Dynamic configuration device and method of integrated optical processor chip

Abstract

The invention belongs to the field of integrated photonics, and discloses a dynamic configuration device and method of an integrated optical processor chip. The configuration device comprises a signal acquisition module and a feedback configuration module, and the signal acquisition module detects an output signal of the integrated optical processor chip, processes the output signal into an electric signal and transmits the electric signal to the feedback configuration module; and the feedback configuration module processes and analyzes the electric signal based on a hybrid cooperation algorithm to generate a feedback control signal, and the feedback control signal is applied to a regulator of the integrated optical processor chip to influence the state of the regulator so as to realize dynamic configuration of the integrated optical processor chip. Compared with existing manual configuration, offline design and other methods, the method realizes real-time state mapping and dynamic configuration of the integrated optical processor chip, is a key technology for promoting practical application of the integrated optical processor chip, and is worthy of popularization and application.

Description

Technical Field

[0001] The present invention belongs to the field of integrated photonics and relates to a dynamic configuration device and method for an integrated optical processor chip. Background Art

[0002] The rapid expansion of wireless networks, the Internet of Things, and cloud-based services has imposed higher requirements on the speed, bandwidth, and power consumption of underlying signal processing. Photonic technology, with its advantages of strong electromagnetic interference resistance, low loss, and large working bandwidth, makes optical signal processing an ideal solution to break through the limitations of traditional electronic speeds. In recent years, the breakthrough progress of photonic integration technology has promoted the rapid development of on-chip optical signal processing, which has shown significant advantages in terms of system volume, energy consumption, and manufacturing cost. Currently, integrated optical processors based on multiple material platforms such as silicon-on-insulator, silicon nitride, indium phosphide, and lithium niobate have been technically verified. Compared with traditional electronic processors, integrated optical processors have an ultra-large working bandwidth, flexible tuning ability, and dynamic reconfiguration ability, and thus have broad application prospects in frontier fields such as wireless communication, satellite systems, sensor networks, and advanced radars.

[0003] With the development of multifunctional integrated systems, the dynamicity and stability of integrated optical processors have become key factors in improving system performance. Although existing research has verified the flexible reconfiguration potential of integrated optical processors, most reconfigurable schemes still rely on manual configuration and are difficult to meet the requirements of real-time dynamic configuration in complex application scenarios. The typical structures of integrated optical processors include series-connected microring resonators, high-order coupled-resonator optical waveguides, Mach-Zehnder interferometer-assisted microring resonators, two-dimensional square topological grids, etc. These structures can implement signal processing means such as frequency-domain filtering, time-domain integration, Hilbert transform, and signal compression. However, limited by the cumbersome manual operation, integrated optical processors based on the above structures generally suffer from problems such as insufficient dynamic configuration, difficult collaborative optimization of adjustable parameters, weak anti-environmental disturbance ability, and difficult long-term operation stability. Generally speaking, traditional methods such as manual configuration and off-line design are difficult to meet the requirements of dynamic reconfiguration, multi-objective collaborative optimization, and stable operation of integrated optical processors, seriously restricting their application prospects. Therefore, constructing a supporting dynamic configuration method is crucial for promoting the practical industrial application of integrated optical processors. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to propose a dynamic configuration device and method for an integrated optical processor chip to realize real-time state mapping and dynamic configuration of the integrated optical processor chip.

[0005] To solve the above technical problem, the present invention first proposes a dynamic configuration device for an integrated optical processor chip, including a signal acquisition module and a feedback configuration module. The signal acquisition module is used to detect the output signal of the integrated optical processor chip and process the output signal into an electrical signal for transmission to the feedback configuration module; The feedback configuration module is used to process and analyze the electrical signal based on a hybrid cooperation algorithm to generate a feedback control signal, and the feedback control signal is applied to the regulator of the integrated optical processor chip to change the state of the regulator so as to achieve dynamic configuration of the integrated optical processor chip.

[0006] Further, the feedback configuration module includes an analog-to-digital converter, a digital signal processor, and a digital-to-analog converter; The analog-to-digital converter, the digital signal processor, and the digital-to-analog converter are connected in sequence; The analog-to-digital converter is connected to the output end of the signal acquisition module and is used to convert the electrical signal output by the signal acquisition module into a digital signal; The digital signal processor extracts and analyzes the digital signal characteristics in real time, performs non-linear parameter optimization processing on the digital signal based on a hybrid cooperation algorithm, and outputs a feedback control signal.

[0007] The output end of the digital-to-analog converter is connected to the regulator of the integrated optical processor chip and is used to perform digital-to-analog conversion on the feedback control signal and output an analog feedback control signal to the regulator.

[0008] Furthermore, the feedback configuration module further includes an electrical filter for filtering high-frequency noise of the electrical signal to achieve more accurate state mapping. The input end of the electrical filter is connected to the output end of the signal acquisition module, and the output end of the electrical filter is connected to the input end of the analog-to-digital converter.

[0009] The signal acquisition module includes a photodetector for detecting the optical signal output by the integrated optical processor chip and converting the optical signal into an electrical signal for transmission to the feedback configuration module.

[0010] More optimally, the output end of the photodetector is connected to an electrical amplifier, and the electrical amplifier is used to amplify the power of the input electrical signal and then transmit it to the feedback configuration module.

[0011] The present invention also proposes a dynamic configuration method for the dynamic configuration device of the integrated optical processor chip, including the following steps: The signal acquisition module detects the output signal of the integrated optical processor chip and processes the output signal into an electrical signal for transmission to the feedback configuration module; The feedback configuration module processes and analyzes the electrical signal based on a hybrid collaboration algorithm to generate a feedback control signal, and then applies the feedback control signal to the regulator of the integrated optical processor chip to change the state of the regulator, so as to achieve dynamic configuration of the integrated optical processor chip.

[0012] Further, the feedback configuration module processes and analyzes the electrical signal based on a hybrid collaboration algorithm to generate a feedback control signal, specifically including: Pretreatment process: Perform multi-dimensional signal analysis on the feedback electrical signal, define the performance characteristics of the integrated optical processor chip, and then define the optimization evaluation index to guide the global search process. Global search process: Randomly initialize the feedback control signal; Use the global optimization algorithm to iteratively perform the optimization of the working state of the integrated optical processor chip, search for and calculate the optimal value of the control signal, calculate the optimization evaluation index after a preset number of rounds of iteration, and determine whether the evaluation index meets the convergence condition to the target value. If not, after enabling the multi-level optimization parameter search constraint mechanism, execute the global optimization algorithm again; If so, cache the optimal control signal and enter the multi-objective configuration process. Multi-objective configuration process: Redefine the multi-objective comprehensive evaluation index of the working state of the integrated optical processor chip; Use the multi-objective optimization algorithm to perform the optimization calculation for a preset number of iterations, and then determine whether the optimization result of the multi-objective optimization algorithm meets the convergence condition. If not, after enabling the multi-level optimization parameter search constraint mechanism, execute the multi-objective optimization algorithm again; If so, select different subsequent working modes according to the application requirements; If it is necessary to dynamically reconstruct the working state of the integrated optical processor chip, redefine the comprehensive evaluation index and repeat the multi-objective configuration process. If it is necessary for the integrated optical processor chip to work stably for a long time, enter the adaptive stabilization process. Adaptive stabilization process: After the integrated optical processor chip is configured to the target working state, use the local fine-tuning compensation algorithm to perform real-time fine-tuning optimization, so as to compensate for the working state change and performance degradation caused by thermal fluctuations and mechanical vibrations, and then ensure the continuous and stable operation of the integrated optical processor chip. When the preset stable working duration is reached, determine whether the work is over. If so, return to the multi-objective configuration process; if not, continue to execute the local fine-tuning compensation algorithm.

[0013] Preferably, the hybrid collaboration algorithm is an intelligent optimization algorithm with multi-stage progression, multi-algorithm collaboration, and multi-objective coordination, which integrates the global optimization algorithm, multi-objective optimization algorithm, and local fine-tuning compensation algorithm based on the phased collaboration mechanism.

[0014] Preferably, the global optimization algorithm, multi-objective optimization algorithm, and local fine-tuning compensation algorithm all adopt the following typical algorithms: genetic algorithm, simulated annealing algorithm, particle swarm algorithm, tabu search algorithm, gradient descent algorithm, or differential evolution algorithm.

[0015] Through the above technical solutions conceived by the present invention, compared with the prior art, the following beneficial effects can be achieved: 1. By constructing a closed-loop feedback of the integrated feedback configuration module, the present invention directly feeds the output signal back to the feedback configuration module to perform real-time optimization calculations without any preprocessing. Compared with existing methods such as manual configuration and offline design, the system has stronger real-time and dynamic performance in configuring the integrated optical processor chip.

[0016] 2. The core advantage of the present invention is that it is compatible with different application scenarios and different integrated optical processor chip structures without changing the system architecture. Only by adjusting the algorithm execution logic and parameters can it adapt to different requirements. The architecture of the present invention is simple, has good scalability, strong configuration ability, and a wide range of applications, and is a key technology to promote the practical application of integrated optical processor chips.

[0017] 3. While enhancing the dynamic reconfiguration ability of the integrated optical processor chip, the present invention can effectively improve the long-term stability of the integrated optical processor chip.

[0018] 4. The hybrid cooperation algorithm proposed by the present invention integrates the global optimization algorithm, multi-objective optimization algorithm, and local fine-tuning compensation algorithm based on a phased cooperation mechanism, and introduces a multi-level voltage constraint mechanism and an evaluation index dynamic redefinition strategy, realizing a full-process closed-loop intelligent control from dynamic optimization configuration to environmental disturbance compensation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The technical solutions of the present invention will be further specifically described below in conjunction with the drawings and specific embodiments.

[0020] Figure 1 is the flowchart of the hybrid cooperation algorithm of the present invention.

[0021] Figure 2 is the application scenario diagram of the dynamically configured integrated optical processor chip of the present invention.

[0022] Figure 3 is the overall structure diagram of the integrated optical processor chip in the working scenario and the dynamic configuration device of a specific embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention fall within the scope of protection of the present invention.

[0024] The present invention realizes the state mapping and dynamic configuration of an integrated optical processor chip by constructing a dynamic configuration device for the integrated optical processor chip based on optoelectronic mapping technology. The configuration device includes a signal acquisition module and a feedback configuration module. The signal acquisition module is used to detect the output signal of the integrated optical processor chip and process the output signal into an electrical signal and send it to the feedback configuration module. The feedback electrical signal output by it indicates the working state of the integrated optical processor chip. The feedback configuration module performs processing and analysis on the electrical signal based on a hybrid cooperation algorithm to generate a feedback control signal. The feedback control signal is applied to the regulator of the integrated optical processor chip to change the state of the regulator, so as to realize the dynamic configuration of the integrated optical processor chip.

[0025] Generally speaking, the dynamic configuration method of the dynamic configuration device of the integrated optical processor chip includes the following steps: The signal acquisition module detects the output signal of the integrated optical processor chip and processes the output signal into an electrical signal and sends it to the feedback configuration module; The feedback configuration module performs processing and analysis on the electrical signal based on a hybrid cooperation algorithm, generates, and then applies the feedback control signal to the regulator of the integrated optical processor chip to change the state of the regulator, so as to realize the dynamic configuration of the integrated optical processor chip.

[0026] Figure 2This is an application scenario diagram of the dynamically configured integrated optical processor chip of the present invention. A continuous-wave laser generates a carrier optical signal [located at I]. The carrier optical signal is then input to the optical input terminal of the electro-optic modulator. The electro-optic modulator modulates the electrical signal into the optical domain to generate a modulated signal [located at II]. The modulated signal is input into the integrated optical processor chip to perform optical-domain processing of the signal and output the processed modulated signal [located at III]. The output signal of the integrated optical processing is photoelectrically converted by the signal acquisition module to generate an electrical signal [located at IV], and is fed to the feedback configuration module. In the feedback configuration module, after the electrical signal is filtered by an electrical filter to remove high-frequency noise, it is input into an analog-to-digital converter to be converted into a digital signal to be analyzed. Then, the digital signal processor performs processing on the digital signal including but not limited to fast Fourier transform, analyzes the current state of the integrated optical processor chip, and generates a feedback control signal based on the hybrid cooperation algorithm. Then, the digital-to-analog converter converts the feedback control signal into an analog feedback control signal and applies it to the regulator of the integrated optical processor chip to change the state of the regulator, thereby reconstructing the integrated optical processor chip to different target operating states, and further realizing dynamic signal processing [located at V].

[0027] Specifically, the feedback configuration module performs processing and analysis on the electrical signal based on the hybrid cooperation algorithm to generate a feedback control signal, combined with Figure 1 as shown, specifically including: Pretreatment process: The digital signal processor performs multi-dimensional signal analysis on the feedback electrical signal of the working state of the acquired mapped integrated optical processor chip, and then defines the performance characteristics of the integrated optical processor chip, and defines an optimization evaluation index for guiding the global search process according to the application requirements; Global search process: The global optimization algorithm constructs a multi-dimensional optimization vector corresponding one-to-one with multiple feedback control signals as optimization parameters. For example, if the integrated optical processor has ten regulators, then ten analog feedback control signals are required to achieve regulation, corresponding to the values of ten-dimensional optimization parameters. After randomly initializing the feedback control signal, the global optimization algorithm performs iterative optimization of the working state of the integrated optical processor chip with the maximization or minimization of the evaluation index as the goal. During the optimization process, the global optimization algorithm will continuously calculate the adjustment direction and adjustment speed of the optimization parameters according to the change of the evaluation index in real time, and then update the analog feedback control signal, so as to realize the rapid optimization of the working state of the integrated optical processor chip and calculate the optimal value of the control signal. After a preset number of iterations, calculate the optimization evaluation index and judge whether the evaluation index meets the convergence condition to the target value. If not, after enabling the multi-level optimization parameter search constraint mechanism, execute the global optimization algorithm again; if so, cache the optimal control signal and enter the multi-objective configuration process; Multi-objective Configuration Process: The digital signal processor redefines the multi-objective comprehensive evaluation index to more accurately map the working state of the integrated optical processor chip. The multi-objective optimization algorithm inherits the multi-dimensional optimization parameters of the global search process and uses the optimal control signal of the global search process as the initial parameter of this process. Considering that the execution of the multi-objective optimization algorithm involves multiple performance characteristics to be optimized simultaneously, the algorithm needs to adjust algorithm parameters such as the parameter scale, update rate, and mutation probability under the Pareto optimality criterion, coordinate and balance the mutual influences of multiple objectives, and at the same time ensure the computational efficiency and optimization robustness of the multi-objective optimization algorithm. Through continuous iterative optimization, the multi-objective comprehensive evaluation index approaches the target value, and the working state mapped to the integrated optical processor chip continuously approaches the target state. After the iteration ends, it is judged whether the optimization result of the multi-objective optimization algorithm meets the convergence condition. If not, the multi-level optimization parameter search constraint mechanism is also enabled, and then the multi-objective optimization algorithm is executed again. If so, it indicates that the current integrated optical processor chip is configured to the target working state, and different subsequent working modes are selected according to application requirements. If it is necessary to dynamically reconstruct the working state of the integrated optical processor chip, the comprehensive evaluation index is redefined and the multi-objective configuration process is repeated. If it is necessary for the integrated optical processor chip to work stably for a long time, the adaptive stabilization process is entered; Adaptive Stabilization Process: The local fine-tuning compensation algorithm inherits the definition methods of multiple performance characteristics in the multi-objective comprehensive evaluation index, and the optimization goal is to minimize the deterioration amount of the multi-objective comprehensive evaluation index. After the integrated optical processor chip is configured to the target working state, the local fine-tuning compensation algorithm performs real-time processing and analysis on the electrical signal, and monitors the deterioration amount of the multi-objective comprehensive evaluation index in real time. When the deterioration amount exceeds the preset threshold, fine-tuning compensation is started. The local fine-tuning compensation algorithm performs fine-tuning search within the neighborhood of the optimal control signal, and compensates for the performance deterioration of the integrated optical chip caused by thermal fluctuations and mechanical vibrations through the dynamic adjustment of the feedback control signal, thereby ensuring the continuous and stable operation of the integrated optical processor chip. When the preset stable working duration is reached, it is judged whether the work is over. If so, the multi-objective configuration process is returned to perform the configuration for other tasks. If not, the local fine-tuning compensation algorithm continues to be executed.

[0028] The multi-level optimization parameter search constraint mechanism is a mechanism that restricts the search range of optimization parameters in a discrete stepped manner. Based on this mechanism, the global search process and the multi-objective configuration process can effectively restrict the search step of the optimization parameter during optimization, thereby limiting the mutation amplitude of the feedback control signal, and further suppressing the performance perturbation caused by the temperature fluctuation triggered by the power jump.

[0029] The hybrid collaborative algorithm is an intelligent optimization algorithm with multi-stage progression, multi-algorithm collaboration, and multi-objective coordination. It integrates a global optimization algorithm, a multi-objective optimization algorithm, and a local fine-tuning compensation algorithm based on a phased collaboration mechanism. The global optimization algorithm, multi-objective optimization algorithm, and local fine-tuning compensation algorithm all adopt the following typical algorithms: genetic algorithm, simulated annealing algorithm, particle swarm algorithm, tabu search algorithm, gradient descent algorithm, or differential evolution algorithm.

[0030] Those skilled in the art can adjust the algorithm execution parameters of the digital signal processor, thereby configuring the integrated optical processor to different working states. The entire process of the hybrid collaborative algorithm is implemented on the digital signal processor.

[0031] Such as Figure 3 The overall structure diagram of the integrated optical processor chip and the dynamic configuration device of the specific embodiment of the present invention in the shown working scenario includes a continuous wave laser, an electro-optic modulator, an integrated optical processor, a splitter, a photodetector, an electrical amplifier, an electrical power splitter, an electrical filter, an analog-to-digital converter, a digital signal processor, and a digital-to-analog converter arranged in sequence. Among them, the continuous wave laser is used to generate a carrier optical signal and input it into the optical input end of the electro-optic modulator; the electro-optic modulator is used to modulate the electrical signal to be measured into the optical domain to generate an optical modulation signal loaded with the electrical signal; the integrated optical processor is used to implement the optical domain processing of the signal; the splitter is used to divide the optical signal into two paths; the photodetector is used to perform photoelectric conversion on the input signal to generate an electrical signal; the electrical amplifier is used to amplify the electrical signal; the electrical power splitter divides the electrical signal into two paths; the electrical filter is used to filter out high-frequency noise. The input end of the electrical filter is connected to the output end of the signal acquisition module, and the output end of the electrical filter is connected to the input end of the analog-to-digital converter. In this specific embodiment, the output end of the electrical filter is connected to the second output port of the electrical power splitter. The analog-to-digital converter is connected to the output end of the signal acquisition module and is used to convert the electrical signal output by the signal acquisition module into a digital signal; the digital signal processor extracts and analyzes the digital signal characteristics in real time, performs non-linear parameter optimization processing on the digital signal based on the hybrid collaborative algorithm, and outputs a feedback control signal. The output end of the digital-to-analog converter is connected to the regulator of the integrated optical processor chip and is used to perform digital-to-analog conversion on the feedback control signal and output an analog feedback control signal to the regulator. The integrated optical processor involved in the present invention has a regulator for self-adjusting the working state. The specific regulator is, for example, a regulating device such as a microelectrode deposited on the phase shifter in the optical processor. The microelectrode adjusts the waveguide refractive index of the optical waveguide in the integrated optical processor through a regulation method not limited to the thermo-optic effect, thereby realizing the configuration of the integrated optical processor.

[0032] The working and dynamic configuration process of the integrated optical processor chip in this specific embodiment is as follows: The continuous-wave laser generates a carrier optical signal and inputs it into the optical input end of the electro-optic modulator; The electro-optic modulator modulates the electrical signal into the optical domain to generate an optical modulation signal loaded with the electrical signal; The optical modulation signal is input into the integrated optical processor to realize the optical-domain processing of the signal; The optical splitter divides the optical signal into two paths. One path is input into the photodetector, and the other path is directly output as the optical signal processed by the integrated optical processor; The photodetector performs photoelectric conversion on the input signal to generate an electrical signal; The electrical signal is amplified by the electrical amplifier and then output to the electrical power divider; The electrical power divider divides the electrical signal into two paths. One path is input into the electrical filter, and the other path is directly output as the electrical signal processed by the integrated optical processor; The electrical signal is filtered by the electrical filter to remove high-frequency noise and then fed into the analog-to-digital converter to be converted into a digital signal to be analyzed; The digital signal processor processes the digital signal, including but not limited to fast Fourier transform, analyzes the current state of the integrated optical processor, and generates a feedback control signal based on the hybrid cooperation algorithm; The feedback control signal is converted into an analog feedback control signal through the digital-to-analog converter. The analog feedback control signal is applied to the regulator of the integrated optical processor chip to change the state of the regulator so as to realize the dynamic configuration of the integrated optical processor chip.

[0033] Finally, it should be noted that the above specific embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A dynamic configuration device for an integrated optical processor chip, characterized in that It includes a signal acquisition module and a feedback configuration module. The signal acquisition module is used to detect the output signal of the integrated optical processor chip and process the output signal into an electrical signal for transmission to the feedback configuration module. The feedback configuration module is used to perform processing and analysis on the electrical signal based on a hybrid cooperation algorithm to generate a feedback control signal, and the feedback control signal is applied to the regulator of the integrated optical processor chip to change the state of the regulator so as to achieve dynamic configuration of the integrated optical processor chip.

2. The dynamic configuration device of the integrated optical processor chip according to claim 1, characterized in that The feedback configuration module includes an analog-to-digital converter, a digital signal processor, and a digital-to-analog converter. The analog-to-digital converter, the digital signal processor, and the digital-to-analog converter are connected in sequence. The analog-to-digital converter is connected to the output end of the signal acquisition module and is used to convert the electrical signal output by the signal acquisition module into a digital signal. The digital signal processor is used to extract and analyze the digital signal features in real time, perform non-linear parameter optimization processing on the digital signal based on the hybrid cooperation algorithm, and output a feedback control signal. The output end of the digital-to-analog converter is connected to the regulator of the integrated optical processor chip and is used to perform digital-to-analog conversion on the feedback control signal and output an analog feedback control signal to the regulator.

3. The dynamic configuration device of the integrated optical processor chip according to claim 2, characterized in that, The feedback configuration module further includes an electrical filter, the input end of the electrical filter is connected to the output end of the signal acquisition module, and the output end of the electrical filter is connected to the input end of the analog-to-digital converter.

4. The dynamic configuration device of the integrated optical processor chip according to any one of claims 1 to 3, characterized in that, The signal acquisition module includes a photodetector, which is used to detect the optical signal output by the integrated optical processor chip and convert the optical signal into an electrical signal for transmission to the feedback configuration module.

5. The dynamic configuration device of the integrated optical processor chip according to claim 4, wherein, The output end of the photodetector is connected to an electrical amplifier, and the electrical amplifier is used to amplify the power of the input electrical signal and then transmit it to the feedback configuration module.

6. A dynamic configuration method for a dynamic configuration device of an integrated optical processor chip according to claim 1, characterized in that, It includes the following steps: The signal acquisition module detects the output signal of the integrated optical processor chip and processes the output signal into an electrical signal for transmission to the feedback configuration module. The feedback configuration module performs processing and analysis on the electrical signal based on the hybrid cooperation algorithm to generate a feedback control signal, and then applies the feedback control signal to the regulator of the integrated optical processor chip to change the state of the regulator so as to achieve dynamic configuration of the integrated optical processor chip.

7. The dynamic configuration method according to claim 6, characterized in that The feedback configuration module performs processing and analysis on the electrical signal based on the hybrid cooperation algorithm to generate a feedback control signal, specifically including: Pretreatment process: Perform multi-dimensional signal analysis on the feedback electrical signal, define the performance characteristics of the integrated optical processor chip, and further define the optimization evaluation index for guiding the global search process. Global search process: Randomly initialize the feedback control signal; use the global optimization algorithm to iteratively execute the optimization of the working state of the integrated optical processor chip, search for and calculate the optimal value of the control signal, calculate the optimization evaluation index after a preset number of rounds of iteration, and judge whether the evaluation index meets the convergence condition to the target value. If not, after enabling the multi-level optimization parameter search and constraint mechanism, execute the global optimization algorithm again; if so, cache the optimal control signal and enter the multi-objective configuration process. Multi-objective configuration process: Redefine the multi-objective comprehensive evaluation index for the working state of the integrated optical processor chip; Use the multi-objective optimization algorithm to perform the optimization calculation for the preset number of iterations, and then judge whether the optimization result of the multi-objective optimization algorithm meets the convergence condition. If not, after enabling the multi-level optimization parameter search and constraint mechanism, execute the multi-objective optimization algorithm again; If so, select different subsequent working modes according to the application requirements; If it is necessary to dynamically reconstruct the working state of the integrated optical processor chip, redefine the comprehensive evaluation index and repeat the multi-objective configuration process. If it is necessary for the integrated optical processor chip to work stably for a long time, enter the adaptive stabilization process; Adaptive stabilization process: After the integrated optical processor chip is configured to the target working state, use the local fine-tuning compensation algorithm to perform real-time compensation optimization, so as to suppress the change of the working state and the degradation of performance caused by thermal fluctuations and mechanical vibrations, and then ensure the continuous and stable operation of the integrated optical processor chip.

8. The dynamic configuration method according to claim 6, wherein The hybrid collaborative algorithm is an intelligent optimization algorithm with multi-stage progression, multi-algorithm collaboration, and multi-objective coordination, which integrates the global optimization algorithm, multi-objective optimization algorithm, and local fine-tuning compensation algorithm based on the phased collaboration mechanism.

9. The dynamic configuration method of the integrated optical processor chip according to claim 8, wherein The global optimization algorithm, multi-objective optimization algorithm, and local fine-tuning compensation algorithm all adopt the following typical algorithms: genetic algorithm, simulated annealing algorithm, particle swarm algorithm, tabu search algorithm, gradient descent algorithm, or differential evolution algorithm.

Images