Fiber Optic Immersion Sensing Method for Maximizing Safety Energy Efficiency

Abstract

The present invention discloses a method for maximizing the safety energy efficiency of optical fiber immersion sensing. This method senses the environmental state information of the optical cable through multiple sensors, uses optical sensing technology to transmit data, and analyzes and calculates the data in the data center. Among them, hierarchical meta-reinforcement learning is used to train the model to continuously optimize the system parameters, so as to achieve the maximization of safety energy efficiency.

Description

Technical Field

[0001] The present invention belongs to the field of power communication, and particularly relates to a fiber optic immersion sensing method for maximizing secure energy efficiency. Background Art

[0002] With the increasing demand for remote monitoring and data collection in the industrial, military, and environmental protection fields, fiber optic power transmission technology has been widely used due to its advantages such as high efficiency, low loss, and no electric sparks. Especially in harsh environments, it replaces traditional cable transmission systems to achieve safe and stable power supply. However, existing fiber optic power transmission systems still face challenges in terms of energy efficiency and security. First of all, the energy efficiency of fiber optic power transmission systems is affected by multiple factors such as fiber material, transmission distance, and system configuration during the transmission process, resulting in relatively large energy losses. In addition, the impact of environmental factors such as immersion on fiber optic transmission systems has not been fully resolved, and there are certain potential safety hazards in the actual application of the system. Although traditional energy management and optimization methods have improved the performance of fiber optic power transmission systems to a certain extent, they often rely on static models and fixed parameters and are difficult to cope with complex dynamic environmental changes.

[0003] Therefore, considering the complexity of the system environment and the diversity of requirements, traditional static optimization methods cannot well meet the requirements of real-time and dynamic adjustment. Summary of the Invention

[0004] The present invention aims to solve the technical challenges faced by existing fiber optic power transmission systems in terms of energy efficiency and security. Traditional fiber optic power transmission technology is easily affected by environmental factors during the energy transmission process, resulting in reduced transmission efficiency and ineffective guarantee of system security. Most existing technologies rely on static optimization models and cannot dynamically adapt to environmental changes and load requirements, which limits the performance of the system under complex and variable conditions. For this reason, the present invention proposes an intelligent optimization method based on fiber optic power transmission by combining a hierarchical meta-reinforcement learning algorithm, which can dynamically adjust system parameters under real-time environmental monitoring to achieve the maximization of energy efficiency and the guarantee of system security. The present invention uses the hierarchical meta-reinforcement learning algorithm to intelligently control the fiber optic power transmission system. For different transmission environments and real-time load changes, by continuously learning and optimizing strategies, it dynamically adjusts parameters such as the power output, transmission mode, and energy distribution of the system to maximize the secure energy efficiency of the system. The system state is sensed in real time through fiber optic sensing technology, and the sensed data is transmitted to the reinforcement learning model through a feedback mechanism, enabling the system to continuously optimize parameter settings in a dynamic environment, improve the overall energy efficiency, and ensure system security.

[0005] Based on this, the present invention proposes an optimization method for an optical fiber power transmission system to maximize safety energy efficiency. This method combines a hierarchical meta-reinforcement learning algorithm and can adjust various parameters of the optical fiber power transmission system in real time according to changes in the environment and load, ensuring the best energy utilization rate and safety of the system under different working conditions. At the same time, through a real-time monitoring and feedback mechanism, the system can intelligently respond to complex and changeable working environments, improving the application efficiency of optical fiber power transmission technology.

[0006] The embodiments of the present invention provide the following technical solutions:

[0007] S1, Construction and initialization of the optical fiber power transmission system. Select a laser with high power output and suitable for long-distance transmission as the light source, and calculate its output power to ensure energy transmission efficiency; at the same time, select single-mode optical fiber to support high-power transmission, evaluate the transmission loss of the optical fiber to ensure system stability; by adjusting the output power of the light source and evenly distributing it to each sensor, ensure the normal operation of each sensor; finally, debug the system to verify the performance of the light source, optical fiber and coupling system, and ensure that the overall energy transmission efficiency meets the design requirements.

[0008] S2, Implement the water immersion sensing function through optical sensing technology and feedback data. Use an optical fiber sensor to detect water immersion and monitor the change of signal intensity with the degree of water immersion; transmit the sensor data to the monitoring center through the optical fiber network, considering the signal loss during the transmission process; the system judges the water immersion event according to the signal change. If the change exceeds the preset threshold, trigger feedback; calibrate the system regularly and adjust the calibration parameters to ensure the accuracy of the sensing signal.

[0009] S3, Calculate the safety guarantee index and energy efficiency. Through a risk assessment model, comprehensively consider the power consumption, temperature and humidity of each node to calculate the safety index of the system; at the same time, calculate the energy efficiency of the system and evaluate the ratio of output power to input power; design an optimization goal to balance safety and energy efficiency, and achieve the optimal overall performance by adjusting system parameters.

[0010] S4, Use the hierarchical reinforcement learning algorithm to optimize system parameters. Design a hierarchical reinforcement learning framework, where the high-level policy formulates long-term plans and the low-level policy executes short-term operations; define the state space and action space, and design a reward function to balance energy efficiency and safety; through training the model, make it quickly adapt to new tasks, optimize system parameters, and improve the overall performance.

[0011] S5, Maximize the safety energy efficiency through a real-time energy monitoring and feedback mechanism. Use sensors to monitor the system status in real time and obtain data such as power and temperature; based on the hierarchical meta-reinforcement learning model, dynamically adjust the control strategy; continuously monitor and adjust until the safety energy efficiency reaches the optimal, and output the best control strategy.

[0012] The present invention discloses the following technical effects: The present invention solves the challenges of energy efficiency and safety of existing fiber optic energy transmission systems by combining a hierarchical meta-reinforcement learning algorithm and fiber optic energy transmission technology. First, the present invention adopts a hierarchical meta-reinforcement learning model to automatically optimize system parameters in a dynamic environment to maximize energy efficiency and ensure the safety of the system. Through real-time fiber optic sensing technology, the system status is fully sensed, and the sensed data is transmitted to the reinforcement learning algorithm for processing, thereby achieving continuous optimization of system performance. Unlike traditional static optimization methods, the present invention can dynamically adjust the energy distribution and transmission mode of the system according to real-time data, significantly improving the adaptability and efficiency of the fiber optic energy transmission system in different environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 A schematic flow chart of a method for sensing optical fiber water immersion with maximum safety and energy efficiency provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0015] As mentioned in the background technology section, how to perceive the system status in real time and optimize it while ensuring the energy efficiency and safety of the optical fiber energy transmission system is a key issue that needs to be solved in this field. Especially in dynamic environments and complex load conditions, the static parameter setting of traditional methods cannot meet the needs of flexible adjustment of the system.

[0016] The core idea of ​​this invention is to achieve deep integration of perception and calculation in the optimization process of fiber optic energy transmission system by combining fiber optic sensing technology with hierarchical meta-reinforcement learning algorithm. Fiber optic sensors are used to monitor the environment and system status in real time, and the collected data is input into the hierarchical meta-reinforcement learning model to dynamically adjust the energy transmission and allocation strategy of the system, thereby maximizing energy efficiency while ensuring the safe operation of the system.

[0017] See also Figure 1 The embodiment of the present invention provides a method for sensing optical fiber water immersion with maximum safety and energy efficiency, the method comprising:

[0018] S1. Construction and initialization of the optical fiber energy transmission system. Select a laser with high power output and suitable for long-distance transmission as the light source, and calculate its output power to ensure the energy transmission efficiency. At the same time, select single-mode optical fiber to support high-power transmission, evaluate the transmission loss of the optical fiber to ensure the system stability. By adjusting the output power of the light source and evenly distributing it to each sensor, ensure the normal operation of each sensor. Finally, debug the system to verify the performance of the light source, optical fiber and coupling system, and ensure that the overall energy transmission efficiency meets the design requirements.

[0019] S2. Realize the water immersion sensing function and feedback data through optical sensing technology. Use an optical fiber sensor to detect water immersion and monitor the change of signal intensity with the degree of water immersion. Transmit the sensor data to the monitoring center through the optical fiber network, considering the signal loss during the transmission process. The system judges the water immersion event according to the signal change. If the change exceeds the preset threshold, trigger the feedback. Calibrate the system regularly and adjust the calibration parameters to ensure the accuracy of the sensing signal.

[0020] S3. Calculate the security guarantee index and energy efficiency. Through the risk assessment model, comprehensively consider the power consumption, temperature and humidity of each node to calculate the security index of the system. At the same time, calculate the energy efficiency of the system and evaluate the ratio of the output power to the input power. Design the optimization goal to balance security and energy efficiency, and achieve the optimal overall performance by adjusting the system parameters.

[0021] S4. Use the hierarchical reinforcement learning algorithm to optimize the system parameters. Design a hierarchical reinforcement learning framework. The high-level policy formulates long-term plans, and the low-level policy executes short-term operations. Define the state space and action space, and design a reward function to balance energy efficiency and security. Through training the model, make it quickly adapt to new tasks, optimize the system parameters, and improve the overall performance.

[0022] S5. Maximize the safe energy efficiency through a real-time energy monitoring and feedback mechanism. Use sensors to monitor the system state in real time and obtain data such as power and temperature. Based on the hierarchical meta-reinforcement learning model, dynamically adjust the control strategy. Continuously monitor and adjust until the safe energy efficiency reaches the optimal, and output the best control strategy.

[0023] Among them, in step S1, select a laser with high power output, small beam divergence angle and suitable for long-distance transmission as the light source. The output power of the light source is:

[0024] (1)

[0025] Among them, is the output power at the output end of the optical fiber, is the efficiency of the light source, which depends on the type and quality of the light source, is the input electric power at the input end of the optical fiber.

[0026] Secondly, a single-mode optical fiber suitable for long-distance and high-power transmission is selected. The transmission loss coefficient of the optical fiber is:

[0027] (2)

[0028] Wherein, is the length of the optical fiber, is the transmission loss coefficient of the optical fiber.

[0029] Then, by adjusting the output power of the light source and the light energy allocated to each terminal, ensure that each sensor operates normally within its working range. Let the total power of the system be , and the power allocated to each terminal sensor be . Then, the power demand of each node can be ensured to be met by adjusting the allocation ratio. The power allocation formula is as follows:

[0030] (3)

[0031] Wherein, is the number of terminal sensors. Through reasonable power allocation, ensure the normal operation of each sensor and avoid system failures caused by insufficient power.

[0032] Finally, conduct initial-stage debugging on the system, including verifying the light source, optical fiber, and coupling system, ensuring that the performance of each module meets the design requirements, and testing the overall efficiency of the optical fiber energy transmission system. The overall energy transmission efficiency of the system can be evaluated by the following formula:

[0033] (4)

[0034] Wherein, is the coupling efficiency between the light source and the optical fiber. By conducting initial debugging on the system, the energy losses in each link can be verified, and the parameters can be adjusted to optimize the system efficiency.

[0035] Among them, in step S2, the intensity change of the optical signal obtained by the optical fiber sensor has a linear relationship with the degree of water immersion. The sensor signal intensity is expressed by the following formula:

[0036] (5)

[0037] Wherein, is the initial signal intensity of the sensor. When water immersion occurs, the presence of water will affect the transmission characteristics of the optical fiber, resulting in a change in the signal intensity.

[0038] Secondly, the data collected by the fiber optic sensor will be transmitted to the monitoring center through the fiber optic network. The transmission loss coefficient of the fiber optic is , and the transmission distance of the fiber optic network is , then the transmission loss of the optical signal can be expressed as:

[0039] (6)

[0040] Among them, is the input electric power at the input end of the fiber optic. The data of the fiber optic sensor is collected by the fiber optic network to the central processing system and analyzed in real time by the processing system.

[0041] Then, the integrated water immersion sensing module needs to judge the occurrence of the water immersion event in real time through the algorithm and give timely feedback. Set the threshold of the signal change to , when the sensor signal change exceeds this threshold, it means that the water immersion event occurs. At this time, the judgment formula is:

[0042] (7)

[0043] Among them, represents the sensor signal change, represents the current sensor signal, represents the sensor signal in the previous time slot. If , then the water immersion sensing event is triggered, and the system responds and notifies the management center.

[0044] Finally, in order to ensure the accuracy and stability of the water immersion sensing module, the system needs to be calibrated and optimized regularly. During the system calibration process, model fitting needs to be carried out according to the experimental data. Let the calibration factor be , then the calibrated sensing signal is:

[0045] (8)

[0046] The calibration factor is dynamically adjusted according to the experimental results and sensor errors to ensure the accurate relationship between the sensing signal and the actual water immersion degree.

[0047] Among them, in step S3, in order to ensure system security, a risk assessment model is introduced to evaluate the security by monitoring the states of each node. During this process, the risk assessment formula of the system is as follows:

[0048] (9)

[0049] Among them, represents the overall security risk of the system, is the Security risk assessment of nodes represents the total number of nodes. The risk assessment of each node is jointly determined by its node power consumption, temperature, and humidity factors. Specifically, the security risk assessment of a node is calculated by the following formula:

[0050] (10)

[0051] where is the real-time power consumption of the th node, is the maximum power carrying capacity of the node, is the real-time temperature of the th node, is the maximum safe temperature of the node, is the real-time humidity of the th node, is the maximum humidity threshold, , and are importance coefficients used to adjust the importance of node power consumption, temperature, and humidity factors, reflecting the relative importance of power consumption, temperature, and humidity for security assurance.

[0052] Secondly, in order to ensure that the system can achieve optimal energy utilization during operation, it is necessary to calculate the overall energy efficiency of the system. The energy efficiency of the system can be represented by the ratio of the transmission power to the input power. The input power of the system is , and the output power is , then the energy efficiency is calculated by the following formula:

[0053] (11)

[0054] Optimization of energy efficiency requires considering the transmission loss of the optical fiber, the power output of the light source, and the energy consumption of the sensing nodes simultaneously. To further optimize the system efficiency, it can be calculated in the following way:

[0055] (12)

[0056] where is the output power of the system, is the transmission loss coefficient of the optical fiber, is the length of the optical fiber. By calculating the power of the effective light source output after attenuation, the energy transmission efficiency of the system can be evaluated.

[0057] Finally, by combining the assessments of security risks and energy efficiency, an optimization objective function is designed to simultaneously maximize both:

[0058] (13)

[0059] where 、 、 and are the importance coefficients for adjusting the node power consumption, temperature, humidity, and total power input / output, used to balance the relationship between security and energy efficiency. represents the secure energy efficiency. By using the hierarchical meta-reinforcement learning method to solve this problem, the light source output, fiber length, node power, etc. are automatically adjusted to maximize the secure energy efficiency.

[0060] Among them, in step S4, in hierarchical reinforcement learning, two levels of policies are designed. The low-level policy executes short-term specific actions, adjusts the sensor sensitivity and optimizes the power transmission efficiency , and the high-level policy is responsible for long-term decisions, determining the sensing frequency and selecting the energy management strategy.

[0061] First, to apply the hierarchical meta-reinforcement learning algorithm, the state space, action space, and reward function need to be defined.

[0062] State space includes all possible environmental and system states in the system. In the fiber optic immersion sensing system, the state space is represented as a vector, including the power transmission state , temperature state , humidity state , immersion state , fiber optic system power transmission efficiency , system risk degree . The state space is represented as follows:

[0063] (14)

[0064] where represents the state of the th device and its environmental variables.

[0065] Secondly, the action space contains all possible control actions. Each control action changes the system parameters to optimize energy efficiency and security.

[0066] The action space of the high-level policy is defined as:

[0067] (15)

[0068] The action space of the low-level policy is:

[0069] (16)

[0070] Among them, and are the action spaces of the high-level and low-level policies respectively, and are the actions of the high-level and low-level policies respectively.

[0071] The high-level policy focuses on long-term statistical features, and the statistical feature set is defined as:

[0072] (17)

[0073] Among them represents the average power transmission state, represents the total sampling time, represents the starting sampling time, represents the variance of temperature fluctuation, The function of temperature changing with time t varies, represents the average temperature, represents the maximum humidity state, represents the expected frequency of flooding events, represents the average power transmission efficiency of the optical fiber system, represents the average system risk level.

[0074] The low-level policy focuses on the instantaneous environmental state, and the statistical feature set is defined as:

[0075] (18)

[0076] Among them is the power transmission state, is the temperature state, is the humidity state, is the flooding state, The gradient of the change in the power transmission efficiency of the optical fiber system, is the rate of change of the system risk level.

[0077] Then, the reward function is a function that measures the quality of the current state of the system and the state after performing an action. Among them represents the state space, represents the action taken by the agent, . The reward function should consider both the safety and energy efficiency of the system. After the system executes each action, a comprehensive reward is calculated to evaluate the effectiveness of the action. The calculation formula is as follows:

[0078] (19)

[0079] where, is the change in energy efficiency after executing the action, is the change in system risk, and are the weight coefficients for adjusting energy efficiency and safety risk.

[0080] Finally, the goal of the model is to learn how to quickly adapt to new tasks from previous experiences. This enables the model to quickly adjust its strategy when facing different flooding perception tasks and improve the training efficiency. The training process is represented by the following formula:

[0081] (20)

[0082] where, represents the parameters of the system, is the task set, is the loss function after learning on the task set, represents the new system parameters obtained through training.

[0083] Among them, in step S5, the system uses the integrated sensor network to monitor the power transmission status of each system component in real time , temperature status , humidity status , flooding status , power transmission efficiency of the fiber optic system , system risk level , where represents the next state.

[0084] First, the status information of the fiber optic system and the environment obtained through real-time monitoring by the sensors is expressed as:

[0085] (21)

[0086] Secondly, the hierarchical meta-reinforcement learning model obtained through training in formula (20) is used to update its decision-making strategy. The state update of the system can be expressed as:

[0087] (22)

[0088] At this time, the control strategy is updated to:

[0089] (23)

[0090] Among them, represents the parameter change obtained by hierarchical meta-reinforcement learning after being adjusted based on the current feedback.

[0091] Finally, repeat formulas (21) to (23), transfer to the next iteration, so as to output when obtains the maximum value as the optimal control strategy.

[0092] The present invention combines a hierarchical meta-reinforcement learning algorithm and a fiber-optic power transmission technology to solve the challenges in energy efficiency and security of existing fiber-optic power transmission systems. First, the present invention adopts a hierarchical meta-reinforcement learning model to automatically optimize system parameters in a dynamic environment to maximize energy efficiency and ensure the security of the system. Through real-time fiber-optic sensing technology, the system state is comprehensively sensed, and the sensed data is transmitted to the reinforcement learning algorithm for processing, thereby realizing continuous optimization of the system performance. Different from traditional static optimization methods, the present invention can dynamically adjust the energy distribution and transmission mode of the system according to real-time data, significantly improving the adaptability and efficiency of the fiber-optic power transmission system in different environments.

[0093] In this article, specific examples are used to elaborate on the principle and implementation manner of the present invention. The description of the above embodiments is only used to help understand the method and its core idea of the present invention; at the same time, for those of ordinary skill in the art, according to the idea of the present invention, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present invention.

Claims

1. A method for sensing optical fiber water immersion with maximum safety and energy efficiency, characterized in that: The steps include: S1, construction and initialization of the optical fiber energy transmission system. Select a high-power output laser suitable for long-distance transmission as the light source, and calculate its output power to ensure energy transmission efficiency. At the same time, select a single-mode optical fiber to support high-power transmission, evaluate the transmission loss of the optical fiber, and ensure system stability. Adjust the output power of the light source and distribute it evenly to each sensor to ensure the normal operation of each sensor. Finally, debug the system, verify the performance of the light source, optical fiber, and coupling system, and ensure that the overall energy transmission efficiency meets the design requirements. S2, realizes water immersion sensing function and feedback data through optical sensing technology, uses optical fiber sensors to detect water immersion, and monitors the change of signal strength with the degree of water immersion; transmits sensor data to the monitoring center through the optical fiber network, taking into account the signal loss during the transmission process; the system determines the water immersion event based on the signal change, and triggers feedback if the change exceeds the preset threshold; regularly calibrates the system and adjusts the calibration parameters to ensure the accuracy of the sensing signal; S3, calculates the security assurance index and energy efficiency. Through the risk assessment model, the power consumption, temperature and humidity of each node are comprehensively considered to calculate the security index of the system. At the same time, the energy efficiency of the system is calculated to evaluate the ratio of output power to input power. Design optimization goals, balance safety and energy efficiency, and achieve optimal overall performance by adjusting system parameters; S4, using hierarchical reinforcement learning algorithm to optimize system parameters, designing a hierarchical reinforcement learning framework, where high-level strategies formulate long-term plans and low-level strategies perform short-term operations; Define the state space and action space, and design the reward function to balance energy efficiency and safety; train the model to quickly adapt to new tasks, optimize system parameters, and improve overall performance; S5, through real-time energy monitoring and feedback mechanism, maximizes safety energy efficiency, uses sensors to monitor system status in real time, obtains power and temperature data; dynamically adjusts control strategy based on hierarchical meta-reinforcement learning model; continuously monitors and adjusts until safety energy efficiency is optimal, and outputs the best control strategy; Among them, in step S4, in hierarchical reinforcement learning, two levels of strategies are designed. The low-level strategy performs short-term specific actions and adjusts the sensor sensitivity. and optimize power transfer efficiency ,The high-level strategy is responsible for long-term decision making, determining the sensing frequency and selecting the energy management strategy; First, in order to apply the hierarchical meta-reinforcement learning algorithm, it is necessary to define the state space, action space, and reward function; State Space Including all possible environments and system states in the system. In the fiber optic water immersion sensing system, the state space is represented as a vector, including the power transmission state , temperature status , humidity status , immersion state , optical fiber system power transmission efficiency , system risk , the state space is represented as follows: （14） in Indicates The state of each device and its environment variables; Second, the action space Contains all possible control actions, each of which changes a parameter of the system in order to optimize energy efficiency and safety; The action space of the high-level policy is defined as: （15） The action space of the low-level strategy is: （16） in, and are the action spaces of high-level and low-level strategies, respectively. and They are the actions of high-level and low-level strategies respectively; High-level strategies focus on long-term statistical characteristics, statistical characteristics collection Defined as: （17） in represents the average power transmission state, represents the total sampling time, Indicates the sampling start time, represents the temperature fluctuation variance, Temperature over time t The function of change, represents the mean temperature, Indicates the maximum humidity state. represents the expected frequency of flooding events, represents the mean power transmission efficiency of the optical fiber system, represents the mean value of system risk; Low-level strategies focus on instantaneous environmental states and statistical feature collections Defined as: （18） in is the power transmission state, is the temperature state, For humidity status, In water-immersed state, The gradient of power transmission efficiency change of optical fiber system, is the rate of change of system risk; Then, the reward function is a function that measures the quality of the system's current state and the state after the action is executed, where represents the state space, represents the action taken by the agent, , the reward function should consider both the safety and energy efficiency of the system; after executing each action, the system calculates a comprehensive reward to evaluate the effectiveness of the action. The calculation formula is as follows: （19） in, is the change in energy efficiency after performing the action, is the change in systemic risk, and is the weight coefficient for adjusting energy efficiency and safety risk; Finally, the goal of the model is to learn how to quickly adapt to new tasks from previous experience, so that the model can quickly adjust its strategy when facing different immersion perception tasks and improve training efficiency. The training process is expressed by the following formula: （20） in, represents the parameters of the system, is a set of tasks, is the loss function after learning on the task set, Represents the new system parameters obtained through training.

2. The optical fiber water immersion sensing method with maximum safety energy efficiency according to claim 1, characterized in that: In step S1, the output power of the light source is: （1） in, is the output power at the output end of the optical fiber, is the efficiency of the light source, which depends on the type and quality of the light source. is the input electrical power at the fiber input end; Secondly, choose to use single-mode optical fiber suitable for long-distance, high-power transmission. The transmission loss coefficient of optical fiber is: （2） in, is the length of the optical fiber, is the transmission loss coefficient of the optical fiber; Then, by adjusting the output power of the light source and the light energy allocated to each terminal, each sensor can be ensured to work normally within its working range. The total system power is , the power allocated to each terminal sensor is , the power demand of each node can be met by adjusting the allocation ratio. The power allocation formula is as follows: (3) in, For the number of terminal sensors, reasonable power allocation is used to ensure the normal operation of each sensor and avoid system failures caused by insufficient power; Finally, the system is debugged in the initial stage, including verification of the light source, optical fiber and coupling system to ensure that the performance of each module meets the design requirements, and the overall efficiency of the optical fiber energy transmission system is tested. It can be evaluated by the following formula: （4） in, It is the coupling efficiency between the light source and the optical fiber. Through the initial debugging of the system, the energy loss of each link can be verified and the parameters can be adjusted to optimize the system efficiency.

3. The optical fiber water immersion sensing method with maximum safety energy efficiency according to claim 1, characterized in that: In step S2, the intensity change of the optical signal obtained by the optical fiber sensor is linearly related to the water immersion degree. It is expressed by the following formula: （5） in, is the initial signal strength of the sensor. When water immersion occurs, the presence of water will affect the transmission characteristics of the optical fiber, resulting in changes in signal strength; Secondly, the data collected by the optical fiber sensor will be transmitted to the monitoring center through the optical fiber network. The transmission loss of the optical fiber is , the transmission distance of the optical fiber network is , then the transmission loss of the optical signal It can be expressed as: （6） in, It is the power of transmitting input signal. The fiber optic sensor data is collected to the central processing system through the fiber optic network and analyzed in real time by the processing system; Then, the integrated flood sensing module needs to use the algorithm to determine the occurrence of flood events in real time and provide timely feedback, setting the threshold of signal change to , when the sensor signal changes beyond this threshold, it indicates that a flooding event has occurred. At this time, the judgment formula is: （7） in, Indicates sensor signal changes, Indicates the current sensor signal, Indicates the sensor signal of the previous time slot. If , a flood sensing event is triggered, and the system responds and notifies the management center; Finally, in order to ensure the accuracy and stability of the water immersion sensing module, the system needs to be calibrated and optimized regularly. During the system calibration process, the model needs to be fitted according to the experimental data. The calibration factor is set as , then the calibrated perception signal for: （8） Calibration Factor Dynamic adjustments are made based on experimental results and sensor errors to ensure an accurate relationship between the sensed signal and the actual water immersion level.

4. The optical fiber water immersion sensing method with maximum safety energy efficiency according to claim 1, characterized in that: In step S3, in order to ensure the security of the system, a risk assessment model is introduced to evaluate the security of each node by monitoring its status. In this process, the risk assessment formula of the system is as follows: （9） in, Represents the overall security risk of the system, It is Security risk assessment of each node. Represents the total number of nodes. The risk assessment of each node is determined by its node power consumption, temperature, and humidity factors. Specifically, the safety risk assessment of the node Calculated by the following formula: （10） in, It is The real-time power consumption of each node, is the maximum power carrying capacity of the node, It is The real-time temperature of each node, is the maximum safe temperature of the node, It is Real-time humidity of each node, is the maximum humidity threshold, , and It is the importance coefficient used to adjust the node power consumption, temperature, and humidity factors, reflecting the relative importance of power consumption, temperature, and humidity to security assurance; Secondly, in order to ensure that the system can achieve optimal energy utilization during operation, the overall energy efficiency of the system must be calculated. The energy efficiency of the system can be expressed as the ratio of transmission power to input power. The input power of the system is , the output power is , then the energy efficiency Calculated by the following formula: （11） The optimization of energy efficiency needs to consider the transmission loss of the optical fiber, the power output of the light source, and the energy consumption of the sensing node. In order to further optimize the system efficiency, the calculation can be performed in the following way: （12） in, is the power output by the light source, is the transmission loss coefficient of the optical fiber, is the length of the optical fiber, and the effective light source output power after attenuation is calculated , the energy transfer efficiency of the system can be evaluated; Finally, by combining the evaluation of safety risk and energy efficiency, an optimization objective function is designed to maximize both: （13） in, , , and It is the importance coefficient used to adjust node power consumption, temperature, humidity, and total power input and output, and is used to balance the relationship between security and energy efficiency. Indicates safety energy efficiency.

5. The optical fiber water immersion sensing method with maximum safety energy efficiency according to claim 1, characterized in that: In step S5, the system again monitors the power transmission status of each system component in real time through the integrated sensor network , temperature status , humidity status , immersion state , optical fiber system power transmission efficiency , system risk ,in Represents the next state; First, the status information of the optical fiber system and environment is obtained through real-time monitoring by sensors. The status information is expressed as: （21） Secondly, the hierarchical meta-reinforcement learning model trained in formula (20) is used to update its decision strategy. The state update of the system can be expressed as: （22） At this time, the control strategy Updated to: （23） in, represents the parameter changes obtained by hierarchical meta-reinforcement learning based on the current feedback adjustment, Finally, repeat formula (21) to formula (23) and enter the next iteration to output When the maximum value is obtained as the optimal control strategy.

Images