A high temperature adaptable VCSEL light source

Abstract

The present invention relates to the technical field of semiconductor lasers and discloses a high-temperature adaptability VCSEL light source, comprising two VCSELs connected in parallel, which can respectively obtain high-efficiency light output in low-temperature and high-temperature environments. The peak value of the material gain of the active region of the VCSEL that works normally at low temperature is located in the long-wave direction of the resonant cavity mode at room temperature; the peak value of the material gain of the VCSEL that works normally at high temperature is located in the short-wave direction of the cavity mode, thereby causing the cavity mode to move toward the material gain peak at low and high temperatures to obtain high mode gain. The lasers emitted by the two VCSELs are combined into one by a fiber combiner, so that the device can achieve stable light output within a large temperature variation range, and a temperature monitoring and prediction layer and an adaptive layer are provided to realize temperature monitoring, prediction, and intelligent adaptive adjustment.

Description

Technical Field

[0001] The invention relates to the technical field of semiconductor lasers and discloses a VCSEL light source with high temperature adaptability. Background Art

[0002] As devices that generate optical signals, semiconductor lasers play an irreplaceable role in fiber-optic communication systems. Depending on the direction of light emission, semiconductor lasers can be divided into edge-emitting lasers (EELs) and surface-emitting lasers (VCSELs). Surface-emitting lasers have gradually become the primary light source for short-distance optical communications due to their low cost, symmetrical beam spot, high electro-optical conversion efficiency, ease of arraying, and on-wafer testing. Unlike edge-emitting lasers, the wavelength of the laser emitted by a surface-emitting laser is determined by the DBR reflector structure on either side of the laser and the material layer structure of the central light-emitting region. Because the resonant cavity thickness is only a few microns, the mode spacing of its longitudinal modes is large, resulting in single-longitudinal-mode lasing. As temperature increases, the refractive index of the DBR material of a surface-emitting laser increases, resulting in an increase in its optical thickness, a redshift in the reflection spectrum, and an increase in the emission wavelength. However, the redshift rate of the VCSEL spectrum with increasing temperature is less than 0.1 nm / °C, typically an order of magnitude lower than that of the edge-emitting laser spectrum, resulting in better wavelength stability.

[0003] To achieve high optical gain at room temperature, conventional surface-emitting lasers typically align the device's cavity mode with the peak gain of the active material, thereby lowering the device's threshold and improving electro-optical conversion efficiency. However, as temperature increases, the material gain decreases, and its redshift rate is much faster than that of the cavity mode. This causes conventional VCSELs to rapidly decrease in mode gain at high temperatures. Even after reaching a certain temperature, the optical power decreases with increasing current, a phenomenon known as thermal inversion. This characteristic severely impacts the VCSEL's adaptability to high-temperature operation.

[0004] To improve the high-temperature operating characteristics of VCSELs, existing technologies use wavelength detuning to separate the cavity mode from the material gain peak in the active region at room temperature. This allows the cavity mode and material gain peak to converge as the device temperature rises, resulting in good high-temperature performance. However, the degree of mismatch in such devices at low temperatures is greater than that of conventional VCSELs, causing their quantum efficiency to drop rapidly at low temperatures, making it difficult to maintain stable operation over a wide temperature range. Automotive devices typically need to withstand a wide temperature range of -40°C to 85°C, a requirement that neither conventional VCSELs nor single high-temperature VCSELs can meet. Therefore, a VCSEL light source that can achieve stable output across a wide temperature range from low to high temperatures is urgently needed.

[0005] For example, the existing patent with publication number CN110301076B discloses a light source array (1), comprising a plurality of VCSELs (2) arranged laterally relative to each other on a substrate (3), wherein each VCSEL (2) comprises a light emitting region (21) surrounded by an electrode structure (22) that does not emit light, wherein a shielding layer (4) is applied on at least the electrode structure (22), the shielding layer (4) only covering a surface (22s) of the electrode structure (22) facing the general light emission direction (5) of the VECSELs (2), the shielding layer (4) being an opaque layer and being adapted to optically match the array (10) in a closed state with an outer surface (11s) of a housing (11) of a device (10) in which the light source array (1) is to be installed. The present invention also describes a device (10) comprising such an array (1) and a method (100) for manufacturing the light source array (1).

[0006] The above patent lacks an active temperature control mechanism, making it difficult to cope with a wide temperature environment. It lacks internal status monitoring function, making it difficult to know in advance whether abnormal temperature changes occur inside the light source. It cannot provide early warning and targeted optimization of factors that may affect the performance of the light source. The light intensity control is not flexible enough and the functional focus is relatively single. Summary of the Invention

[0007] The purpose of this section is to summarize some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of this application to avoid obscuring the purpose of this section, the abstract and the title of the invention, and such simplifications or omissions should not be used to limit the scope of the present invention.

[0008] To solve the above technical problems, the main purpose of the present invention is to provide a high-temperature adaptability VCSEL light source, comprising:

[0009] One end of the light combiner is connected to a metal electrode through an optical coupling optical fiber, and one end of the metal electrode is sequentially provided with a passivation layer, a P-type DBR reflector, an oxide layer, a multi-quantum well active region, an N-type DBR reflector, an N-type substrate layer, and a lower metal electrode;

[0010] One end of the light combiner is parallel to a light coupling optical fiber connected to a phase matching layer;

[0011] The sensing layer monitors and predicts the temperature of the VCSEL light source by establishing a temperature model;

[0012] Adaptive layer, adaptively adjusts the VCSEL light source intensity through temperature stratification.

[0013] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0014] The VCSEL light source includes a low-temperature vertical cavity surface emitting laser and a high-temperature vertical cavity surface emitting laser connected in parallel, wherein the low-temperature vertical cavity surface emitting laser and the high-temperature vertical cavity surface emitting laser are filled with BCB material;

[0015] The high-temperature vertical cavity surface emitting laser comprises:

[0016] One end of the passivation layer is connected to the metal electrode, and the other end is provided with a P-type DBR reflector;

[0017] An oxide layer is provided at the other end of the P-type DBR reflector, and the other end of the oxide layer is connected to the N-type DBR reflector through a multi-quantum well active region;

[0018] An N-type substrate layer is provided at the other end of the N-type DBR reflector, and the N-type substrate layer is connected to a heat sink via a lower metal electrode.

[0019] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0020] The phase matching layer is located at the top of the P-type DBR reflector in the vertical cavity surface emitting laser working at low temperature and the vertical cavity surface emitting laser working at high temperature;

[0021] A multi-quantum well active region and a P-type DBR reflector are respectively arranged at both ends of the spacer layer;

[0022] The spacer layer is also provided between the multi-quantum well active region and the N-type DBR reflector;

[0023] The spacer layer is located between the multi-quantum well active region and the P-type DBR reflector and the N-type DBR reflector, and is used to adjust the length of the resonant cavity.

[0024] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0025] The sensing layer includes a temperature sensor, a temperature model and a data-driven model;

[0026] The temperature model detects the temperature distribution prediction data inside the VCSEL light source under different working conditions by conducting multi-layer heat conduction on the VCSEL light source and the heat exchange with the external environment;

[0027] Methods for obtaining the temperature distribution inside the VCSEL light source include:

[0028] S1. Divide the VCSEL light source into N layers from bottom to top according to its physical structure, marked as layer 1 to layer N respectively;

[0029] S2. Obtain heat flux density vector data and simplify it into one-dimensional heat conduction.

[0030]

[0031] Among them, Q is the heat flux density of each layer of the VCSEL light source, k is the thermal conductivity, T is the temperature of the VCSEL light source material layer, and x is the one-dimensional heat conduction direction;

[0032] S3. Output the temperature distribution prediction data inside the VCSEL light source through the data-driven model.

[0033] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0034] Data acquisition is used to receive temperature distribution data inside the VCSEL light source under different working conditions;

[0035] Data preprocessing includes data cleaning and normalization;

[0036] Multi-layer temperature prediction includes a multi-layer perception structure. Assume that the input layer has n temperature feature nodes inside the VCSEL light source, the hidden layer has L layers, and the number of hidden layer nodes in the zth layer is d z , where z = 1, 2, 3, ..., and the output layer has k nodes;

[0037] Randomly initialize the connection weight matrix W and bias vector b between each layer. For the connection between the zth layer and the z+1th layer, the dimension of the weight matrix W is d z+1 ×d z , bias vector b z Dimension d z+1 ×z;

[0038] The temperature distribution data inside the VCSEL light source is calculated and kernel operation is performed to further predict the temperature distribution data inside the VCSEL light source. The kernel operation expression is as follows:

[0039] T1=f(W z-1 Y+b z-1 )

[0040] Among them, T1 is the predicted value of the temperature distribution data inside the VCSEL light source, f(~) is the activation function, Y is the real-time temperature data of each layer inside the VCSEL light source, b is the bias, W z-1 is the weight of the hidden layer z-1, b z-1 is the bias of the hidden layer z-1;

[0041] The temperature distribution data set inside the entire VCSEL light source is calculated through a multi-layer temperature prediction process to obtain a predicted output multi-layer temperature matrix, whose dimension is the same as the target temperature value matrix, and each row corresponds to a predicted temperature value vector of a sample.

[0042] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0043] The loss is calculated by predicting the temperature inside the VCSEL light source. The loss is obtained by calculating the difference between the predicted temperature value and the actual temperature value. The data-driven model is then corrected by loss gradient compensation. The loss calculation expression is as follows:

[0044]

[0045] Where MES is the loss value between the predicted temperature and the actual temperature inside the VCSEL light source, a is the number of VCSEL light source layers, b is the temperature of the corresponding layer, T' a,b To predict the output of the multi-layer temperature matrix, T a,b is the actual temperature matrix, and M is the total number of data.

[0046] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0047] The weight of each part of the data-driven model is corrected through chaining, and the calculation expression is as follows:

[0048]

[0049] The weight of each layer is calculated in turn, and the weight of each layer is updated. The above steps are repeated for a certain number of iterations until the stopping condition is met.

[0050] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0051] The adaptive layer includes fuzzy reasoning, defuzzification, VCSEL light source control and control execution unit;

[0052] The light source fuzzy control receives the difference between the predicted VCSEL light source temperature and the pre-set high temperature threshold as the basic data, and uses the temperature change rate and optical power deviation as the input data of the light fuzzy control algorithm. Based on fuzzy reasoning, the fuzzified output of the VCSEL light source temperature variable and optical power variable is automatically compared with the fuzzy control rules. After defuzzification, the VCSEL light source heat dissipation components and output light are adjusted and controlled.

[0053] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0054] The calculation expression of the defuzzification is as follows:

[0055]

[0056] Where λ is the membership coefficient, ΔQ(x) is the temperature difference or optical power deviation control adjustment control rule, x is the input VCSEL light source temperature adjustment value, X is the total range of the temperature adjustment value, and Δ is the defuzzified VCSEL light source control instruction.

[0057] As a preferred solution of a high temperature adaptability VCSEL light source of the present invention, wherein:

[0058] The calculation expression for VCSEL light source control is as follows:

[0059]

[0060] e(t)=C on -C out (t)

[0061] Among them, u(t) is the driving current adjustment value on the VCSEL light source electrode, K p To adjust the light intensity proportional coefficient, K i To adjust the light intensity integral coefficient, K d is the differential coefficient for adjusting the light intensity, e(t) is the defuzzified VCSEL light source control instruction, t is the time, C on is the target light intensity, C out (t) is the predicted light intensity

[0062] The control execution unit is used to receive VCSEL light source control instructions and control the VCSEL light source.

[0063] Beneficial effects of the present invention:

[0064] By connecting a low-temperature VCSEL and a high-temperature VCSEL in parallel and filling them with BCB material, the light source can operate normally over a wide temperature range. By collecting temperature distribution data under different operating conditions and performing preprocessing such as cleaning and normalization, and then using a multi-layer perception structure to perform kernel operations to predict temperature, and finally reverse-correcting the model based on the predicted loss, the prediction of the temperature distribution within the light source is becoming increasingly accurate. Accurate temperature prediction provides a reliable basis for subsequent adaptive adjustments, helps achieve precise temperature control, and efficiently and accurately regulates the light source, thus ensuring that the light source can still stably and efficiently output the required laser light in complex and changing temperature environments, thereby enhancing the practical value of the light source and its applicability in different application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. Those skilled in the art can also derive other drawings based on these drawings without inventive effort. Among them:

[0066] Figure 1 This is a structural diagram of a high-temperature adaptability VCSEL light source of the present invention;

[0067] Figure 2 This is a simulation diagram of the spectrum and active region gain spectrum of a low-temperature VCSEL light source with high temperature adaptability at room temperature (23°C) according to the present invention;

[0068] Figure 3 The spectrum of the high-temperature VCSEL light source with high temperature adaptability and the gain spectrum of the active region at room temperature (23°C) of the present invention;

[0069] Figure 4 This is a simulation diagram of the PIV characteristics of a low-temperature VCSEL at an ambient temperature of -20°C, a high-temperature adaptability VCSEL light source of the present invention;

[0070] Figure 5 This is a simulation diagram of the PIV characteristics of a high-temperature VCSEL light source with high temperature adaptability at an ambient temperature of 80°C.

[0071] Figure numerals: 1. Metal electrode; 2. Passivation layer; 3. P-type DBR reflector; 4. Oxide layer; 5. Multi-quantum well active region; 6. N-type DBR reflector; 7. N-type substrate layer; 8. Lower metal electrode; 9. Spacer layer; 10. Phase matching layer; 11. Heat sink; 12. BCB material; 13. Light coupling into optical fiber; 14. Fiber combiner. DETAILED DESCRIPTION

[0072] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0073] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0074] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0075] Example 1:

[0076] like Figure 1 As shown, a high temperature adaptability VCSEL light source includes:

[0077] One end of the light combiner 14 is connected to the metal electrode 1 through the light coupling optical fiber 13. One end of the metal electrode 1 is sequentially provided with a passivation layer 2, a P-type DBR reflector 3, an oxide layer 4, a multi-quantum well active region 5, an N-type DBR reflector 6, an N-type substrate layer 7, and a lower metal electrode 8;

[0078] The VCSEL light source includes a low-temperature vertical cavity surface emitting laser and a high-temperature vertical cavity surface emitting laser connected in parallel, and the low-temperature vertical cavity surface emitting laser and the high-temperature vertical cavity surface emitting laser are filled with BCB material 12;

[0079] A high-temperature VCSEL and a low-temperature VCSEL are connected in parallel. Their output lasers are combined into one parallel output through two directly coupled single-mode optical fibers via an external fiber combiner. This allows the optical powers of the VCSEL light source to compensate for each other at both low and high temperatures, ultimately achieving stable light output over a wide temperature range.

[0080] To ensure wavelength matching between the two lasers, the low- and high-temperature VCSELs are designed with similar cavity lengths and DBR structures. The active region gain peak of the low-temperature VCSEL at room temperature is designed to be located at the short-wavelength position of the cavity mode, while the active region gain peak of the high-temperature VCSEL at room temperature is designed to be located at the long-wavelength position of the cavity mode. This allows the gain spectrum to blueshift and redshift relative to the cavity mode, respectively, as the ambient temperature decreases or increases, resulting in higher modal gain for both devices at low and high temperatures. By increasing the thickness of the active region quantum well or quantum dot layer and reducing the In content to minimize lattice mismatch, the material gain spectrum is flattened, thereby increasing the wavelength detuning of the high-temperature VCSEL and ensuring its normal operation at higher temperatures.

[0081] Furthermore, the VCSEL surface and sidewalls are electrically isolated by vapor-depositing an insulating layer of silicon oxide or silicon nitride to prevent further oxidation caused by air penetrating the gaps between the oxide layers after prolonged high-temperature operation. After fabrication, the high- and low-temperature VCSELs are press-welded to the same metal heatsink substrate, with a spacing greater than 100 μm to reduce thermal crosstalk. The gap is then filled with BCB material, and finally, the top electrodes of the two VCSEL devices are connected by gold vapor deposition.

[0082] High temperature vertical cavity surface emitting lasers, including:

[0083] One end of the passivation layer 2 is connected to the metal electrode 1, and the other end is provided with a P-type DBR reflector 3;

[0084] An oxide layer 4 is provided at the other end of the P-type DBR reflector 3, and the other end of the oxide layer 4 is connected to an N-type DBR reflector 6 via a multi-quantum well active region 5;

[0085] The other end of the N-type DBR reflector 6 is provided with an N-type substrate layer 7, which is connected to a heat sink 11 through a lower metal electrode 8.

[0086] One end of the light combiner 14 is parallel to the light coupling optical fiber 13 and connected to the phase matching layer 10;

[0087] The sensing layer monitors and predicts the temperature of the VCSEL light source by establishing a temperature model;

[0088] Adaptive layer, adaptively adjusts the VCSEL light source intensity through temperature stratification.

[0089] The phase matching layer 10 is located on the top of the P-type DBR reflector 3 in the vertical cavity surface emitting laser working at low temperature and the vertical cavity surface emitting laser working at high temperature;

[0090] The two ends of the spacer layer 9 are respectively provided with a multi-quantum well active region 5 and a P-type DBR reflector 3;

[0091] The spacer layer 9 is also provided between the multi-quantum well active region 5 and the N-type DBR reflector 6;

[0092] The spacer layer 9 is located between the multi-quantum well active region 5 and the P-type DBR reflector 3 and the N-type DBR reflector 6, and is used to adjust the resonant cavity length;

[0093] Furthermore, there are multiple spacer layers 9 and they are present on both the top of the low-temperature VCSEL and the top of the high-temperature VCSEL.

[0094] Further, such as Figure 1The VCSEL light source of the present invention includes a low-temperature (left) and high-temperature (right) vertical cavity surface emitting laser connected in parallel. Both VCSELs include an upper metal electrode 1, a passivation layer 2, a P-type DBR reflector 3, an oxide layer 4, a multi-quantum well active region 5, an N-type DBR reflector 6, an N-type substrate layer 7, a lower metal electrode 8, a spacer layer 9, and a phase matching layer 10.

[0095] Furthermore, the lower electrodes of the two VCSELs are welded to the same heat sink 11, and the central wavelengths of the spectra of the two VCSELs are both set near 980nm. The quantum well material of the active region of the low-temperature VCSEL uses In0.23Ga0.77As material, and the active region of the high-temperature VCSEL uses In0.18Ga0.82As material with a lower In content, so that the material gain peak is located at the short-wave position of the central wavelength. The gap between the two VCSELs is filled with BCB material 12, and the light output port couples the output light into the optical fiber 13 by direct coupling. The two beams are combined by the optical fiber combiner 14 and output as one beam, so that the output optical power is approximately the sum of the output optical powers of the two VCSELs.

[0096] Furthermore, the low-temperature VCSEL uses a 6-micron oxide aperture, while the high-temperature VCSEL uses a 10-micron oxide aperture for current limiting. The low-temperature VCSEL utilizes a single oxide layer at the top of the active region, made of Al0.98Ga0.02As. The high-temperature VCSEL has oxide layers at both the top and bottom of the active region, with the top oxide layer made of Al0.97Ga0.03As and the bottom oxide layer made of Al0.98Ga0.02As. The devices are grown on an n-type GaAs substrate, and the high and low refractive index materials of the DBR reflector are GaAs and Al0.9Ga0.1As, respectively.

[0097] Figure 2 This is a simulation diagram of the correspondence between the low-temperature spectrum of the VCSEL at room temperature (23°C) and the gain spectrum of the active area;

[0098] Figure 3 This is a simulation diagram of the correspondence between the high-temperature spectrum of the VCSEL at room temperature (23°C) and the gain spectrum of the active area;

[0099] Figure 4 PIV characteristics of low-temperature VCSEL at -20°C ambient temperature;

[0100] Figure 5 PIV characteristics of high-temperature VCSEL at 80°C ambient temperature;

[0101] Furthermore, a specific example of setting parameters for a high temperature adaptable VCSEL light source is provided, including

[0102] The beam divergence angle of a VCSEL is directly related to the thickness of the resonant cavity. The thicker the resonant cavity, the smaller the beam divergence angle, and thus the larger its numerical aperture and the higher the coupling efficiency with the optical fiber. However, an increase in the cavity length often leads to a decrease in the longitudinal mode spacing and a decrease in the high-frequency performance of the device. Therefore, after comprehensive consideration, a resonant cavity length between 2λ and 5λ is finally selected.

[0103] Both low- and high-temperature VCSELs use an oxide layer to limit the lateral diffusion of current. Since the active region has higher gain at low temperatures, high optical output power and efficiency are easily achieved. However, at high temperatures, the gain decreases. Considering power matching at both high and low temperatures, the oxide aperture size of the low-temperature VCSEL is selected to be ≤8μm to appropriately increase the device's differential resistance and reduce the optical output power. At the same time, a smaller oxide aperture effectively reduces the number of transverse modes in the device, improving beam quality. For high-temperature VCSELs, a small oxide aperture can severely affect device heat dissipation. Therefore, the oxide aperture is set to ≥8μm to improve heat dissipation and avoid exceeding the thermal damage threshold during high-temperature operation, which could lead to device damage. Furthermore, considering the lower gain of the high-temperature VCSEL, this design incorporates an oxide layer on both the upper and lower sides of the active region to enhance current limiting capability and lower the device threshold. In actual device fabrication, VCSEL mesa etching is often difficult to achieve steeply, resulting in the lower oxide layer being wider than the upper one. To address this, the Al content of the lower oxide layer is designed to be slightly higher than that of the upper oxide layer to increase the wet oxygen oxidation rate, resulting in comparable upper and lower oxide aperture sizes for optimal current confinement. The oxide layer is positioned at the nodes of the VCSEL's standing wave field to minimize current crowding at the oxide layer's edges.

[0104] An external driver circuit supplies the device with a current greater than the sum of the maximum threshold currents of the two VCSELs over the full temperature range to drive normal operation. In low-temperature environments, the low-temperature VCSEL achieves high-power output, while the high-temperature VCSEL has lower power. In high-temperature environments, the low-temperature VCSEL's power decreases, while the high-temperature VCSEL achieves higher-power optical output. At room temperature, both lasers achieve moderate optical output power. The two VCSELs are directly coupled to optical fibers, combined by a fiber combiner, and then output through a single fiber. This allows the device to achieve relatively constant optical power output in low-temperature, room-temperature, and high-temperature environments.

[0105] Example 2:

[0106] A high temperature adaptable VCSEL light source, comprising:

[0107] The sensing layer monitors and predicts the temperature of the VCSEL light source by establishing a temperature model;

[0108] Among them, the sensing layer includes temperature sensors, temperature models and data-driven models;

[0109] Furthermore, the temperature sensor uses a miniature high-precision thermistor as the temperature sensor. Based on the structural characteristics and thermal conductivity of the VCSEL light source, multiple thermistors are distributed and arranged in different key positions;

[0110] For example:

[0111] Several thermistors are evenly spaced on the surface of the N-type substrate layer 7. Since the substrate layer is closely connected to the entire VCSEL chip and conducts the heat generated by the chip during operation, the thermistors here can directly obtain the approximate temperature of the entire chip. 3-5 thermistors can be arranged here to form a preliminary temperature monitoring point network.

[0112] On the surface of the heat sink 11 near the spacer layer 9, several thermistors (e.g., 2-4) are reasonably arranged along the edge and center area of ​​the heat sink. The heat sink is mainly used for heat dissipation. Monitoring its temperature changes helps to understand the heat dissipation efficiency and the heat distribution in the entire heat dissipation channel, thereby indirectly reflecting the thermal state of the VCSEL light source during operation.

[0113] At appropriate locations around the multi-quantum well active region 5, extremely small thermistors (1-2) are embedded using micro-nano processing technology. This is because the multi-quantum well active region is the core area for laser generation, and the temperature here has a significant impact on the performance of the light source. Although it is difficult to deploy a large number of sensors due to space limitations, even a small number of high-precision sensors can obtain critical temperature data here, helping to accurately understand the thermal environment of the active region.

[0114] Build a multiplexed temperature acquisition circuit, connect each thermistor to the circuit, provide a stable drive current to each thermistor through a constant current source, and use a high-precision analog-to-digital converter (ADC) to convert the voltage changes caused by temperature changes across the thermistor into digital signals. Select an ADC with a resolution of at least 12 bits or even higher to ensure that it can accurately distinguish the voltage difference corresponding to small temperature changes and ensure that the temperature measurement accuracy is within ±0.1°C. The acquisition circuit is equipped with a microcontroller (such as a low-power, high-performance single-chip microcomputer). The microcontroller collects the temperature data corresponding to each thermistor in sequence at a certain time interval (for example, every 100 milliseconds) and temporarily stores this data in an internal buffer area;

[0115] Furthermore, the temperature model detects the temperature distribution inside the VCSEL light source under different working states by conducting multi-layer heat conduction to the VCSEL light source and the heat exchange with the external environment;

[0116] First, the VCSEL light source is divided into N layers from bottom to top according to its physical structure, labeled as layer 1 to layer N. Each layer has its own specific material properties. For example, the N-type substrate layer 7 and the multi-quantum well active region 5 correspond to different layers and have different thermal parameters.

[0117] Secondly, according to Fourier's law, the heat flux density vector data is obtained and simplified into one-dimensional heat conduction.

[0118]

[0119] Among them, Q is the heat flux density of each layer of the VCSEL light source, k is the thermal conductivity, T is the temperature of the VCSEL light source material layer, and x is the one-dimensional heat conduction direction;

[0120] This involves transient changes in temperature over time. A simplified analysis is provided first, and the analysis can be subsequently extended to transient conditions. All of these follow the principle of energy conservation, i.e., thermal equilibrium. This embodiment provides a specific implementation method to address steady-state changes. Specific parameters should be considered in accordance with specific practical applications, including temperature changes, material selection, heat loss, and other factors.

[0121] The data-driven model receives the temperature distribution inside the VCSEL light source under different operating conditions, predicts the temperature conditions at different parts of the VCSEL light source, and continuously verifies and updates the model to ensure that its prediction accuracy continues to improve with the increase in data volume and the enrichment of working scenarios. Ultimately, a reliable temperature model is formed that can be used to monitor and predict the temperature of the VCSEL light source in real time.

[0122] Furthermore, the data-driven model includes data acquisition, data preprocessing, multi-layer temperature prediction, and inverse optimization of prediction losses;

[0123] Data acquisition is used to receive temperature distribution data inside the VCSEL light source under different working conditions;

[0124] Data preprocessing includes data cleaning and normalization;

[0125] Specifically, data cleaning involves calculating the mean and standard deviation of the temperature distribution data inside the VCSEL light source under different working states. Outliers are removed by setting a threshold (for example, based on the 3-times standard deviation principle, determining whether the temperature inside the VCSEL light source under a certain working state of a sample meets the 3-times standard deviation principle. If so, the sample is considered an outlier and the entire row of data corresponding to the sample is discarded). This results in a cleaned data set.

[0126] Furthermore, eigenvalues ​​are extracted from the temperature distribution data set inside the cleaned VCSEL light source, and the eigenvectors corresponding to the eigenvalues ​​are calculated to construct a characteristic matrix. The normalized eigenvalues ​​are calculated based on the minimum-maximum normalization of the input characteristic matrix.

[0127] Multi-layer temperature prediction includes a multi-layer perception structure. Assume that the input layer has n nodes (corresponding to the number of temperature features inside the input VCSEL light source), the hidden layer has L layers, and the number of nodes in the zth hidden layer is d z (z=1,2,3,...), the output layer has k nodes;

[0128] Randomly initialize the connection weight matrix W and bias vector b between each layer. For the connection between the zth layer and the z+1th layer, the dimension of the weight matrix W is d z+1 ×d z , bias vector b z The dimension is d z+1 ×z. For example, the weight matrix W0 of the first hidden layer (z = 0, the input layer is considered as layer 0) connects n nodes in the input layer to d1 nodes in the first hidden layer, and its dimension is d1×n;

[0129] The temperature distribution data inside the VCSEL light source is calculated and kernel operation is performed to further predict the temperature distribution data inside the VCSEL light source. The kernel operation expression is as follows:

[0130] T1=f(W z-1 Y+b z-1 )

[0131] Among them, T1 is the predicted value of the temperature distribution data inside the VCSEL light source, f(~) is the activation function, Y is the real-time temperature data of each layer inside the VCSEL light source, b is the bias, W z-1 is the weight of the hidden layer z-1, b z-1 is the bias of the hidden layer z-1;

[0132] The temperature distribution data set inside the entire VCSEL light source is calculated through the multi-layer temperature prediction process, and a predicted output multi-layer temperature matrix is ​​obtained. Its dimension is the same as the target temperature value matrix, and each row corresponds to the predicted temperature value vector of a sample;

[0133] The loss is calculated by predicting the temperature inside the VCSEL light source. The loss is obtained by calculating the difference between the predicted temperature value and the actual temperature value. The data-driven model is then corrected by loss gradient compensation. The loss calculation expression is as follows:

[0134]

[0135] Where MES is the loss value, a is the number of VCSEL light source layers, b is the temperature of the corresponding layer, T' a,b To predict the output of the multi-layer temperature matrix, T a,b is the actual temperature matrix, M is the total number of data;

[0136] The weight of each part of the data-driven model is adjusted through chaining, and the calculation expression is as follows:

[0137]

[0138] The weights of each layer are calculated in turn, and the weights of each layer are updated. The above steps are repeated for a certain number of iterations until the stopping condition is met (such as the loss function value of the validation set no longer decreases significantly).

[0139] Example 3:

[0140] Adaptive layer, adaptively adjusts the VCSEL light source intensity through temperature stratification;

[0141] An optical power detector is installed at the output end of the light combiner 14. The optical power detector uses a photodiode with high sensitivity, wide dynamic range, and fast response characteristics. It can accurately measure the optical power output from the VCSEL light source after it is combined by the light combiner and convert the optical power signal into an electrical signal in real time. A corresponding signal conditioning circuit is designed to amplify and filter the weak electrical signal output by the photodiode. The processed signal is input into the subsequent control module to ensure that the optical power signal received by the control module is of sufficient accuracy and stability to accurately determine the difference between the real-time change in optical power and the preset target optical power.

[0142] The adaptive layer also includes a control strategy output unit, a VCSEL light source control and a control unit;

[0143] The control strategy output unit is used to divide the temperature range into multiple stratified intervals and adaptively adjust the algorithm according to the performance changes of the VCSEL light source in different temperature ranges and application requirements;

[0144] Furthermore, according to the performance changes of the VCSEL light source in different temperature ranges and application requirements, the temperature range is divided into multiple hierarchical intervals. For example: the low-temperature interval (T < T1, where T1 is a certain low-temperature threshold). In this interval, the performance of the light source is relatively stable, but the optical output power may be slightly low, and it is necessary to appropriately increase the drive current to enhance the light intensity; the medium-temperature interval (T1 ≤ T < T2), which is the main interval for the normal operation of the light source, and the optical power and efficiency are in a good balance state, and relatively stable drive conditions can be maintained; the high-temperature interval (T ≥ T2). When the temperature is too high, the performance of the light source may drop sharply or even pose a risk of damage. At this time, it is necessary to quickly reduce the drive current or take measures to enhance heat dissipation to protect the light source and maintain a reasonable optical output intensity;

[0145] The adaptive adjustment algorithm is based on the above temperature hierarchical intervals and uses a fuzzy control algorithm to achieve the adaptive adjustment of the intensity of the VCSEL light source. Taking the hierarchical interval where the temperature value predicted by the temperature model is located as the judgment basis, when in the low-temperature interval, the PID controller gradually increases the drive current applied to the metal electrode 1 and the lower metal electrode 8 according to the deviation between the optical power and the target optical power, so that the injection current of the VCSEL light source increases, thereby enhancing the optical output intensity. At the same time, the change of the optical power is monitored in real time to prevent damage to the light source caused by over-driving; in the medium-temperature interval, a relatively stable drive current is maintained, and only fine-tuning is performed according to the small fluctuations of the optical power to ensure the stability of the optical output; after entering the high-temperature interval, the light source fuzzy control algorithm comprehensively judges based on multiple factors such as the degree to which the temperature exceeds the high-temperature threshold and the decrease of the optical power, quickly reduces the drive current, and at the same time sends a control signal to enhance heat dissipation to the heat dissipation system;

[0146] Send a control signal to enhance heat dissipation. After the temperature drops back to the appropriate interval, gradually resume the normal drive state and optical output intensity. Through this hierarchical adaptive control strategy, the intensity of the VCSEL light source can be reasonably adjusted and stably output in different temperature environments.

[0147] The light source fuzzy control algorithm receives the difference between the predicted temperature of the VCSEL light source and the pre-set high-temperature threshold as the basic data, uses the temperature change rate and the optical power deviation as the input data of the optical fuzzy control algorithm, and according to fuzzy reasoning, outputs the fuzzyization of the VCSEL light source temperature variable and optical power variable, automatically compares with the fuzzy control rules, and adjusts and controls the heat dissipation components and output light of the VCSEL light source;

[0148] Furthermore, the rule base includes but is not limited to: when the temperature deviation is large and still increasing rapidly, and at the same time the optical power deviation is also large and the optical power is seriously insufficient (much lower than the target optical power), it is necessary to greatly reduce the drive current to protect the light source, and greatly increase the rotation speed of the cooling fan to enhance heat dissipation and quickly reduce the temperature;

[0149] If the temperature deviation is small, the temperature change rate is close to zero (stable at a level slightly above the high temperature threshold), and the optical power is only slightly below the target value, then you only need to slightly reduce the drive current and slightly increase the cooling fan speed to fine-tune the temperature and optical power.

[0150] When the temperature deviation is negative (below the high temperature threshold and the deviation is small), but the temperature drops rapidly, the optical power is higher than the target value and the deviation is moderate, the drive current can be slightly increased to increase the optical power and the cooling fan speed can be slightly reduced to prevent the temperature from being too low and affecting the light source performance.

[0151] Furthermore, fuzzy reasoning is to infer the fuzzy value of the output variable through the Mamdani reasoning method based on the fuzzy value of the input variable and the fuzzy control rule base, for example:

[0152] For the precise values ​​of given input variables (specific values ​​of temperature deviation, temperature deviation change rate and optical power deviation obtained through actual measurement), we first determine their membership in their respective corresponding fuzzy subsets through the previously defined temperature difference and optical power deviation. The membership is understood as the deviation rate.

[0153] Then, for each fuzzy control rule, the input temperature variable membership is ANDed (usually taking the minimum value) to obtain the rule's activation strength (or trigger strength). For example, for a rule, if the temperature deviation has a membership of 0.6 in the "positive large (PB)" subset, the temperature deviation change rate has a membership of 0.8 in the "positive large (PB)" subset, and the optical power deviation has a membership of 0.7 in the "negative large (NB)" subset, then the activation strength of this rule is the minimum membership of 0.6.

[0154] Then, according to the incentive strength of each rule, the conclusion part of the rule (the fuzzy subset corresponding to the output variable) is “truncated” to obtain the fuzzy implication relationship of the output variable corresponding to each rule.

[0155] Each rule is based on the thermal characteristics and optical power characteristics of the VCSEL light source to establish a fuzzy control rule base. The rule format is usually "IF (condition) THEN (conclusion)";

[0156] Finally, the fuzzy implication relations of the output variables corresponding to all rules are “OR” operated (usually by taking the maximum value) to obtain the fuzzy set of the total output variables;

[0157] The output variable obtained through fuzzy inference is in the form of a fuzzy set, which needs to be converted into a precise value for practical application in VCSEL light source temperature regulation or optical power regulation. The defuzzification calculation expression is as follows:

[0158]

[0159] Where λ is the membership coefficient, ΔQ(x) is the temperature difference or optical power deviation control adjustment control rule, x is the input VCSEL light source temperature adjustment value, X is the total range of the temperature adjustment value, and Δ is the defuzzified VCSEL light source control instruction;

[0160] The defuzzification calculation includes the numerator part ∫ x∈X x·λΔQ(x)dx and the denominator ∫ x∈X λΔQ(x)dx;

[0161] Numerator part ∫ x∈X x·λΔQ(x)dx considers each possible control variable value and its degree of conformity under the current fuzzy inference result, and then performs weighted summation;

[0162] Furthermore, the denominator ∫ x∈X λΔQ(x)dx integrates the membership coefficients within the entire value range, which essentially calculates the sum of the membership coefficients of all possible value points. It plays a normalization role, so that the defuzzified result can reasonably fall within the value range of the output variable, ensuring that the value of the calculated control instruction is consistent with the actual control quantity range and reasonably reflects the fuzzy reasoning result.

[0163] By dividing the numerator by the denominator, we determine the precise value that best represents the overall state of the fuzzy set. This value is the control instruction used to actually adjust the temperature or optical power of the VCSEL light source. For example, the precise value of the drive current adjustment can be directly applied to the VCSEL light source's drive circuit to achieve accurate current regulation, thereby affecting the temperature and optical power. The precise value of the cooling fan speed adjustment can control the actual speed of the cooling fan, assisting in temperature regulation.

[0164] The calculation expression for VCSEL light source control is as follows:

[0165]

[0166] e(t)=C on -C out (t)

[0167] Among them, u(t) is the driving current adjustment value on the VCSEL light source electrode, K p To adjust the light intensity proportional coefficient, K i To adjust the light intensity integral coefficient, K d is the differential coefficient for adjusting the light intensity, e(t) is the defuzzified VCSEL light source control instruction, t is the time, C on is the target light intensity, C out(t) is the predicted light intensity;

[0168] The further defuzzified VCSEL light source control instruction only represents a logic, and how to output the VCSEL light source control instruction is output by the adaptive layer.

[0169] The detailed design of the sensing layer and adaptive layer can effectively monitor and predict the temperature of the VCSEL light source and adaptively adjust the light source intensity according to the temperature conditions, thereby improving the working stability and performance of the VCSEL light source under various complex working conditions.

[0170] It is important to note that the construction and arrangement of the present application shown in a number of different exemplary embodiments are merely illustrative. Although only two embodiments are described in detail in this disclosure, it should be readily understood by those who refer to this disclosure that, without departing substantially from the novel teachings and advantages of the subject matter described in this application, many modifications are possible, for example, the size, scale, structure, shape and proportion of various elements, as well as parameter values ​​(e.g., temperature, pressure, etc.), mounting arrangements, use of materials, color, directional changes, etc. For example, an element shown as integrally formed can be composed of multiple parts or elements, the position of the element can be inverted or otherwise changed, and the nature or number or position of the discrete elements can be altered or changed. Therefore, all such modifications are intended to be included within the scope of the present invention. The order or sequence of any process or method steps can be changed or reordered according to alternative embodiments. Any "device plus function" clause is intended to cover the structure of the execution function described herein, and is not only structurally equivalent but also equivalent structures. Without departing from the scope of the present invention, other replacements, modifications, changes and omissions can be made in the design, operating conditions and arrangement of the exemplary embodiments. Therefore, the invention is not limited to the specific embodiments, but extends to various modifications that still fall within the scope of the appended claims.

[0171] Additionally, in order to provide a concise description of exemplary embodiments, all features of an actual embodiment (ie, those features that are not relevant to the best mode presently contemplated for carrying out the invention or those that are not relevant to implementing the invention) may not be described.

[0172] It should be understood that in the development of any actual embodiment, as in any engineering or design project, numerous implementation-specific decisions may be made. Such a development effort may be complex and time-consuming, but for those of ordinary skill having the benefit of this disclosure, the development effort will be a routine task of design, fabrication, and production without undue experimentation.

[0173] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A high temperature adaptability VCSEL light source, characterized in that: include: One end of the light beam combiner (14) is connected to a metal electrode (1) via a light coupling optical fiber (13); one end of the metal electrode (1) is sequentially provided with a passivation layer (2), a P-type DBR reflector (3), an oxide layer (4), a multi-quantum well active region (5), an N-type DBR reflector (6), an N-type substrate layer (7), and a lower metal electrode (8); The oxide layer (4) is located at the node of the VCSEL standing wave field, and an oxide layer is provided on the upper and lower sides of the active area of ​​the high-temperature vertical cavity surface emitting laser, and the component content of the lower oxide layer is higher than that of the upper oxide layer. One end of the light beam combiner (14) is parallelly provided with a light coupling optical fiber (13) connected to the phase matching layer (10); The sensing layer monitors and predicts the temperature of the VCSEL light source by establishing a temperature model; The data-driven model includes data acquisition, data preprocessing, multi-layer temperature prediction, and inverse optimization of predicted losses; Data acquisition is used to receive temperature distribution data inside the VCSEL light source under different working conditions; Data preprocessing includes data cleaning and normalization; Multi-layer temperature prediction includes a multi-layer perception structure. Assume that the input layer has n temperature feature nodes inside the VCSEL light source, the hidden layer has L layers, and the number of hidden layer nodes in the zth layer is d z , where z = 1, 2, 3, ..., and the output layer has k nodes; Randomly initialize the connection weight matrix W and bias vector b between each layer. For the connection between the zth layer and the z+1th layer, the dimension of the weight matrix W is d z+1 ×d z , bias vector b z Dimension d z+1 ×z; The temperature distribution data inside the VCSEL light source is calculated and kernel operation is performed to further predict the temperature distribution data inside the VCSEL light source. The kernel operation expression is as follows: T1=f(W z-1 Y+b z-1 ) Among them, T1 is the predicted value of the temperature distribution data inside the VCSEL light source, f(~) is the activation function, Y is the real-time temperature data of each layer inside the VCSEL light source, b is the bias, W z-1 is the weight of the hidden layer z-1, b z-1 is the bias of the hidden layer z-1; The temperature distribution data set inside the entire VCSEL light source is calculated through a multi-layer temperature prediction process to obtain a predicted output multi-layer temperature matrix. Its dimension is the same as the target temperature value matrix, and each row corresponds to a predicted temperature value vector of a sample. Adaptive layer, adaptively adjusts the VCSEL light source intensity through temperature stratification.

2. The high temperature adaptability VCSEL light source according to claim 1, characterized in that: The VCSEL light source comprises a low-temperature vertical cavity surface emitting laser and a high-temperature vertical cavity surface emitting laser connected in parallel, wherein the low-temperature vertical cavity surface emitting laser and the high-temperature vertical cavity surface emitting laser are filled with a BCB material (12); The high-temperature vertical cavity surface emitting laser comprises: One end of the passivation layer (2) is connected to the metal electrode (1), and the other end is provided with a P-type DBR reflector (3); An oxide layer (4) is provided at the other end of the P-type DBR reflector (3), and the other end of the oxide layer (4) is connected to an N-type DBR reflector (6) via a multi-quantum well active region (5); An N-type substrate layer (7) is provided at the other end of the N-type DBR reflector (6), and the N-type substrate layer (7) is connected to a heat sink (11) via a lower metal electrode (8).

3. The high temperature adaptability VCSEL light source according to claim 2, characterized in that: The phase matching layer (10) is located at the top of the P-type DBR reflector (3) in the vertical cavity surface emitting laser operating at low temperature and the vertical cavity surface emitting laser operating at high temperature; A multi-quantum well active region (5) and a P-type DBR reflector (3) are respectively arranged at both ends of the spacer layer (9); The spacer layer (9) is also arranged between the multi-quantum well active region (5) and the N-type DBR reflector (6); The spacer layer (9) is located between the multi-quantum well active region (5) and the P-type DBR reflector (3) and the N-type DBR reflector (6), and is used to adjust the length of the resonant cavity.

4. The high temperature adaptability VCSEL light source according to claim 3, characterized in that: The sensing layer includes a temperature sensor, a temperature model and a data-driven model; The temperature model detects the temperature distribution prediction data inside the VCSEL light source under different working conditions by conducting multi-layer heat conduction on the VCSEL light source and the heat exchange with the external environment; Methods for obtaining the temperature distribution inside the VCSEL light source include: S1. Divide the VCSEL light source into N layers from bottom to top according to its physical structure, marked as layer 1 to layer N respectively; S2. Obtain heat flux density vector data and simplify it into one-dimensional heat conduction. Among them, Q is the heat flux density of each layer of the VCSEL light source, k is the thermal conductivity, T is the temperature of the VCSEL light source material layer, and x is the one-dimensional heat conduction direction; S3. Output the temperature distribution prediction data inside the VCSEL light source through the data-driven model.

5. The high temperature adaptability VCSEL light source according to claim 4, characterized in that: The loss is calculated by predicting the temperature inside the VCSEL light source. The loss is obtained by calculating the difference between the predicted temperature value and the actual temperature value. The data-driven model is then corrected by loss gradient compensation. The loss calculation expression is as follows: Where MES is the loss value between the predicted temperature and the actual temperature inside the VCSEL light source, a is the number of VCSEL light source layers, b is the temperature of the corresponding layer, T' a,b To predict the output of the multi-layer temperature matrix, T a,b is the actual temperature matrix, and M is the total number of data.

6. The high temperature adaptability VCSEL light source according to claim 5, characterized in that: The weight of each part of the data-driven model is corrected through chaining, and the calculation expression is as follows: The weight of each layer is calculated in turn, and the weight of each layer is updated. The above steps are repeated for a certain number of iterations until the stopping condition is met.

7. The high temperature adaptability VCSEL light source according to claim 6, characterized in that: The adaptive layer includes fuzzy reasoning, defuzzification, VCSEL light source control and control execution unit; The light source fuzzy control receives the difference between the predicted VCSEL light source temperature and the pre-set high temperature threshold as the basic data, and uses the temperature change rate and optical power deviation as the input data of the light fuzzy control algorithm. Based on fuzzy reasoning, the fuzzified output of the VCSEL light source temperature variable and optical power variable is automatically compared with the fuzzy control rules. After defuzzification, the VCSEL light source heat dissipation components and output light are adjusted and controlled.

8. The high temperature adaptability VCSEL light source according to claim 7, characterized in that: The calculation expression of the defuzzification is as follows: Where λ is the membership coefficient, ΔQ(x) is the temperature difference or optical power deviation control adjustment control rule, x is the input VCSEL light source temperature adjustment value, X is the total range of the temperature adjustment value, and Δ is the defuzzified VCSEL light source control instruction.

9. The high temperature adaptability VCSEL light source according to claim 8, characterized in that: The calculation expression for VCSEL light source control is as follows: e(t)=C on -C out (t) Among them, u(t) is the driving current adjustment value on the VCSEL light source electrode, K p To adjust the light intensity proportional coefficient, K i To adjust the light intensity integral coefficient, K d is the differential coefficient for adjusting the light intensity, e(t) is the defuzzified VCSEL light source control instruction, t is the time, C on is the target light intensity, C out (t) is the predicted light intensity The control execution unit is used to receive VCSEL light source control instructions and control the VCSEL light source.

Images