents the energy of the laser light, and its unit is called "watt-second" and is expressed as [Ws]. Therefore, the cumulative value E of the oscillation command u(k) is also the cumulative value of the laser power.

[0137] Although the above equation (20) is an equation in which a definite integral is expressed by a rectangular numerical integral, a more accurate calculation method such as a numerical integral using a trapezoidal approximation may be used. Also, instead of the cumulative value of the oscillation command u(k), information obtained by calculating the cumulative value of the time during which the laser beam is output, the integrated value of the amount of heat generated by the laser beam, the error between the predicted temperature value and the actual temperature, the magnitude of variation in the measured temperature value, etc. may be output to the estimation validity determination unit 25.

[0138] The estimation validity determination unit 25 receives the cumulative value of the oscillation command u(k) as the estimated validity determined by the estimation validity determination unit 24. The estimation validity determination unit 25 determines the validity of the health estimation result by comparing the cumulative value of the oscillation command u(k) calculated since the laser processing machine 2 started operating with a pre-set estimation validity determination threshold. Specifically, if the cumulative value of the oscillation command u(k) is smaller than the estimation validity determination threshold, the estimation validity determination unit 25 determines that the validity of the health estimation result is low. Furthermore, if the cumulative value of the oscillation command u(k) is equal to or greater than the estimation validity determination threshold, the estimation validity determination unit 25 determines that the validity of the health estimation result is high.

[0139] The above-mentioned "after the laser processing machine 2 starts to operate" may mean, for example, "after the user turns on the power of the laser processing machine 2," "after the parameter estimation unit 21 performs initialization processing of the internal variables," "after the parameter estimation unit 21 starts to acquire the oscillation command u(k) or the temperature measurement value y(k) used to determine the estimated parameters," or "after the temperature prediction unit 221 acquires the temperature prediction value T ＾ ｉｎｆ This can be rephrased as "after starting the calculation of

[0140] 17 is a diagram showing a simulation result for explaining points to consider when implementing the optical component health estimation device 20 according to the sixth embodiment. ＾ ｉｎｆ 17 shows the relationship between the error of the temperature prediction value T ＾ ｉｎｆ As shown in the dashed ellipse, when the cumulative value of the oscillation command u(k) is relatively large, the error of the temperature prediction value T ＾ ｉｎｆ The area enclosed by the solid ellipse is the temperature prediction value T when data from a long-term operation at maximum output is also used. ＾ ｉｎｆ In the example of FIG. 17, the cumulative value of the oscillation command u(k) is extremely large, about 4.2×10 ６ [W / s]. When the cumulative value of the oscillation command u(k) is extremely large, as in this case, the obtained temperature prediction value T ＾ ｉｎｆ can be estimated as the true temperature reached.

[0141] In the example of FIG. 17, the estimated validity determination threshold is set to 0.5×10 ６ [W / s]. By setting in this way, it is possible to exclude conditions where the cumulative value of the oscillation command u(k) is small, and thereby it is possible to extract conditions where the error in the estimation result is small. As a result, it is possible to improve the estimation accuracy of the health estimation result. Information regarding the validity of the estimation result is output to the outside of the optical component health estimation device 20. In this way, the information regarding the validity of the estimation result can be used as an index indicating whether the health estimation result of the optical component 80 has been determined with high accuracy.

[0142] 16, information indicating the validity of the estimation result is output to the outside, but it goes without saying that this information can be used inside the optical component healthiness estimation device 20. For example, if the validity of the estimation result is negative, the estimation process is performed excluding time-series data for which the validity is negative. By excluding time-series data for which the validity is negative, it is possible to suppress a decrease in estimation accuracy.

[0143] The method of the sixth embodiment aims to exclude time series data that may cause a decrease in estimation accuracy, and any method that meets this aim may be used. For example, it is sufficient to exclude time series data when the machining power is low, time series data when an operation with an extremely short machining time is performed for a short period of time, time series data when the change in the measured temperature is small, etc.

[0144] As described above, the optical component health estimation device according to the sixth embodiment includes an estimation validity determining unit and an estimation validity determining unit in addition to the optical component health estimation devices according to the first to fifth embodiments. The estimation validity determining unit determines, as the estimation validity, at least one of the cumulative value of oscillation commands, the cumulative value of the time during which laser light is output, the integrated value of the amount of heat generated by the laser light, the error between a predicted temperature value and an actual temperature, and the magnitude of variation in the measured temperature values. The estimation validity determining unit determines the validity of the health estimation result based on the estimation validity and a predetermined threshold value, i.e., an estimation validity determination threshold. According to the optical component health estimation device according to the sixth embodiment, information regarding the validity of the estimation result can be used as an index indicating whether the health estimation result of the optical component has been determined with high accuracy. Furthermore, according to the optical component health estimation device according to the sixth embodiment, if the validity of the estimation result is negative, the estimation process can be performed while excluding time-series data for which the validity is negative. Therefore, according to the optical component healthiness estimation device of embodiment 6, it is possible to enjoy the effects of embodiments 1 to 5 while further improving the accuracy of estimating the healthiness of the optical component 80.

[0145] The configurations shown in the above embodiments are merely examples, and may be combined with other known technologies, or different embodiments may be combined with each other. It is also possible to omit or modify parts of the configurations as long as they do not deviate from the gist of the invention.

[0146] 1 Laser processing system, 2 Laser processing machine, 10 Control device, 12 Oscillator control unit, 14 Calculation unit, 16 Processing machine control unit, 20 Optical component soundness estimation device, 21 Parameter estimation unit, 22 Soundness estimation unit, 24 Estimation effectiveness determination unit, 25 Estimation effectiveness determination unit, 40 Laser oscillator, 41, 41-1, 41-2, 41-3, 41-4 Laser module, 42 Combiner unit, 43, 43-1, 43-2, 43-3, 43-4 Drive power supply, 50 Optical fiber, 60 Processing head, 65 Relative movement unit, 70 Work support unit, 72 Work, 80 Optical component, 82 Temperature sensor, 210 Differential unit, 211 Differential processing unit, 212 Vector connection unit, 213 Parameter coefficient vector update unit, 214 Covariance matrix update unit, 215, 216 Delay unit, 217 Vector extraction unit, 218 limiter, 221 temperature prediction unit, 222 estimation result output unit, 223 lifetime model determination unit, 224 lifetime estimation unit, 400 processor, 402 memory, 403 processing circuit, 404 interface, 406 display unit, 411 semiconductor laser, 412 excitation combiner.

Claims

1. An optical component health assessment device comprising: a parameter assessment unit that determines an estimation parameter, which is an estimate of a parameter that represents the relationship between the state of the laser light and the temperature, based on either a laser light emission command to a laser oscillator or a measurement value of the laser light in the laser oscillator, and a measurement value of the temperature of a component heated by the generation or emission of laser light; and a health assessment unit that estimates the health of an optical component based on the estimation parameter and outputs a health assessment result that represents the degree of the health.

2. The optical component health assessment device according to claim 1, characterized in that the estimation parameters include a process gain, which is a value indicating the ratio of the amount of increase in temperature of a component heated by the generation or emission of the laser light to the amount of heat generated in the component.

3. The optical component soundness estimation device according to claim 2, wherein the estimation parameters further include a time constant that is a value indicating the speed of response to the temperature.

4. An optical component health estimation device as described in any one of claims 1 to 3, characterized in that the health estimation unit comprises: a temperature prediction unit that calculates a temperature prediction value, which is a predicted value of the temperature, based on the estimation parameters and an assumed laser light state, which is a state of the laser light given from outside for two or more times; and an estimation result output unit that determines and outputs the health estimation result based on the temperature prediction value and a predetermined threshold value.

5. The optical component health estimation device described in any one of claims 1 to 4, characterized in that the health estimation unit comprises: a life model determination unit that determines a life model for estimating the relationship between the estimated parameters obtained at two or more points in time and time information that is information on the time corresponding to the estimated parameters; and a life estimation unit that estimates a remaining life, which is the time difference from a first time point when the optical component has not yet reached its life, to a second time point when the optical component is predicted to reach the end of its life, based on the life model, and outputs the remaining life as the health estimation result.

6. An optical component soundness estimation device according to any one of claims 1 to 5, characterized in that the parameter estimation unit determines the estimation parameters using a recursive least squares method.

7. The optical component soundness estimation device according to claim 6, wherein the parameter estimation section is provided with a limiter that limits the value of the estimated parameter.

8. An optical component healthiness estimation device according to any one of claims 1 to 7, characterized in that it comprises an estimation effectiveness determination unit that determines as an estimated effectiveness at least one of the cumulative value of the oscillation command, the cumulative value of the time for which the laser light is output, the integrated value of the power of the laser light, the error between the predicted temperature value that is the predicted value of the temperature and the actual temperature, and the magnitude of variation in the measured temperature value; and an estimation effectiveness determination unit that determines the validity of the healthiness estimation result based on the estimation effectiveness and an estimation effectiveness determination threshold that is a preset threshold value.

9. A laser processing system comprising: a laser processing machine that processes a workpiece with laser light; a control device that controls the laser processing machine; and an optical component healthiness estimation device that estimates the healthiness of optical components, wherein the optical component healthiness estimation device comprises: a parameter estimation unit that determines an estimation parameter that is an estimate of a parameter that represents the relationship between the state of the laser light and the temperature, based on either a laser light emission command for a laser oscillator or a measurement value of the laser light in the laser oscillator, and a measurement value of the temperature of a member heated by the generation or emission of laser light; and a healthiness estimation unit that estimates the healthiness of optical components based on the estimation parameter and outputs a healthiness estimation result that represents the degree of the healthiness.

10. A method for estimating the soundness of optical components provided in a laser processing system, comprising: a first step of determining an estimation parameter, which is an estimate of a parameter representing the relationship between the state of the laser light and the temperature, based on either a laser light oscillation command given to a laser oscillator or a measurement value of the laser light in the laser oscillator, and a measurement value of the temperature of a member heated by the generation or emission of the laser light; and a second step of estimating the soundness of the optical component based on the estimation parameter, and outputting a soundness estimation result representing the degree of the soundness.

Images