Electric polarization controller

Abstract

The invention provides an electric polarization controller. The electric polarization controller comprises a converter and a polarization assembly, the polarization assembly comprises a controller and a polarizer; the converter is used for generating a corresponding adjustment instruction according to the received generation instruction and sending the adjustment instruction to the controller; the controller is used for adjusting corresponding modulation parameters of the polarizer according to the received adjustment instruction; and the polarizer is used for generating a corresponding polarization state according to the received light beam and the current modulation parameter. According to the invention, the required polarization state can be accurately and automatically generated.

Description

Technical Field

[0001] This application relates to the field of polarization controller technology, and more particularly to an electric polarization controller. Background Technology

[0002] An electric polarization controller (EPC) is a dynamic optical device that uses an electric field or current to control the polarization state of a beam in real time. Its core function is to convert any input polarization state (such as linear polarization, circular polarization, elliptic polarization, etc.) into any desired output polarization state (i.e., target polarization state).

[0003] However, when using existing EPC technologies, if the EPC is required to generate the desired polarization state, it is usually necessary to adjust the internal parameters of the EPC. For example, in an optical fiber extrusion-type electric polarization controller, the driving voltage of each set of piezoelectric ceramic actuators in the controller needs to be adjusted. By applying stress to the single-mode fiber through each set of piezoelectric ceramic actuators, the fiber generates a controllable birefringence effect, thereby generating the desired polarization state.

[0004] However, when adjusting the internal parameters of an EPC, staff usually adjust the parameters manually based on their own experience. Therefore, it usually takes multiple adjustments to obtain the required polarization state. The operation is cumbersome, takes a long time, and the accuracy is difficult to guarantee. Summary of the Invention

[0005] In view of this, this application provides an electric polarization controller that can accurately generate the desired polarization state.

[0006] This application provides an electric polarization controller, including: a converter and a polarization component; The polarization component includes a controller and a polarizer; The converter is configured to generate a corresponding adjustment instruction based on the received generation instruction, and send the adjustment instruction to the controller; The controller is configured to adjust the corresponding modulation parameters of the polarizer according to the received adjustment instructions; The polarizer is used to generate a corresponding polarization state based on the received beam and the current modulation parameters.

[0007] Furthermore, the polarization component is a waveplate-type electric polarization controller with multiple waveplates, a paddle-type electric polarization controller with multiple blades, or an optical fiber extrusion-type electric polarization controller with multiple piezoelectric elements.

[0008] Furthermore, the generation instruction includes multiple polarization state data; the polarization state data are Stokes parameters.

[0009] Furthermore, the adjustment command includes multiple modulation parameters; the modulation parameters are angle parameters or voltage parameters.

[0010] Furthermore, when the polarization component is a waveplate-type electric polarization controller, the modulation parameter is the rotation angle of each waveplate. When the polarization component is a paddle-type electric polarization controller, the modulation parameter is the deflection angle of each paddle. When the polarization component is an optical fiber extrusion electric polarization controller, the modulation parameter is the driving voltage of each piezoelectric element.

[0011] Furthermore, the correspondence between generation instructions and adjustment instructions is pre-set in the converter.

[0012] Furthermore, the polarization state data of the polarization component is collected multiple times to obtain the polarization component dataset; The converter is trained using a neural network based on the polarization component dataset to obtain the correspondence between generation instructions and adjustment instructions.

[0013] Furthermore, the neural network is a fully connected neural network, a deep neural network, or a Transformer network.

[0014] Furthermore, when the neural network is a deep neural network, the network structure includes an input layer, an output layer, and multiple hidden layers.

[0015] Furthermore, the input and output layers of the neural network have a dimension of 3, and the number of hidden layers is N, where N is a natural number greater than 1.

[0016] As can be seen from the above technical solution, in the electric polarization controller of this application, since the converter can generate a corresponding adjustment command according to the received generation command, the controller can adjust the corresponding modulation parameters of the polarizer according to the received adjustment command, and the polarizer can generate a corresponding polarization state according to the received beam and the current modulation parameters, thereby accurately generating the required polarization state. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the structure of an electric polarization controller in a specific embodiment of this application.

[0018] Figure 2 This is a schematic diagram of the neural network structure in a specific embodiment of this application. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0020] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0021] Figure 1 This is a schematic diagram of the structure of the electric polarization controller in a specific embodiment of this application. Figure 1 As shown, the electric polarization control device in this application may include: a converter 11 and a polarization component 12; The polarization component 12 includes a controller 121 and a polarizer 122; The converter 11 is used to generate a corresponding adjustment instruction according to the received generation instruction, and send the adjustment instruction to the controller 121; The controller 121 is used to adjust the corresponding modulation parameters of the polarizer 122 according to the received adjustment command; The polarizer 122 is used to generate a corresponding polarization state based on the received beam and the current modulation parameters.

[0022] In the technical solution of this application, when it is necessary to generate the required polarization state, a corresponding generation command can be output to the aforementioned electric polarization control device to inform it of the required polarization state; the converter can generate a corresponding adjustment command according to the generation command, and then output the adjustment command to the controller in the polarization component; the controller can adjust the corresponding modulation parameters (e.g., angle, voltage, etc.) of the polarizer in the polarization component according to the adjustment command; and when the polarizer receives a beam of light input from the outside (e.g., a laser), it can generate the corresponding polarization state according to the received beam and the current modulation parameters.

[0023] In addition, in the technical solution of this application, various polarization components can be used according to the needs of actual application scenarios.

[0024] For example, as an example, in a specific embodiment of this application, the polarization component may be a waveplate-type electric polarization controller with multiple waveplates, a paddle-type electric polarization controller with multiple blades, or an optical fiber extrusion-type electric polarization controller with multiple piezoelectric elements.

[0025] For example, in one specific embodiment of this application, the polarization component may be a waveplate-type electric polarization controller with two, three, or four waveplates. When the controller (e.g., a servo motor) receives an adjustment command, it can adjust the rotation direction of each waveplate in the polarizer of the waveplate-type electric polarization controller according to the adjustment command, thereby adjusting the corresponding modulation parameters.

[0026] For example, in one specific embodiment of this application, the polarization component may also be a paddle-type electric polarization controller with 2, 3 or 4 paddles. The controller (e.g., a servo motor) can adjust the deflection angle of each paddle in the polarizer of the paddle-type electric polarization controller according to the received adjustment command, thereby adjusting the corresponding modulation parameters.

[0027] For example, in one specific embodiment of this application, the polarization component may also be an optical fiber extrusion electric polarization controller with two, three or four piezoelectric elements. The controller (e.g., a voltage controller) can adjust the driving voltage of each piezoelectric element (e.g., piezoelectric ceramic) in the polarizer of the optical fiber extrusion electric polarization controller according to the received adjustment command, thereby adjusting the corresponding modulation parameters.

[0028] Additionally, as an example, in one specific embodiment of this application, the generation instructions may include multiple polarization state data.

[0029] For example, as an example, in one specific embodiment of this application, the polarization state data may be Stokes parameters.

[0030] Since Stokes parameters can quantitatively describe the desired optical polarization state, when a certain polarization state needs to be generated, multiple Stokes parameters of that polarization state can be input to the converter through a generation command, and the converter can then generate corresponding adjustment commands based on the generation command.

[0031] Additionally, as an example, in one specific embodiment of this application, the adjustment instruction may include multiple modulation parameters.

[0032] For example, as an example, in one specific embodiment of this application, the modulation parameter may be a parameter such as an angle parameter or a voltage parameter.

[0033] For example, as an example, when the polarization component is a waveplate-type electric polarization controller, the modulation parameter can be the rotation angle of each waveplate.

[0034] For example, when the polarization component is a paddle-type electric polarization controller, the modulation parameter can be the deflection angle of each paddle.

[0035] For example, when the polarization component is a fiber-optic extrusion electric polarization controller, the modulation parameters can be the driving voltages of the individual piezoelectric elements.

[0036] Therefore, when the converter outputs an adjustment command to the controller in the polarization component, the controller can adjust the corresponding modulation parameters in the polarizer according to the modulation parameters in the adjustment command; when the polarizer receives a beam of light input from the outside (e.g., a laser), it can generate the corresponding polarization state according to the received beam and the current modulation parameters.

[0037] Furthermore, as an example, in one specific embodiment of this application, a correspondence between generation instructions and adjustment instructions can be pre-set in the converter. Therefore, when the converter receives a generation instruction, it can generate an adjustment instruction corresponding to the generation instruction according to the pre-set correspondence.

[0038] In addition, in the technical solution of this application, the correspondence between generation instructions and adjustment instructions can be pre-set in the converter through various specific implementation methods according to the needs of actual applications.

[0039] For example, as an example, in a specific embodiment of this application, the polarization state data of the polarization component can be collected multiple times to obtain a polarization component dataset. Then, based on the polarization component dataset, a neural network can be used to train the converter to obtain the correspondence between the generation instructions and the adjustment instructions.

[0040] For example, the modulation parameters of the polarizer can be set first through the controller in the polarization component (e.g., setting the rotation direction of each waveplate in a waveplate-type electric polarization controller, setting the deflection angle of each paddle in a paddle-type electric polarization controller, or setting the driving voltage of each piezoelectric element in a fiber extrusion-type electric polarization controller, etc.). Then, a laser beam is input into the polarizer, and the polarization state output by the polarizer is acquired using a measuring instrument (e.g., a polarization analyzer), thereby obtaining the polarization state data of the polarization component (e.g., the Stokes parameters corresponding to the polarization state). This acquisition process is repeated multiple times, and the obtained multiple polarization state data and corresponding modulation parameters are used as the polarization component dataset.

[0041] After obtaining the polarization component dataset, the dataset can be divided into a training set, a validation set, and a test set according to a preset ratio (e.g., 6:2:2). Then, based on the training set, validation set, and test set, a neural network is used for training to obtain the correspondence between modulation parameters and polarization state data, thereby obtaining the correspondence between generation commands and adjustment commands.

[0042] For example, in a specific embodiment of this application, when the polarization component is a waveplate-type electric polarization controller, and the modulation parameters are the rotation angles of each waveplate, and the polarization state data are Stokes parameters, the rotation direction (i.e., modulation parameters) of each waveplate in the polarizer can be set first through the controller in the polarization component. Then, a laser beam is input to the polarizer, and the polarization state output by the polarizer is acquired using a polarization analyzer, thereby obtaining the Stokes parameters corresponding to the polarization state output by the polarization component. This acquisition process is repeated n times. Since each set of modulation parameters in the acquisition process corresponds to a set of Stokes parameters, n acquisition processes can yield n sets of modulation parameters and n sets of Stokes parameters. Based on these n sets of modulation parameters and n sets of Stokes parameters, the polarization component dataset of the waveplate-type electric polarization controller can be obtained.

[0043] The polarization component dataset is divided into training, validation, and test sets according to a preset ratio (e.g., 6:2:2). Then, a neural network is trained using these sets to obtain the correspondence between modulation parameters and Stokes parameters. Since the generation instructions can include Stokes parameters, and the adjustment instructions can include modulation parameters, the correspondence between generation and adjustment instructions can be derived from the training-derived modulation parameter-Stokes parameter correspondence. Therefore, the correspondence between generation and adjustment instructions can be pre-set in the converter, allowing it to generate corresponding adjustment instructions upon receiving a generation instruction.

[0044] Similarly, when the polarization component is a paddle-type electric polarization controller, an optical fiber extrusion electric polarization controller, or other polarization components, the correspondence between generation instructions and adjustment instructions can also be preset in the converter using the above method, which will not be elaborated here.

[0045] Additionally, as an example, in one specific embodiment of this application, the neural network may be a fully connected neural network, a deep neural network (DNN), a Transformer network, or other suitable neural networks.

[0046] For example, as an example, in a specific embodiment of this application, when the neural network is a DNN, the network structure may include: an input layer, an output layer, and multiple hidden layers.

[0047] In the technical solution of this application, the dimensions of the input layer and output layer of the neural network, as well as the number of hidden layers, can be preset according to the needs of actual application.

[0048] For example, as an example, in a specific embodiment of this application, the input layer and output layer of the neural network have a dimension of 3, and the number of hidden layers is N, where N is a natural number greater than 1.

[0049] For example, in a specific embodiment of this application, when the neural network is a DNN, the polarization component is a paddle-type electric polarization controller with three paddles, and the modulation parameters are the deflection angles of the three paddles (A1, A2, A3), and the polarization state data are three Stokes parameters (S1, S2, S3), the input and output layers of the MLP or DNN can both have a dimension of 3, and the hidden layers are N layers (N=5), such as... Figure 2 As shown.

[0050] The 3-dimensional input layer corresponds to 3 modulation parameters (A1, A2, A3), and the 3-dimensional output layer corresponds to 3 Stokes parameters (S1, S2, S3). The dimensions of the 5 hidden layers can be 512, 512, 256, 128, and 32, respectively.

[0051] In another specific embodiment of this application, each hidden layer undergoes Batch Norm normalization and uses the LeakyReLU activation function for nonlinear operations.

[0052] Furthermore, in this MLP, mean squared error (MSE) can be used as the loss function; the Adam optimizer can be used for optimization; the initial learning rate can be set to 0.001 and dynamically adjusted; in addition, a Dropout layer can be introduced during training to prevent overfitting.

[0053] For example, in a specific embodiment of this application, when the neural network is a Transformer model, the polarization component is a paddle-type electric polarization controller with three paddles, and the modulation parameters are the deflection angles (A1, A2, A3) of the three paddles, and the polarization state data are three Stokes parameters (S1, S2, S3), an embedding layer can be used to map the deflection angle information to a high-dimensional space while preserving angle specificity; then, position encoding is used to enhance the Transformer model's ability to perceive the order of the paddle deflection angles; subsequently, five encoder layers can be used to capture the nonlinear modulation relationship between deflection angles; finally, all feature information can be dimensionality reduced and concatenated, and then mapped onto the three Stokes parameters (S1, S2, S3) using a fully connected neural network.

[0054] Therefore, the aforementioned neural network can be used to train the converter to obtain the correspondence between the generated instructions and the adjustment instructions.

[0055] After conducting multiple experiments, the results show that by training the converter with the aforementioned neural network to obtain the correspondence between the generation command and the adjustment command, and setting this correspondence in the converter, the converter can generate a very accurate adjustment command based on the received generation command (for example, the accuracy probability within ±5% has exceeded 99.9%). Therefore, after the controller adjusts the modulation parameters of the polarizer according to the adjustment command, the polarizer can accurately generate the polarization state corresponding to the generation command based on the received beam.

[0056] In summary, in the technical solution of this application, since the converter can generate a corresponding adjustment instruction based on the received generation instruction, the controller can adjust the corresponding modulation parameters of the polarizer based on the received adjustment instruction, and the polarizer can generate a corresponding polarization state based on the received beam and the current modulation parameters, the required polarization state can be generated accurately and automatically.

[0057] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0058] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. An electric polarization controller, characterized in that, The electric polarization controller includes: a converter and a polarization component; The polarization component includes a controller and a polarizer; The converter is configured to generate a corresponding adjustment instruction based on the received generation instruction, and send the adjustment instruction to the controller; The controller is configured to adjust the corresponding modulation parameters of the polarizer according to the received adjustment instructions; The polarizer is used to generate a corresponding polarization state based on the received beam and the current modulation parameters.

2. The electric polarization controller according to claim 1, characterized in that: The polarization component is a waveplate-type electric polarization controller with multiple waveplates, a paddle-type electric polarization controller with multiple blades, or an optical fiber extrusion-type electric polarization controller with multiple piezoelectric elements.

3. The electric polarization controller according to claim 2, characterized in that: The generation instruction includes multiple polarization state data; the polarization state data are Stokes parameters.

4. The electric polarization controller according to claim 3, characterized in that: The adjustment command includes multiple modulation parameters; the modulation parameters are angle parameters or voltage parameters.

5. The electric polarization controller according to claim 4, characterized in that, When the polarization component is a waveplate-type electric polarization controller, the modulation parameter is the rotation angle of each waveplate. When the polarization component is a paddle-type electric polarization controller, the modulation parameter is the deflection angle of each paddle. When the polarization component is an optical fiber extrusion electric polarization controller, the modulation parameter is the driving voltage of each piezoelectric element.

6. The electric polarization controller according to claim 1, characterized in that: The correspondence between generation instructions and adjustment instructions is preset in the converter.

7. The electric polarization controller according to claim 6, characterized in that: The polarization state data of the polarization component is collected multiple times to obtain the polarization component dataset; The converter is trained using a neural network based on the polarization component dataset to obtain the correspondence between generation instructions and adjustment instructions.

8. The electric polarization controller according to claim 7, characterized in that: The neural network is a fully connected neural network, a deep neural network, or a Transformer network.

9. The electric polarization controller according to claim 8, characterized in that: When the neural network is a deep neural network, the network structure includes: an input layer, an output layer, and multiple hidden layers.

10. The electric polarization controller according to claim 9, characterized in that: The neural network has an input layer and an output layer of dimension 3, and the hidden layer has N layers, where N is a natural number greater than 1.

Images