Model creation method, input waveform generation method, and model creation device

Abstract

This model creation method is a method for creating a model that reflects the output characteristics of an optical amplifier. The method includes: an excitation step of irradiating a laser medium of an optical amplifier with excitation light; an amplification step of inputting input light into the laser medium irradiated with the excitation light, amplifying the input light, and causing output light to emit from the laser medium; a detection step of detecting an output waveform of the output light emitted in the amplification step; and a model creation step of creating a model on the basis of a time integration value of the output waveform detected in the detection step and an amplification factor of the output waveform relative to the input waveform of the input light input in the amplification step.

Description

Model creation method, input waveform generation method and model creation device

[0001] The present disclosure relates to a model creation method, an input waveform generation method, and a model creation device.

[0002] Optical amplifiers are known that input light to a laser medium, amplify it, and emit output light. Patent Document 1, for example, describes a laser characteristic prediction device that predicts the characteristics of output light relative to an injection current as a technology for predicting the output characteristics of such an optical amplifier. The laser characteristic prediction device described in Patent Document 1 repeatedly predicts the intensity of excitation light irradiated onto the laser medium, the intensity of output light, and the absorption coefficient of the laser medium, and predicts the predicted output light intensity when the absorption coefficient converges within a predetermined range as the characteristics of the output light relative to the injection current to be predicted.

[0003] Japanese Patent Application Laid-Open No. 2008-211085

[0004] In laser processing using the output light of an optical amplifier, it has been found that, for example, the processing rate, processing quality, and processing efficiency depend on the waveform shape of the output light. Therefore, it is desirable to be able to accurately predict the output characteristics of an optical amplifier in order to find the optimal waveform shape of the output waveform. In this regard, the above-mentioned techniques do not sufficiently take into account, for example, individual differences and aging, making it difficult to accurately predict the output characteristics of an optical amplifier. Furthermore, the above-mentioned techniques require various predictions to be performed repeatedly, and the required calculations are complicated.

[0005] Therefore, an object of the present disclosure is to provide a model creation method, an input waveform generation method, and a model creation device that are capable of predicting the output characteristics of an optical amplifier easily and accurately.

[0006] The present inventors have focused on the fact that there is a certain relationship between the amplification factor of an output waveform with respect to the input waveform of input light (hereinafter also simply referred to as "amplification factor") and the time integral value of the output waveform, and have found that by utilizing this relationship, it is possible to create a model that reflects the output characteristics of an optical amplifier from actual data that reflects, for example, individual differences and aging changes, without repeating various predictions. The present disclosure has been made based on this finding.

[0007] That is, the model creation method of the present disclosure is [1] "a method for creating a model that reflects the output characteristics of an optical amplifier, comprising: an excitation step of irradiating excitation light onto a laser medium of the optical amplifier; an amplification step of inputting input light into the laser medium irradiated with the excitation light, amplifying the input light, and emitting output light from the laser medium; a detection step of detecting an output waveform of the output light emitted in the amplification step; and a model creation step of creating the model based on a time integral value of the output waveform detected in the detection step and an amplification factor of the output waveform relative to the input waveform of the input light input in the amplification step."

[0008] This model creation method utilizes the fact that there is a fixed relationship between the amplification factor and the time integral value of the output waveform, making it possible to create a model that reflects the output characteristics of an optical amplifier without repeatedly making various predictions from actual data, thereby making it possible to predict the output characteristics of an optical amplifier easily and accurately.

[0009] The model creation method of the present disclosure may be [2] "the model creation method according to [1], wherein the model is a function model expressing parameters related to the amplification factor, and the function model includes a function related to energy accumulated in the laser medium by irradiation with the pumping light and a function related to energy consumed by amplification in the laser medium." In this case, a function model that takes into account the energy accumulated in the laser medium by irradiation with the pumping light and the energy consumed by amplification in the laser medium can be created as a model that reflects the output characteristics of the optical amplifier.

[0010] The model creation method of the present disclosure may be [3] "the model creation method according to [2]," in which the function model includes a linear term of the time integral value. In this case, the linear term of the time integral value can be used as the function model.

[0011] The model creation method of the present disclosure may be [4] "the model creation method according to [2] or [3], in which the function model includes a nonlinear term of the time integral value." In this case, the nonlinear term of the time integral value can be used as the function model.

[0012] The model creation method of the present disclosure may be [5] "the model creation method according to [1]," wherein the model has a lookup table in which the time integral value and a parameter related to the amplification factor are associated with each other. In this case, the lookup table can be used to create a model that reflects the output characteristics of the optical amplifier.

[0013] The model creation method of the present disclosure may be [6] "the model creation method according to any one of [1] to [5], wherein, in the amplifying step, the input light is input to the optical amplifier and amplified for each of a plurality of repetition rates different from each other, and the output light is emitted from the optical amplifier; in the detecting step, an output waveform of the output light is detected for each of the plurality of repetition rates; and in the model creation step, the model is created based on the time integral value and the amplification factor for each of the plurality of repetition rates." In this case, even when the repetition rates of the input light are different, it is possible to create a model that accurately reflects the output characteristics of the optical amplifier.

[0014] The model creation method of the present disclosure may be [7] "the model creation method according to any one of [1] to [6], wherein the input waveform of the input light input in the amplification step is a waveform whose first-order time derivative is continuous." In this case, the time integral value of the output waveform becomes continuous, making it possible to create a model that reflects the output characteristics of the optical amplifier with high accuracy.

[0015] The input waveform generation method of the present disclosure is [8] "a method for generating an input waveform of the input light to be input to the optical amplifier using the model created by the model creation method described in any one of [1] to [7], the method comprising an input waveform generation step of generating the input waveform that realizes the target waveform based on the target waveform and the model created in the model creation step."

[0016] This input waveform generation method can generate an input waveform using a model that can accurately predict the output characteristics of an optical amplifier, thereby enabling the generation of an input waveform that accurately achieves a target waveform.

[0017] The input waveform generation method of the present disclosure may be [9] "the model generation method according to [8], wherein the model generation method includes a storage step of storing the created model, and in the input waveform generation step, each time the target waveform is changed, the model stored in the storage step is read out, and the input waveform that realizes the target waveform is generated based on the target waveform and the model." In this case, by storing the model, it is possible to generate an appropriate input waveform even when the target waveform is changed.

[0018] The model creation device of the present disclosure is

[10] "a device for creating a model that reflects the output characteristics of an optical amplifier, comprising: an excitation unit that irradiates excitation light onto a laser medium of the optical amplifier; an amplification unit that inputs input light to the laser medium irradiated with the excitation light, amplifies the input light, and emits output light from the laser medium; a detection unit that detects the output waveform of the output light emitted by the amplification unit; and a model creation unit that creates the model based on the time integral value of the output waveform detected by the detection unit and the amplification factor of the output waveform relative to the input waveform of the input light input to the amplification unit."

[0019] This model creation device also utilizes the fact that there is a fixed relationship between the amplification factor and the time integral value of the output waveform, making it possible to create a model that reflects the output characteristics of an optical amplifier without repeatedly making various predictions from actual data, thereby making it possible to predict the output characteristics of an optical amplifier easily and accurately.

[0020] According to the present disclosure, it is possible to provide a model creation method, an input waveform generation method, and a model creation device that are capable of predicting the output characteristics of an optical amplifier easily and accurately.

[0021] FIG. 1 is a block diagram showing the configuration of a laser device according to an embodiment. FIG. 2(a) is a schematic diagram showing a laser medium. FIG. 2(b) is a schematic diagram showing a laser medium during amplification. FIG. 3 is a graph showing consumed energy and stored energy. FIG. 4 is a flowchart showing a model creation method according to an embodiment. FIG. 5(a) is a diagram showing an input waveform used for creating a model. FIG. 5(b) is a diagram showing an output waveform of detected output light. FIG. 6(a) is a diagram showing actual measurement data related to the time integral value of the output waveform and the amplification factor. FIG. 6(b) is a diagram showing a fitting function for the actual measurement data of FIG. 6(a). FIG. 7 is another block diagram showing the configuration of a laser device according to an embodiment. FIG. 8(a) is a diagram showing an example of a target waveform. FIG. 8(b) is a diagram showing the result of applying a model to the target waveform. FIG. 9(a) is a diagram showing a corrected input waveform. FIG. 9(b) is a diagram showing an output waveform. FIG. 10 is a flowchart showing a method for generating an input waveform according to an embodiment. FIG. 11(a) is a graph showing an output waveform when the target waveform is a uniform intensity triangular wave. FIG. 11(b) is another graph showing an output waveform when the target waveform is a uniform intensity triangular wave. FIG. 12(a) is a graph showing an output waveform when the target waveform is a damped intensity triangular wave. FIG. 12(b) is another graph showing an output waveform when the target waveform is a damped intensity triangular wave. FIG. 13(a) is a graph showing an output waveform when the target waveform is a uniform intensity Gaussian pulse. FIG. 13(b) is another graph showing an output waveform when the target waveform is a uniform intensity Gaussian pulse. FIG. 14(a) is a graph showing an output waveform when the target waveform is a damped intensity Gaussian pulse. FIG. 14(b) is another graph showing an output waveform when the target waveform is a damped intensity Gaussian pulse. FIG. 15(a) is a graph showing an output waveform when the target waveform is Gaussian. FIG. 15(b) is another graph showing an output waveform when the target waveform is Gaussian. FIG. 16(a) is a graph showing an output waveform when the target waveform is a triangular wave. Fig. 16(b) is another graph showing an output waveform when the target waveform is a triangular wave Fig. 17 is a block diagram showing the configuration of a laser device according to a modified example.

[0022] Hereinafter, embodiments will be described in detail with reference to the drawings. In the following description, the same or equivalent elements will be designated by the same reference numerals, and redundant description will be omitted.

[0023] As shown in FIG. 1, the laser device 1 of this embodiment includes a computer 10, a driver circuit 2, a semiconductor laser element 3, an optical isolator 4, an optical amplifier 5, an optical branching unit 6, and a detector 7.

[0024] The computer 10 includes, for example, a processor (CPU) and storage media such as RAM and ROM. The computer 10 is connected to the driver circuit 2 and provides the driver circuit 2 with waveform data relating to the waveform of the optical pulse to be emitted from the semiconductor laser element 3. The driver circuit 2 generates a drive current having a time waveform corresponding to the waveform data input from the computer 10 and supplies the generated drive current to the semiconductor laser element 3.

[0025] The semiconductor laser element 3 emits input light La based on the drive current supplied from the driver circuit 2. The input light La is light before amplification by the optical amplifier 5, and has an input waveform corresponding to waveform data input from the computer 10. The input light La is pulsed laser light input to the optical amplifier 5. For example, the semiconductor laser element 3 is a distributed feedback (DFB) laser diode. The output of the semiconductor laser element 3 is, for example, several nanojoules. The input waveform is a pulse waveform. The input waveform is, for example, a current signal corresponding to the optical intensity of the input light La. The input waveform is, for example, a time waveform of current represented on a graph with the horizontal axis being time (nsec) and the vertical axis being current (mA).

[0026] The optical isolator 4 is interposed on the optical path between the semiconductor laser element 3 and the optical amplifier 5. The optical isolator 4 prevents light amplified by the optical amplifier 5 from returning to the semiconductor laser element 3. The optical amplifier 5 amplifies the input light La output from the semiconductor laser element 3. The optical amplifier 5 directly amplifies the input light La as light without converting it into an electrical signal. The optical amplifier 5 may be configured, for example, by a solid-state laser amplifier. The solid-state laser amplifier may be configured, for example, by glass or yttrium aluminum garnet (YAG) doped with an impurity such as Nd. The optical amplifier 5 has a laser medium 51 to which the input light La is input. The laser medium 51 is irradiated with excitation light L0 from an excitation light source 8. In the optical amplifier 5, the excitation light L0 is input to the laser medium 51 together with the input light La, thereby amplifying the input light La, and the amplified input light La is output from the laser medium 51 as output light Lb.

[0027] The optical branching unit 6 branches the output light Lb by reflecting or transmitting a portion of the output light Lb emitted from the optical amplifier 5 and transmitting or reflecting the remainder. The optical branching unit 6 may be composed of, for example, a glass plate. The detector 7 detects the output waveform of the output light Lb. For example, the detector 7 may be composed of a photodiode that receives a portion of the output light Lb and a circuit (such as a waveform measuring device or an oscilloscope) that converts the photocurrent flowing through the photodiode into a voltage signal. The detector 7 provides the output waveform of the output light Lb to the computer 10. The output waveform is a pulse waveform. For example, the output waveform is a voltage signal corresponding to the light intensity of the output light Lb. For example, the output waveform is a time waveform of voltage represented on a graph with the horizontal axis being time (nsec) and the vertical axis being voltage (mV).

[0028] In the laser device 1 of this embodiment, a model reflecting the output characteristics of the optical amplifier 5 is created. The model models the laser amplification phenomenon. The model is a prediction model for correcting distortion of the pulse waveform caused by the optical amplifier 5. The computer 10 has an input unit 11, an input waveform generation unit 12, a model creation unit 13, and a storage unit 14.

[0029] The input unit 11 accepts input of various types of information. The input unit 11 is, for example, a device that accepts user operation input, such as a mouse or a keyboard. The input unit 11 may be commonly configured, for example, by a touch panel, and in this case, may be configured by a GUI (Graphical User Interface). The input unit 11 may accept various types of input for generating an input waveform for creating a model.

[0030] The input waveform generation unit 12 generates (shapes) an input waveform for model creation in accordance with various inputs received by the input unit 11. The input waveform for model creation is a waveform having a continuous first-order derivative with respect to time, as shown in FIG. 5A, for example. As an example, the input waveform for model creation may be a waveform having a period in which it rises sharply and then becomes a constant value. The input waveform for model creation may include, for example, a square wave (rectangular wave). The input waveform for model creation is not particularly limited and may include, for example, a triangular wave or a Gaussian pulse. The input waveform for model creation may be a waveform having a continuous first-order derivative with respect to time, excluding the rising and / or falling portions. The input waveform generation unit 12 provides waveform data related to the generated input waveform to the driver circuit 2.

[0031] The model creation unit 13 creates a model based on the time integral value and amplification factor of the output waveform detected by the detector 7. Specifically, the model creation unit 13 calculates the time integral value of the output waveform detected by the detector 7. The time integral value corresponds to the output energy of the output light Lb. The model creation unit 13 calculates a parameter related to the amplification factor of the output waveform detected by the detector 7. The amplification factor is the amplification factor of the output waveform detected by the detector 7 with respect to the input waveform for model creation generated by the input waveform generation unit 12. The parameter related to the amplification factor in this embodiment is 1 / amplification factor (the reciprocal of the amplification factor), which corresponds to input waveform / output waveform (the value obtained by dividing the input waveform by the output waveform). The time integral value and amplification factor of the output waveform correspond to the actual data detected by the detector 7.

[0032] The model creation unit 13 generates a function model representing parameters related to the amplification factor as a model. The model creation unit 13 performs fitting based on the time integral value of the output waveform detected by the detector 7 and the amplification factor to generate the function model. Specifically, the model creation unit 13 can obtain the function model by fitting measured data on a graph in which the horizontal axis represents the time integral value of the output waveform and the vertical axis represents 1 / amplification factor (=input waveform / output waveform). The fitting method is not particularly limited, and various methods may be used. The memory unit 14 stores the model MD (see FIG. 7) created by the model creation unit 13.

[0033] Here, the function model is composed of a fitting function including a function f(t) relating to the stored energy (recovery energy) stored in the laser medium 51 by irradiation with the excitation light L0, and a function F(x) relating to the energy consumed by amplification in the laser medium 51. The fitting function will be explained below. Note that "t" is time, and "x" is the time integral value of the output waveform. Function model = F(x) + f(t)

[0034] 2(a) and 2(b) are schematic diagrams showing a laser medium 51. As shown in FIGS. 2(a) and 2(b), energy accumulates in the laser medium 51 due to irradiation with excitation light L0. The stored energy or recovered energy, which is the energy accumulated (remaining energy) in the laser medium 51, is gradually consumed when input light La is input to the laser medium 51 and amplified. The amplification factor of the optical amplifier 5 changes due to a change in the ratio between the consumed energy and the stored energy. For example, when the stored energy is 10 and an energy of 1 is extracted, compared with when the stored energy is 2 and an energy of 1 is extracted, it is more difficult to extract energy in the latter case (i.e., the amplification factor is smaller).

[0035] FIG. 3 is a simplified graph plotting "energy consumed / stored energy" as the energy consumed gradually increases. As shown in FIG. 3, the energy consumed / stored energy ratio has a region where it increases approximately linearly (within the dashed-dotted line frame in the figure) and a region where it increases nonlinearly (within the dashed-dotted line frame in the figure). When creating a fitting function, the calculation can be conveniently performed by separating the function into a linear term and a nonlinear term as a function of energy consumed. In reality, the laser medium 51 continues to be irradiated with the excitation light L0 even while the input light La is being input to the laser medium 51, and the energy in the laser medium 51 is consumed and recovered at the same time. Because the recovered energy is a factor unrelated to the energy consumed, fitting that simply takes into account the energy consumed (the function F(x) related to the energy consumed) may result in an inaccurate fitting.

[0036] Therefore, in this embodiment, a function f(t) related to the recovery energy, which is a time term, is adopted. Because the excitation light L0 is continuously irradiated onto the laser medium 51, the recovery energy is a function of t. In the fitting function, x changes with time, so the function f(t) can essentially be treated as a function of x. When creating the fitting function, accurate fitting can be performed by taking into account not only the simple function F(x) of energy consumption, but also the function f(t) related to the recovery energy, which is a time term.

[0037] Both F(x) and f(t) have regions that can be treated approximately linearly and regions that can be treated nonlinearly. Accurate fitting can be achieved by treating the fitting function (function model) as being divided into linear and nonlinear terms. That is, the function model may include a linear term of a time integral value. The function model may include a nonlinear term of a time integral value. The function model may also include a constant term. The constant term may be determined by a threshold current required for amplification. In this embodiment, the function model can be represented by a function including a linear term L(x) and a nonlinear term NL(x) on a graph in which the horizontal axis represents the time integral value of the output waveform and the vertical axis represents 1 / amplification factor (= input waveform / output waveform) (see FIG. 6B). The linear term L(x) and the nonlinear term NL(x) each include components of F(x) and f(t). Function model = L(x) + NL(x).

[0038] In this embodiment, a model MD is created and stored for each of a plurality of repetition frequencies of the input light La (hereinafter simply referred to as "multiple repetition frequencies"). Specifically, the input waveform generation unit 12 generates an input waveform for each of a plurality of different repetition frequencies and provides the waveform data to the driver circuit 2. As a result, the input light La is emitted from the semiconductor laser element 3 and input to the optical amplifier 5 for each of the plurality of different repetition frequencies, and the optical amplifier 5 emits the output light Lb. The detector 7 detects the output waveform of the output light Lb for each of the plurality of different repetition frequencies. The model creation unit 13 creates a model MD for each of the plurality of different repetition frequencies. The storage unit 14 stores the model MD created by the model creation unit 13 for each of the plurality of different repetition frequencies. The plurality of repetition frequencies is not particularly limited and may be a predetermined number of different frequencies. For example, the plurality of repetition frequencies are 1 kHz, 2.5 kHz, 5 kHz, and 10 kHz.

[0039] In the above, the excitation light source 8 constitutes the excitation section. The input waveform generation section 12, the driver circuit 2, the semiconductor laser element 3, and the optical amplifier 5 constitute the amplification section. The optical branching section 6 and the detector 7 constitute the detection section. The driver circuit 2, the semiconductor laser element 3, the optical amplifier 5, the optical branching section 6, the detector 7, the excitation light source 8, the input waveform generation section 12, and the model creation section 13 constitute the model creation device 1A.

[0040] Next, a model creation method performed by the model creation device 1A will be described with reference to the flowchart of FIG.

[0041] First, the input waveform generator 12 generates an input waveform for creating a model (see, for example, FIG. 5A) in response to various inputs received by the input unit 11 (step S1). The excitation light source 8 irradiates the laser medium 51 with excitation light L0 (step S2: excitation step). The input light La having the input waveform generated in step S1 is emitted from the semiconductor laser element 3, and the input light La is input to the laser medium 51 irradiated with the excitation light L0 in step S2. This amplifies the input light La and causes it to be emitted as output light Lb from the laser medium 51 (step S3: amplification step).

[0042] The output light Lb emitted from the laser medium 51 in step S3 is detected by the detector 7 via the optical branching unit 6. This detects the output waveform of the output light Lb (see, for example, FIG. 5B) (step S4: detection step). The model creation unit 13 creates a model based on the time integral of the output waveform detected in step S4 and the amplification factor of the output waveform relative to the input waveform input in step S3 (step S5: model creation step). In step S5, the model creation unit 13 fits the measured data D0 (see FIG. 6A) on a graph with the time integral of the output waveform on the horizontal axis and the 1 / amplification factor on the vertical axis. This creates a function model including a fitting function (= linear term L(x) + nonlinear term NL(x)) as a model (see FIG. 6B).

[0043] The created model is stored in the storage unit 14 (step S6: storage step). The model creation unit 13 determines whether or not the generation and storage of models has been completed for each of a predetermined number of repetition frequencies (step S7). If the answer is NO in step S7, the repetition frequency of the input light La emitted from the semiconductor laser element 3 is changed to a repetition frequency for which the generation and storage of models has not yet been completed (step S8). After step S8, the process returns to step S3. On the other hand, if the answer is YES in step S7, the generation and storage of models has been completed for all of the predetermined number of repetition frequencies, and the process ends.

[0044] The timing of performing the above-described model creation method is not particularly limited. For example, it may be performed at specific intervals taking into account aging degradation of the optical amplifier 5. For example, it may be performed automatically when the deviation between the input waveform and the output waveform exceeds a preset threshold. For example, it may be performed when the user inputs a command to start maintenance via the input unit 11.

[0045] 7, the laser device 1 of this embodiment uses the created model MD to generate the input waveform of the input light La to be input to the optical amplifier 5. Specifically, the laser device 1 generates the input waveform from the target waveform using the model MD so that the output waveform of the output light Lb output from the optical amplifier 5 becomes the target waveform.

[0046] The input unit 11 receives input related to a target waveform, which is a desired output waveform. The target waveform is a time waveform of a current represented on a graph with the horizontal axis representing time (nsec) and the vertical axis representing current (mA). The target waveform is, for example, a double-pulse waveform including two rectangular waves, as shown in FIG. 8( a). Each time the target waveform is changed (e.g., each time a new target waveform is received by the input unit 11), the input waveform generation unit 12 reads the model MD from the storage unit 14 and generates an input waveform that realizes the target waveform based on the target waveform and the model MD.

[0047] When generating an input waveform in the input waveform generator 12, first, a function model (fitting function) constituting the model MD is applied to the target waveform. As a result of applying the model MD, a parameter related to the amplification factor is obtained. Here, the parameter related to the amplification factor is 1 / amplification factor, and as shown in FIG. 8( b), it is represented on a graph with the horizontal axis representing the time integral value of the output waveform and the vertical axis representing 1 / amplification factor. As an example, the result of applying the model MD to the target waveform is a waveform in which the 1 / amplification factor increases linearly as the time integral value of the output waveform increases.

[0048] The input waveform generation unit 12 corrects the target waveform using the result of applying the model MD to the target waveform. Specifically, the input waveform (hereinafter also referred to as the "corrected input waveform") is obtained by multiplying the target waveform by 1 / amplification factor, which is the result of applying the model MD. This is because the corrected input waveform can be obtained by multiplying 1 / amplification factor by the target waveform, considering that the corrected input waveform can be expressed as 1 / amplification factor = corrected input waveform / target waveform. As shown in FIG. 9(a), the corrected input waveform is represented on a graph with time on the horizontal axis and current on the vertical axis. As an example, an input waveform that realizes the target waveform is a waveform in which the peaks of each rectangular wave of the double pulse include a waveform in which the current increases linearly with increasing time. In this embodiment, an input waveform is generated from the target waveform using the model MD for each of multiple repetition frequencies so that the output waveform of the output light Lb becomes the target waveform. As described above, the input unit 11, the input waveform generation unit 12, and the storage unit 14 constitute an input waveform generation device 1B that generates an input waveform using the model MD.

[0049] Next, a method for irradiating laser light by the laser device 1 will be described with reference to the flowchart of Fig. 10. Here, the description will also include a method for generating an input waveform performed by the input waveform generating device 1B.

[0050] First, a target waveform, for example, as shown in Fig. 8A, is input via the input unit 11 (step S11). The input waveform generation unit 12 reads out a model MD from the storage unit 14 (step S12). Based on the target waveform input in step S11 and the model MD read out in step S12, the input waveform generation unit 12 generates an eye-corrected input waveform (step S13: input waveform generation step).

[0051] The excitation light source 8 irradiates the laser medium 51 with excitation light L0 (step S14). The input light La having the corrected input waveform generated in step S13 is emitted from the semiconductor laser element 3, and the input light La is input to the laser medium 51 irradiated with the excitation light L0 in step S14. This amplifies the input light La and causes it to be emitted as output light Lb from the laser medium 51 (step S15). As a result, the output light Lb is output to the outside of the laser device 1 and is used, for example, for laser processing and various measurements. Furthermore, for example, the detector 7 can confirm that the output waveform of the output light Lb corresponds to the target waveform (has the same shape or a similar shape) (see FIG. 9B).

[0052] Generally, in an optical amplifier 5 in which a laser medium 51 is excited by pumping light L0 and input light La is amplified by the laser medium 51, the amplification factor of the laser medium 51 gradually decreases during emission of the input light La, resulting in an output waveform that does not match the desired waveform. Therefore, the model creation method of this embodiment utilizes the fact that there is a fixed relationship between the amplification factor and the time integral value of the output waveform, making it possible to create a model MD that reflects the output characteristics of the optical amplifier 5 from actual data without repeating various predictions. Taking into account the energy consumption and recovery in the laser medium 51, the output characteristics of the optical amplifier 5 can be modeled based on the time integral value and the amplification factor. This makes it possible to easily and accurately predict the output characteristics of the optical amplifier 5. Even if the amplification characteristics of individual optical amplifiers 5 vary, an accurate model MD can be created. Furthermore, because the rate of change of the amplification factor relative to the time integral value of the output waveform is the same regardless of the input waveform, once the model MD is created, it is possible to generate an input waveform corresponding to any target waveform.

[0053] In the model creation method of this embodiment, the model MD is a function model representing parameters related to the amplification factor, and includes a function f(t) related to stored energy (recovery energy) and a function F(x) related to consumed energy. In this case, a function model that takes into account the stored energy and consumed energy can be created as the model MD that reflects the output characteristics of the optical amplifier 5.

[0054] In the model creation method of this embodiment, the function model includes a linear term L(x) of the time integral value. In this case, the linear term L(x) of the time integral value can be used as the function model. In the model creation method of this embodiment, the function model includes a nonlinear term NL(x) of the time integral value. In this case, the nonlinear term NL(x) of the time integral value can be used as the function model.

[0055] In the model creation method of this embodiment, in the amplification step, input light La is input to the optical amplifier 5 and output light Lb is emitted for each of a plurality of different repetition frequencies, in the detection step, the output waveform of the output light Lb is detected for each of the plurality of repetition frequencies, and in the model creation step, a model MD is created for each of the plurality of repetition frequencies. In this case, even if the repetition frequency of the input light La is different, it is possible to create a model MD with high accuracy.

[0056] In the model creation method of this embodiment, the input waveform of the input light in the amplification step is a waveform whose first-order time derivative is continuous. In this case, the time integral value of the output waveform becomes continuous, making it possible to create the model MD with high accuracy.

[0057] The input waveform generation method of this embodiment includes an input waveform generation step of generating an input waveform based on a target waveform and a model MD. In this manner, in this embodiment, an input waveform can be generated using a model MD that can accurately predict the output characteristics of the optical amplifier 5, thereby enabling the generation of an input waveform that accurately achieves the target waveform. As a result, data-driven laser processing with excellent processing rate, processing quality, processing efficiency, etc. can be achieved. This eliminates the need for an operator to repeatedly adjust the input waveform using an oscilloscope or the like, thereby reducing the number of steps required.

[0058] The model creation method of this embodiment includes a storage step for storing the created model MD. In the input waveform generation method of this embodiment, the model MD stored in the storage step is read out each time the target waveform is changed, and an input waveform is generated based on the target waveform and the model MD. In this case, by storing the model MD, it is possible to generate an appropriate input waveform even if the target waveform is changed.

[0059] The model creation device 1A of this embodiment also utilizes the fact that there is a fixed relationship between the amplification factor and the time integral value of the output waveform, making it possible to create a model MD that reflects the output characteristics of the optical amplifier 5 without repeating various predictions from actual data. Therefore, it becomes possible to predict the output characteristics of the optical amplifier 5 easily and accurately.

[0060] In this embodiment, as described above, the function model (fitting function) constituting the model MD is expressed as 1 / amplification factor = input waveform / output waveform, which is preferable because it simplifies the calculation and allows the desired input waveform to be generated by multiplying the desired output waveform (i.e., target waveform) by 1 / amplification factor.

[0061] In this embodiment, the input waveform for model creation preferably has a waveform in which the input value does not become zero for that time width (for example, a long pulse of rectangular wave pulse light). If there is a period in which the input value becomes zero (the output value becomes zero), both the input and output values ​​will become zero during that period, and therefore, that period must be excluded during fitting, which may complicate the calculations and reduce the fitting accuracy. In addition, the rising portion of the input waveform for model creation (the portion from 0 to a certain value) does not need to be used in creating the model MD.

[0062] 11(a) and 11(b) are graphs showing output waveforms when the target waveform is a uniform intensity triangular wave. FIG. 11(a) shows the output waveform when input light La, which uses the target waveform as the input waveform without using the model MD, is input to the optical amplifier 5. FIG. 11(b) shows the output waveform when input light La, which has a corrected input waveform generated from the target waveform using the model MD, is input to the optical amplifier 5. In the figures, the horizontal axis represents time, and the vertical axis represents voltage or current. The deviation between the target waveform and the output waveform on the vertical axis in the figures is due to, for example, differences in the units or scale of the measurement values, and does not contribute to waveform matching or approximation. The same applies to the following graphs.

[0063] When the model MD is not used, as shown in Fig. 11(a), the output waveform does not have the same intensity as the target waveform due to attenuation in the optical amplifier 5. In contrast, when the input waveform is generated using the model MD, as shown in Fig. 11(b), the output waveform is corrected and approaches the same intensity as the target waveform.

[0064] 12(a) and 12(b) are graphs showing output waveforms when the target waveform is an attenuation intensity triangular wave. FIG. 12(a) shows the output waveform when input light La, which uses the target waveform as the input waveform without using the model MD, is input to the optical amplifier 5. FIG. 12(b) shows the output waveform when input light La, which has an input waveform generated from the target waveform using the model MD, is input to the optical amplifier 5. When the model MD is not used, as shown in FIG. 12(a), the slope of the line connecting the peaks of the output waveform does not match that of the target waveform. In contrast, when the input waveform is generated using the model MD, as shown in FIG. 12(b), the output waveform is corrected, and the slope of the line connecting the peaks of the output waveform and the target waveform nearly match.

[0065] 13(a) and 13(b) are graphs showing output waveforms when the target waveform is a uniform intensity Gaussian pulse. FIG. 13(a) shows the output waveform when input light La, which uses the target waveform as the input waveform without using the model MD, is input to the optical amplifier 5. FIG. 13(b) shows the output waveform when input light La, which has a corrected input waveform generated from the target waveform using the model MD, is input to the optical amplifier 5. When the model MD is not used, as shown in FIG. 13(a), the output waveform does not have the same intensity as the target waveform due to attenuation in the optical amplifier 5. In contrast, when the input waveform is generated using the model MD, the output waveform is corrected and approaches the same intensity as the target waveform, as shown in FIG. 13(b).

[0066] 14(a) and 14(b) are graphs showing output waveforms when the target waveform is an attenuated intensity Gaussian pulse. FIG. 14(a) shows the output waveform when input light La, which uses the target waveform as the input waveform without using the model MD, is input to the optical amplifier 5. FIG. 14(b) shows the output waveform when input light La, which is a corrected input waveform generated from the target waveform using the model MD, is input to the optical amplifier 5. When the model MD is not used, as shown in FIG. 14(a), the slope of the line connecting the peaks of the output waveform does not match that of the target waveform. In contrast, when the input waveform is generated using the model MD, as shown in FIG. 14(b), the output waveform is corrected, and the slope of the line connecting the peaks of the output waveform and the target waveform nearly match.

[0067] 15(a) and 15(b) are graphs showing output waveforms when the target waveform is Gaussian. FIG. 15(a) shows the output waveform when input light La, which uses the target waveform as the input waveform without using the model MD, is input to the optical amplifier 5. FIG. 15(b) shows the output waveform when input light La, which has a corrected input waveform generated from the target waveform using the model MD, is input to the optical amplifier 5. When the model MD is not used, as shown in FIG. 15(a), the output waveform has a shape that leans forward (toward the shorter time) with respect to the target waveform. In contrast, when the input waveform is generated using the model MD, as shown in FIG. 15(b), it can be seen that the output waveform is corrected and the lean is eliminated.

[0068] 16(a) and 16(b) are graphs showing output waveforms when the target waveform is a triangular wave. FIG. 16(a) shows the output waveform when input light La, which uses the target waveform as the input waveform without using the model MD, is input to the optical amplifier 5. FIG. 16(b) shows the output waveform when input light La, which has a corrected input waveform generated from the target waveform using the model MD, is input to the optical amplifier 5. When the model MD is not used, as shown in FIG. 16(a), the output waveform bulges up to the peak position (apex) compared to the target waveform, and the portion after the peak position (the side with a longer time) has a curved shape. In contrast, when the input waveform is generated using the model MD, as shown in FIG. 16(b), the output waveform is corrected to a linear shape both before and after the peak position.

[0069] Furthermore, no substantial change in the output waveform was observed when the output (current value) of the input light La was changed. Therefore, in this embodiment, even when the output (current value) of the input light La is changed, one model MD can be used.

[0070] The above-described aspect of the present invention is not limited to the above-described embodiment.

[0071] 17 is a block diagram showing the configuration of a modified laser device 101. As shown in Fig. 17, the laser device 101 (input waveform generating device 101B) differs from the above-described laser device 1 (input waveform generating device 1B: see Fig. 7) in that it does not include the optical branching unit 6, the detector 7, and the model creating unit 13.

[0072] In such a modified example, an optimal model is read out from the plurality of models MD stored in the storage unit 14 in accordance with the amplification conditions (e.g., target waveform, repetition frequency of input light La, etc.) set via the input unit 11, and a corrected input waveform is generated from the model MD and the target waveform by the input waveform generation unit 12. Then, excitation light L0 is irradiated onto the laser medium 51 from the excitation light source 8, and input light La of the corrected input waveform is emitted from the semiconductor laser element 3 and input to the laser medium 51. As a result, the input light La is amplified, and output light Lb is output from the laser medium 51 to the outside of the laser device 1.

[0073] In recent laser devices, model creation devices, laser processing methods, and model creation methods, it is desirable to generate an input waveform that accurately realizes a target waveform, thereby achieving laser processing with excellent processing rate, processing quality, processing efficiency, etc. Therefore, an object of the present disclosure is to accurately realize the target waveform. In this regard, the laser device 101 and input waveform generation device 101B according to the modified example can read out an optimal model MD to generate an input waveform, making it possible to generate an input waveform that accurately realizes the target waveform.

[0074] In the above embodiment, the model MD may have a lookup table in which the time integral value is associated with a parameter related to the amplification factor. In this case, the model MD can be created using the lookup table.

[0075] The above embodiment can also be defined as a program for creating a model MD. That is, the present disclosure may be a model creation program for creating a model MD that reflects the output characteristics of an optical amplifier 5, the model creation program causing a computer 10 to execute an amplification step of inputting input light La to a laser medium 51 of the optical amplifier 5 that is irradiated with excitation light L0, amplifying the input light La, and emitting output light Lb from the laser medium 51, and a model creation step of creating a model MD based on a time integral value of the output waveform of the output light Lb and an amplification factor of the output waveform with respect to the input waveform of the input light La.

[0076] In the above embodiment, the created model MD is stored in the storage unit 14, but this is not limiting and the model MD may be stored externally (external storage medium, cloud server, etc.) In the above embodiment, the model MD is obtained by reading it from the storage unit 14, but it may also be obtained by downloading it from an external source or by reading it from a database attached to the device.

[0077] The respective components in the above-described embodiment and modified examples are not limited to the materials and shapes described above, and various materials and shapes can be applied. Furthermore, the respective components in the above-described embodiment and modified examples can be arbitrarily applied to the respective components in other embodiments or modified examples.

[0078] 1A...model creation device, 2...driver circuit (amplification section), 3...semiconductor laser element (amplification section), 5...optical amplifier (amplification section), 6...optical branching section (detection section), 7...detector (detection section), 8...excitation light source (excitation section), 12...input waveform generation section (amplification section), 13...model creation section, 51...laser medium, L0...excitation light, La...input light, Lb...output light, MD...model.

Claims

1. A method for creating a model that reflects the output characteristics of an optical amplifier, comprising: an excitation step of irradiating excitation light onto a laser medium of the optical amplifier; an amplification step of inputting input light into the laser medium irradiated with the excitation light, amplifying the input light, and emitting output light from the laser medium; a detection step of detecting an output waveform of the output light emitted in the amplification step; and a model creation step of creating the model based on the time integral value of the output waveform detected in the detection step and the amplification factor of the output waveform relative to the input waveform of the input light input in the amplification step.

2. A model creation method as described in claim 1, wherein the model is a functional model representing parameters related to the amplification rate, and the functional model includes a function related to the energy stored in the laser medium by irradiation with the excitation light, and a function related to the energy consumed by amplification in the laser medium.

3. The model creation method according to claim 2, wherein the function model includes a linear term of the time integral value.

4. The model creation method according to claim 2 or 3, wherein the function model includes a nonlinear term of the time integral value.

5. The model creation method according to claim 1, wherein the model has a look-up table in which the time integral value and a parameter related to the amplification factor are associated with each other.

6. A model creation method as described in claim 1 or 2, wherein in the amplification step, the input light is input to the optical amplifier for amplification for each of a plurality of repetition frequencies different from one another, and the output light is emitted from the optical amplifier; in the detection step, an output waveform of the output light is detected for each of the plurality of repetition frequencies; and in the model creation step, the model is created based on the time integral value and the amplification factor for each of the plurality of repetition frequencies.

7. A model creation method according to claim 1 or 2, wherein the input waveform of the input light input in the amplifying step is a waveform whose first-order time derivative is continuous.

8. A method for generating an input waveform of the input light to be input to the optical amplifier using the model created by the model creation method according to claim 1 or 2, comprising an input waveform generation step of generating the input waveform that realizes the target waveform based on a target waveform and the model created in the model creation step.

9. The method for generating an input waveform according to claim 8, wherein the model creation method includes a storage step for storing the created model, and in the input waveform generation step, each time the target waveform is changed, the model stored in the storage step is read out, and the input waveform that realizes the target waveform is generated based on the target waveform and the model.

10. An apparatus for creating a model that reflects the output characteristics of an optical amplifier, comprising: an excitation unit that irradiates excitation light onto a laser medium of the optical amplifier; an amplification unit that inputs input light to the laser medium irradiated with the excitation light, amplifies the input light, and emits output light from the laser medium; a detection unit that detects the output waveform of the output light emitted by the amplification unit; and a model creation unit that creates the model based on the time integral value of the output waveform detected by the detection unit and the amplification factor of the output waveform relative to the input waveform of the input light input to the amplification unit.

Images