Air-sea cross-domain communication gateway energy optimization management system and method of coupling environment

Abstract

The invention provides an air-sea cross-domain communication gateway energy optimization management system and method in a coupling environment, a closed-loop control mechanism is constructed by fusing data noise reduction of extended Kalman filtering and particle filtering and an LSTM power prediction and reinforcement learning scheduling algorithm, and fine regulation and control of whole sea conditions are realized by combining a multi-mode power supply network. Compared with a traditional static strategy, the energy use time is effectively prolonged; the environmental state judgment accuracy is improved through a nonlinear motion model and a trend prediction mechanism, a transition scene regulation blank is effectively filled up through a medium sea condition mode, the mode switching time delay is lower, and the adaptive capacity of the system to a complex and changeable marine environment is remarkably enhanced; through a surge suppression circuit, redundant power supply design and a hybrid communication link scheduling strategy, the success rate of coping with power supply fluctuation and hardware faults, the communication link switching accuracy and the cross-domain data transmission success rate of the system are all improved, and meanwhile the system has wide applicability and can be expanded to various mobile platforms such as unmanned aerial vehicles and unmanned ships.

Description

Technical Field

[0001] This invention relates to the field of air-sea cross-domain communication technology, and in particular to an energy optimization management system and method for air-sea cross-domain communication gateways in coupled environments. Background Technology

[0002] With the continuous growth in demand for marine resource development and monitoring, the application of air-sea cross-domain communication equipment (such as communication buoys, underwater robots, and unmanned surface vessels) in complex marine environments is becoming increasingly widespread. These devices are typically deployed in open sea areas far from shore, undertaking important tasks such as data transmission, environmental monitoring, and navigation relay. Their stable operation highly depends on an efficient energy management system. In existing technologies, energy management for air-sea cross-domain communication equipment largely relies on static strategies, such as sleep-wake mechanisms based on fixed time periods, power consumption control based on single thresholds, or simple battery power level management. These methods are relatively simple in design and easy to implement, and can, to some extent, meet the basic energy-saving requirements under calm sea conditions.

[0003] However, the marine environment is highly dynamic and unpredictable, and conventional static strategies are insufficient to effectively address the surge in energy consumption caused by extreme conditions (such as strong winds, waves, rapid currents, and severe equipment shaking). Specifically, existing technologies have the following shortcomings: First, they lack environmental awareness and adaptive adjustment capabilities. When equipment encounters strong winds and waves, the communication module needs to frequently increase its transmission power to maintain link stability, leading to a sharp increase in energy consumption. Traditional power management modules cannot dynamically adjust their power supply strategies based on real-time motion status, which can easily result in energy waste or system downtime. Second, existing inertial measurement units are mostly used for equipment attitude monitoring or navigation and positioning, and are not deeply coupled with energy management modules. Their motion data is not fully utilized for power consumption prediction and scheduling optimization. Third, although some studies have attempted to optimize equipment functions through motion data, their control algorithms are mostly unidirectional open-loop, lacking robust handling of nonlinear and non-Gaussian noise, as well as the ability to autonomously learn and predict extreme environments, resulting in low energy utilization efficiency and poor system reliability under complex operating conditions.

[0004] Therefore, there is an urgent need in this field for an intelligent management system that can deeply integrate environmental perception, multi-algorithm fusion and dynamic energy scheduling to solve the key technical challenges of energy efficiency, environmental adaptability and operational reliability of cross-domain communication equipment in complex marine environments. Summary of the Invention

[0005] The purpose of this invention is to provide an energy optimization management system and method for air-sea cross-domain communication gateways in coupled environments, so as to solve the problems existing in the prior art.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] This invention provides an energy optimization management system for air-sea cross-domain communication gateways in coupled environments, comprising a main control module and an energy management module, wherein the main control module includes:

[0008] An inertial measurement unit, which is used to acquire environmental sensing data in real time;

[0009] An embedded processor module, which employs a floating-point arithmetic unit, combines extended Kalman filtering and particle filtering fusion algorithms for data parsing and noise reduction fusion, and incorporates a reinforcement learning scheduling algorithm to coordinate task priorities;

[0010] An environmental state determination unit is used to construct a nonlinear motion model based on the environmental sensing data and analyze the environmental state through threshold comparison and trend prediction.

[0011] A communication link scheduler, which is used to dynamically switch the communication mode of the module using an optimal communication link selection algorithm;

[0012] A recording and fault diagnosis module is used to store system operation logs, self-test results, and abnormal events;

[0013] The energy management module includes:

[0014] A dynamic power conversion unit is used to provide a stable voltage and is equipped with an LSTM power prediction model to monitor the remaining battery capacity and energy consumption trend of the battery pack.

[0015] A multi-mode power supply network is used to switch the battery pack power supply mode according to the environmental conditions. The operating modes include daily mode, medium sea state mode and extreme event mode.

[0016] An emergency energy storage unit is used to provide buffer power supply in the event of a power failure.

[0017] Preferably, the inertial measurement unit is located at the center of the ocean buoy and includes an accelerometer, a gyroscope, a transducer, an antenna module, and a temperature sensor. Acceleration data and angular velocity data are collected by the accelerometer and the gyroscope, and ambient temperature data is collected by the temperature sensor. Together, they serve as the environmental sensing data, and the environmental sensing data is transmitted to the embedded processor module through the antenna module.

[0018] Preferably, the embedded processor module initially eliminates Gaussian noise in the environmental sensing data through extended Kalman filtering, and then processes nonlinear and non-Gaussian noise through particle filtering to complete multi-dimensional data fusion; the environmental state determination unit constructs a nonlinear motion model based on the fused data, performs motion state calculation and trend prediction, obtains vibration intensity data, tilt angle change rate data and change trend parameters, and determines the environmental state by comparing thresholds and trend analysis.

[0019] Preferably, when the vibration intensity data and the rate of change of tilt angle data do not exceed the first threshold and the trend of change is stable, the daily mode is executed; when the data exceeds the first threshold but does not exceed the second threshold, or the trend of change shows fluctuating growth, the moderate sea state mode is executed; when the data exceeds the second threshold, the extreme event mode is executed.

[0020] The daily mode includes: using a single lithium battery pack for power supply, shutting down non-core circuits through optocoupler isolation relays, using Huffman coding to compress data and transmit data back, using an LSTM power prediction model to predict low-energy periods, and appropriately reducing the sampling rate of the inertial measurement unit.

[0021] The medium sea state mode includes: using a main lithium battery pack and an auxiliary energy storage unit for power supply, retaining redundancy in the core communication link, using improved Huffman coding to compress data, and dynamically balancing the sampling rate and communication frequency through reinforcement learning algorithms;

[0022] The extreme event modes include: activating the TVS surge suppression circuit and redundant lithium battery pack, cutting off non-core functions, and prioritizing power supply to the data acquisition and BeiDou communication modules of the inertial measurement unit.

[0023] Preferably, the module communication methods include 4G transmission, Beidou communication, data transmission radio, and underwater acoustic communication. The communication link scheduler collects key indicators of the link in each communication method every 1 second, normalizes the key indicators and assigns weights. Communication methods with scores higher than the threshold enter the candidate pool, and the highest score in the candidate pool is taken as the main communication method, and the second highest score is taken as the alternative communication method.

[0024] During communication, a hybrid model of Bayesian network and reinforcement learning is used to update weight allocation in real time based on historical data and real-time environmental parameters. When the score of the main communication method is lower than the threshold for three consecutive times, the system switches to the alternative communication method and records the fault log. The reinforcement learning algorithm dynamically adjusts the transmission power of each communication method based on energy consumption and communication quality feedback.

[0025] Preferably, the dynamic power conversion unit includes a DC / DC conversion chip, a coulomb counter, and an LSTM power prediction module. The DC / DC conversion chip outputs an adjustable voltage based on a specified input voltage. The coulomb counter monitors the remaining battery power in real time. The LSTM power prediction module predicts the energy consumption trend for the next hour based on historical energy consumption data and environmental parameters. The embedded processor module combines the remaining battery power and the energy consumption trend, and uses a reinforcement learning algorithm to adjust the sampling rate of the inertial measurement unit and the transmission power of the communication link.

[0026] This invention also provides a method for optimizing the energy management of an air-sea cross-domain communication gateway in a coupled environment, comprising the following steps:

[0027] S1. An inertial measurement unit is installed in an ocean buoy to collect acceleration data, angular velocity data, and ambient temperature data in real time;

[0028] S2. The data is analyzed and denoised using a fusion algorithm of extended Kalman filter and particle filter. A nonlinear motion model is constructed based on the fused data, and the environmental state is judged by threshold comparison and trend prediction.

[0029] S3. Switch the power supply mode of the system modules in real time based on the environmental condition determination results;

[0030] S4. Employ the optimal communication link selection algorithm to dynamically switch module communication modes and adjust transmission power;

[0031] S5. The remaining battery charge and energy consumption trend are monitored by combining a coulomb counter and an LSTM power prediction model, and the sampling rate of the inertial measurement unit is adjusted by combining a reinforcement learning algorithm.

[0032] Preferably, in step S2, the process of constructing a nonlinear motion model based on fused data and determining the environmental state through threshold comparison and trend prediction includes:

[0033] A nonlinear motion model is constructed based on fused data to calculate vibration intensity data, tilt rate of change data, and trend parameters. When the data does not exceed the first threshold and the trend is stable, the normal mode is executed. When the data exceeds the first threshold but does not exceed the second threshold, or when the trend fluctuates and increases, the moderate sea state mode is executed. When the data exceeds the second threshold, the extreme event mode is executed.

[0034] Preferably, the daily mode includes: using a single lithium battery pack for power supply, shutting down non-core circuits through optocoupler isolation relays, compressing and transmitting data using Huffman coding, using an LSTM power prediction model to predict low-energy consumption periods, and adjusting the sampling rate of the inertial measurement unit to 10-20Hz.

[0035] The medium sea state mode includes: using a main lithium battery pack and an auxiliary energy storage unit for power supply, retaining redundancy in the core communication link, using improved Huffman coding to compress data, and using a reinforcement learning algorithm to dynamically adjust the sampling rate to 20-50Hz based on energy consumption feedback, while simultaneously optimizing the communication transmission power.

[0036] The extreme event mode includes: activating the TVS surge suppression circuit and redundant lithium battery pack, cutting off non-core functions, fixing the sampling rate to 50Hz, and prioritizing power supply to the data acquisition of the inertial measurement unit and the Beidou communication module.

[0037] Preferably, in step S4, the process of dynamically switching the module communication mode and adjusting the transmission power using the optimal communication link selection algorithm includes:

[0038] Every 1 second, key indicators such as signal strength, bit error rate, latency, and energy consumption level of each communication method are collected, normalized, and weighted. Communication methods with scores higher than the threshold are entered into the candidate pool, and the communication method with the highest score in the candidate pool is selected as the primary communication method, and the second highest score is selected as the alternative communication method.

[0039] Based on historical data and real-time environmental parameters, the weight allocation is updated through a hybrid model of Bayesian network and reinforcement learning. The reinforcement learning algorithm dynamically adjusts the transmission power of each communication method according to feedback on communication quality and energy consumption.

[0040] If the score of the primary communication method falls below the threshold for three consecutive times, the system will switch to the alternative communication method and record a fault log.

[0041] The present invention achieves the following beneficial technical effects compared to the prior art:

[0042] This invention provides an energy optimization management system and method for air-sea cross-domain communication gateways in coupled environments. By integrating extended Kalman filtering and particle filtering for data noise reduction, LSTM power prediction, and reinforcement learning scheduling algorithms, a closed-loop control mechanism of "perception-prediction-scheduling" is constructed. Combined with a multi-mode power supply network, it achieves refined regulation across all sea states. Compared with traditional static strategies, energy usage time is extended by more than 30%. Through nonlinear motion models and trend prediction mechanisms, the accuracy of environmental state judgment is improved to over 98%. The medium sea state mode effectively fills the regulatory gap in transitional scenarios, with mode switching latency ≤300ms, significantly enhancing the system's adaptability to complex and variable marine environments. Through surge suppression circuits, redundant power supply design, and hybrid communication link scheduling strategies, the system achieves a success rate of ≥99.5% in dealing with power fluctuations and hardware failures, a communication link switching accuracy of ≥97%, and a cross-domain data transmission success rate of 99.7%. It also has broad applicability and can be extended to various mobile platforms such as UAVs and unmanned vessels. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram illustrating the functional principle of the main control module in an embodiment of the present invention;

[0045] Figure 2 This is a schematic diagram illustrating the functional principle of the energy management module in an embodiment of the present invention;

[0046] Figure 3 This is a schematic diagram of the method flow in an embodiment of the present invention;

[0047] Figure 4 This is a schematic diagram of the feedback mechanism flow in an embodiment of the present invention;

[0048] Figure 5 This is a schematic diagram of the inertial measurement unit structure in an embodiment of the present invention. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] The purpose of this invention is to provide an energy optimization management system and method for air-sea cross-domain communication gateways in coupled environments, so as to solve the problems existing in the prior art.

[0051] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0052] Example 1:

[0053] This embodiment proposes an energy optimization management system for a coupled air-sea cross-domain communication gateway, such as... Figure 1 and Figure 2 As shown, the system mainly consists of two parts: a main control module and an energy management module. By sensing the movement of the marine environment in real time and combining multiple algorithm fusions with a dynamic scheduling mechanism, the system achieves intelligent energy management and stable cross-domain communication.

[0054] The main control module, as the core decision-making unit of the system, is responsible for data integration, environmental status assessment, communication scheduling, and energy coordination management. Specifically, the inertial measurement unit (IMU) is located at the center of the ocean buoy and includes an accelerometer, gyroscope, transducer, antenna module, and temperature sensor. The accelerometer has a range of ±16g and an accuracy of ±0.01g, used to collect three-dimensional acceleration data; the gyroscope has a range of ±2000° / s and an accuracy of ±0.1° / s, used to collect three-dimensional angular velocity data; and the temperature sensor has a range of -40℃ to 85℃ and an accuracy of ±0.5℃, used to collect ambient temperature data. These three types of data together constitute the environmental sensing data. The transducer is responsible for underwater data transmission, and the antenna module is used for aerial communication link connection, transmitting the collected environmental sensing data to the embedded processor module in real time.

[0055] The embedded processor module is a low-power embedded module, specifically the STM32H743VI, which integrates a quad-core ARMv8 floating-point arithmetic unit with a NEON coprocessor, supporting multi-threaded parallel processing. This module first uses an extended Kalman filter to initially eliminate Gaussian noise in the acceleration and angular velocity data. Then, it uses a particle filter to process nonlinear and non-Gaussian interference, such as sudden noise caused by wave impacts, with the particle number set to 500, to achieve high-precision fusion of multi-dimensional data, with the fused data error not exceeding 3%. Simultaneously, the module incorporates a reinforcement learning scheduling algorithm, using the lowest energy consumption, data integrity of at least 95%, and communication latency of no more than 1 second as the reward function to construct a task priority queue and dynamically allocate CPU resources. For example, under moderate sea conditions, the priority of data acquisition tasks is set to 0.6, communication tasks to 0.3, and non-core tasks to 0.1, thereby avoiding excessive energy consumption by a single task.

[0056] The environmental state determination unit constructs a nonlinear motion model based on fused data and performs dynamic state calculations using a Kalman filter algorithm to obtain vibration intensity data, tilt rate of change data, and trend parameters over a 10-second period. The system sets two threshold levels: the first threshold corresponds to normal sea conditions, where vibration intensity does not exceed 0.5g and tilt rate of change does not exceed 5° / s; the second threshold corresponds to extreme sea conditions, where vibration intensity exceeds 1.2g and tilt rate of change exceeds 15° / s. The environmental state is determined by combining data values ​​and trends. For example, when the vibration intensity is 0.6g and the tilt rate of change is 8° / s, both exceeding the first threshold, but the growth rate within 10 seconds is 10% and the trend is stable, it is determined to be a moderate sea state; when the vibration intensity reaches 1.3g, exceeding the second threshold, it is determined to be an extreme sea state.

[0057] The communication link scheduler uses a hybrid model of Bayesian networks and reinforcement learning to achieve optimal communication link selection. The specific process includes five steps: First, key indicators of each communication link are collected in real time, including RSSI signal strength, BER bit error rate, transmission latency, and energy consumption level, with a collection interval of 1 second; Second, each indicator is normalized to the range of 0 to 1, and initial weights are assigned according to business needs: signal strength weight is 0.3, BER weight is 0.3, latency weight is 0.2, and energy consumption weight is 0.2; Third, a comprehensive score is calculated, and links with scores higher than 0.7 enter the candidate pool, and the primary communication method and alternative communication methods are selected; Fourth, based on historical data such as the link performance over the past 72 hours and real-time environmental parameters such as temperature and wave intensity, the weights are updated through a Bayesian network, and the reinforcement learning algorithm adjusts the transmission power according to communication quality and energy consumption feedback. For example, when the 4G link BER exceeds the standard, the transmission power is increased by 0.5dBm, and when the energy consumption is too high, the transmission power is reduced by 0.3dBm; Fifth, when the score of the primary communication method is lower than 0.7 for three consecutive times, the system switches to the alternative communication method and records the fault log.

[0058] The logging and fault diagnosis module has a built-in 8GB NAND Flash memory for recording system operation logs, including sampling rate, power supply mode, communication link status, etc., as well as storing self-test results and abnormal events, such as power fluctuations, communication interruptions, and sensor failures. This module supports remote fault diagnosis and firmware upgrades, and the log storage period is no less than 90 days.

[0059] The energy management module achieves efficient energy supply and emergency backup through multi-level dynamic regulation. The dynamic power conversion unit includes a DC / DC converter chip, a coulomb counter, and an LSTM power prediction module. The DC / DC converter chip, model TPS5430, has an input voltage range of 12 to 24V and an adjustable output voltage range of 3.3 to 12V, with a minimum conversion efficiency of no less than 92%, providing stable voltage for each module. The coulomb counter, model INA226, monitors the remaining battery power in real time with a measurement accuracy of ±1%. The LSTM power prediction module constructs a 3-layer LSTM neural network based on energy consumption data from the past 24 hours and current environmental parameters such as sea state and communication frequency. The input layer has 10 neurons, the hidden layer has 20 neurons, and the output layer has 1 neuron, accurately predicting the energy consumption trend for the next hour with a prediction error of no more than 5%.

[0060] The multi-mode power supply network switches operating modes based on environmental conditions. In normal mode, a single 100Ah lithium battery pack (12V) is used. Non-core circuits, such as redundant storage modules and backup communication modules, are shut down via optocoupler-isolated relays. Huffman coding is used to compress data with a compression ratio of at least 3:1. An LSTM power prediction model predicts low-energy consumption periods, such as 2-4 AM when the sea is calm, adjusting the inertial measurement unit sampling rate to 10-20Hz, at which point system power consumption does not exceed 1.5W. In medium sea state mode, the main lithium battery pack and an auxiliary energy storage unit (a 500F supercapacitor array) are used for combined power supply. Redundancy in core communication links, such as BeiDou and data radio backups, is maintained. Improved Huffman coding is used to compress data, adaptively adjusting the coding dictionary based on data type, with a compression ratio of at least 2.5:1. A reinforcement learning algorithm dynamically adjusts the sampling rate to 20-50Hz based on energy consumption feedback, controlling system power consumption to 2-3W. In extreme event mode, the TVS surge suppression circuit and redundant lithium battery pack are activated. The TVS surge suppression circuit is model SMBJ6.5CA. Non-core functions, such as ambient temperature acquisition and non-real-time data storage, are cut off. Priority is given to powering the inertial measurement unit and Beidou communication module. The sampling rate is fixed at 50Hz and the system power consumption does not exceed 4W. The redundant battery pack can ensure that the core functions can work continuously for no less than 8 hours.

[0061] The emergency energy storage unit employs a 500F supercapacitor array with a charge / discharge efficiency of no less than 95%. In the event of power anomalies such as instantaneous overvoltage or lithium battery pack failure, it responds within 10 milliseconds, providing buffer power to critical circuits including the embedded processor, inertial measurement unit, and BeiDou communication module for at least 30 seconds, allowing time for backup power switching. Simultaneously, the system employs a self-checking protection mechanism, monitoring battery health every hour. If capacity degradation exceeds 20% or internal resistance exceeds 100mΩ, an alarm is triggered, and the system switches to backup lithium battery pack power.

[0062] Example 2:

[0063] Based on the above system, this embodiment proposes an energy optimization management method for air-sea cross-domain communication gateways in coupled environments, such as... Figure 3 and Figure 4As shown, the process includes the following steps: An inertial measurement unit (IMU) is placed in an ocean buoy to collect acceleration, angular velocity, and ambient temperature data in real time; the collected data is analyzed and denoised using a fusion algorithm combining extended Kalman filtering and particle filtering; a nonlinear motion model is constructed based on the fused data; the environmental state is determined by threshold comparison and trend prediction; the power supply mode of the system modules is switched in real time according to the environmental state determination results; the communication mode of the modules is dynamically switched and the transmission power is adjusted using an optimal communication link selection algorithm; the remaining battery power and energy consumption trend are monitored jointly using a coulomb counter and an LSTM power prediction model; and the sampling rate of the IMU is adjusted using a reinforcement learning algorithm.

[0064] In specific application scenarios, the system first deploys and initializes the equipment. The communication buoy with integrated inertial measurement unit (IMU) is deployed into the target sea area; the buoy automatically deploys, the transducer sinks underwater, and the antenna module surfaces. After power-on, the system executes a self-test program to verify the status of the IMU sensor, power module, communication links (including 4G, BeiDou, and underwater acoustic communication), and energy storage unit. The self-test results are logged and uploaded to the shore-based platform.

[0065] After deployment, the system enters normal operation under normal conditions. The determination of the normal operating mode is based on real-time acquisition of vibration intensity and tilt rate of change data by the IMU. When the vibration intensity does not exceed 0.5g, the tilt rate of change does not exceed 5° / s, and the trend is stable, the system executes the normal operating mode, using a single lithium battery pack for power, shutting down non-core circuits, compressing data using Huffman coding and transmitting it back, and using an LSTM power prediction model to predict low-energy consumption periods, adjusting the sampling rate of the inertial measurement unit to 10-20Hz to maximize energy saving. When the vibration intensity or tilt rate of change exceeds the first threshold but not the second threshold, or when the trend shows fluctuating growth, the system switches to a medium sea state mode, using a combined power supply from the main lithium battery pack and a supercapacitor array, maintaining redundancy in the core communication link, using improved Huffman coding to compress data, and a reinforcement learning algorithm dynamically adjusts the sampling rate to 20-50Hz based on energy consumption feedback, simultaneously optimizing communication transmission power to achieve a dynamic balance between energy consumption and performance. When the vibration intensity exceeds 1.2g or the tilt angle change rate exceeds 15° / s, the system immediately switches to extreme event mode, activates the TVS surge suppression circuit and redundant lithium battery pack, cuts off non-core functions, fixes the sampling rate at 50Hz, and prioritizes power supply to the data acquisition of the inertial measurement unit and the Beidou communication module to ensure that critical data is not lost.

[0066] During communication, the system collects key indicators such as signal strength, bit error rate, latency, and energy consumption level for each communication method every second. These indicators are normalized and weighted. Communication methods with scores higher than 0.7 are added to a candidate pool, with the highest-scoring method selected as the primary communication method and the second-highest-scoring method as a backup. Based on historical data and real-time environmental parameters, the weight allocation is updated using a hybrid model of Bayesian network and reinforcement learning. The reinforcement learning algorithm dynamically adjusts the transmission power of each communication method based on communication quality and energy consumption feedback. When the score of the primary communication method falls below the threshold for three consecutive times, the system switches to a backup communication method and records a fault log.

[0067] After the detection task is completed, the main control module sends a termination command, shutting down underwater acoustic communication and sensors, leaving only the BeiDou module in standby mode. The system enters a low-power state, periodically waking up to report a heartbeat signal until the next task is triggered. If a fault occurs during operation, such as the supercapacitor detecting a momentary overvoltage, the TVS diode array will initiate discharge, and the system will switch to the backup battery. If a communication interruption occurs, such as after three consecutive link connection failures, the main control module will restart the data transmission radio and send a fault code to the nearest buoy network relay via underwater acoustic communication.

[0068] Reference Figure 5 The inertial measurement unit (IMU) module is located at the center of the buoy, integrated with the power supply module, facilitating system power-on and periodic self-testing. Once deployed in water, the IMU floats on the surface. The transducer is underwater, while the antenna module floats on the surface for receiving and transmitting information.

[0069] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0070] It should be noted that the components mentioned in the above embodiments are all general standard parts or components known to those skilled in the art. Their structures and principles can be learned by those skilled in the art through technical manuals or conventional experimental methods.

[0071] This invention has illustrated its principles and implementation methods using specific examples. The descriptions of these embodiments are merely illustrative of the method and its core ideas; furthermore, those skilled in the art will recognize that modifications may be made to the specific implementation methods and application scope based on the principles of this invention. Therefore, the content of this specification should not be construed as limiting the invention.

Claims

1. An energy optimization management system for air-sea cross-domain communication gateways in coupled environments, characterized in that, It includes a main control module and an energy management module, wherein the main control module includes: An inertial measurement unit, which is used to acquire environmental sensing data in real time; An embedded processor module, which employs a floating-point arithmetic unit, combines an extended Kalman filter and a particle filter fusion algorithm for data parsing and noise reduction fusion, and incorporates a reinforcement learning scheduling algorithm to coordinate task priorities; An environmental state determination unit is used to construct a nonlinear motion model based on the environmental sensing data and analyze the environmental state through threshold comparison and trend prediction. A communication link scheduler, which is used to dynamically switch the communication mode of the module using an optimal communication link selection algorithm; A recording and fault diagnosis module is used to store system operation logs, self-test results, and abnormal events; The energy management module includes: A dynamic power conversion unit is used to provide a stable voltage and is equipped with an LSTM power prediction model to monitor the remaining battery capacity and energy consumption trend of the battery pack. A multi-mode power supply network is used to switch the battery pack power supply mode according to the environmental conditions. The operating modes include daily mode, medium sea state mode and extreme event mode. An emergency energy storage unit is used to provide buffer power supply in the event of a power failure.

2. The energy optimization management system for air-sea cross-domain communication gateways in a coupled environment as described in claim 1, characterized in that, The inertial measurement unit is located at the center of the ocean buoy and includes an accelerometer, a gyroscope, a transducer, an antenna module, and a temperature sensor. The accelerometer and gyroscope collect acceleration and angular velocity data, and the temperature sensor collects ambient temperature data, which together serve as the environmental sensing data. The antenna module transmits the environmental sensing data to the embedded processor module.

3. The energy optimization management system for air-sea cross-domain communication gateways in a coupled environment as described in claim 1, characterized in that, The embedded processor module initially eliminates Gaussian noise in the environmental sensing data through extended Kalman filtering, and then processes nonlinear and non-Gaussian noise through particle filtering to complete multi-dimensional data fusion. The environmental state determination unit constructs a nonlinear motion model based on the fused data, performs motion state calculation and trend prediction, obtains vibration intensity data, tilt angle change rate data and change trend parameters, and determines the environmental state by comparing thresholds and trend analysis.

4. The energy optimization management system for air-sea cross-domain communication gateways in a coupled environment according to claim 3, characterized in that, When the vibration intensity data and the rate of change of tilt angle data do not exceed the first threshold and the trend of change is stable, the daily mode is executed; when the data exceeds the first threshold but does not exceed the second threshold, or the trend of change shows fluctuating growth, the medium sea state mode is executed. When the data exceeds the second threshold, the extreme event mode is executed; The daily mode includes: using a single lithium battery pack for power supply, shutting down non-core circuits through optocoupler isolation relays, using Huffman coding to compress data and transmit data back, using an LSTM power prediction model to predict low-energy periods, and appropriately reducing the sampling rate of the inertial measurement unit. The medium sea state mode includes: using a main lithium battery pack and an auxiliary energy storage unit for power supply, retaining redundancy in the core communication link, using improved Huffman coding to compress data, and dynamically balancing the sampling rate and communication frequency through reinforcement learning algorithms; The extreme event modes include: activating the TVS surge suppression circuit and redundant lithium battery pack, cutting off non-core functions, and prioritizing power supply to the data acquisition and BeiDou communication modules of the inertial measurement unit.

5. The energy optimization management system for air-sea cross-domain communication gateways in a coupled environment according to claim 1, characterized in that, The module communication methods include 4G transmission, Beidou communication, data transmission radio, and underwater acoustic communication. The communication link scheduler collects key indicators of the link in each communication method every 1 second, normalizes the key indicators and assigns weights. Communication methods with scores higher than the threshold enter the candidate pool, and the highest score in the candidate pool is taken as the main communication method, and the second highest score is taken as the alternative communication method. During communication, a hybrid model of Bayesian network and reinforcement learning is used to update weight allocation in real time based on historical data and real-time environmental parameters. When the score of the main communication method is lower than the threshold for three consecutive times, the system switches to the alternative communication method and records the fault log. The reinforcement learning algorithm dynamically adjusts the transmission power of each communication method based on energy consumption and communication quality feedback.

6. The energy optimization management system for air-sea cross-domain communication gateways in a coupled environment according to claim 1, characterized in that, The dynamic power conversion unit includes a DC / DC conversion chip, a coulomb counter, and an LSTM power prediction module. The DC / DC conversion chip outputs an adjustable voltage based on the input specified voltage. The coulomb counter monitors the remaining battery power in real time, and the LSTM power prediction module predicts the energy consumption trend for the next hour based on historical energy consumption data and environmental condition parameters. The embedded processor module combines the remaining battery power and energy consumption trend, and adjusts the sampling rate of the inertial measurement unit and the transmission power of the communication link through a reinforcement learning algorithm.

7. The method for optimizing the energy management of an air-sea cross-domain communication gateway in a coupled environment according to any one of claims 1-6, characterized in that, Includes the following steps: S1. An inertial measurement unit is installed in an ocean buoy to collect acceleration data, angular velocity data, and ambient temperature data in real time; S2. The data is analyzed and denoised using a fusion algorithm of extended Kalman filter and particle filter. A nonlinear motion model is constructed based on the fused data, and the environmental state is judged by threshold comparison and trend prediction. S3. Switch the power supply mode of the system modules in real time based on the environmental condition determination results; S4. Employ the optimal communication link selection algorithm to dynamically switch module communication modes and adjust transmission power; S5. The remaining battery charge and energy consumption trend are monitored by combining a coulomb counter and an LSTM power prediction model, and the sampling rate of the inertial measurement unit is adjusted by combining a reinforcement learning algorithm.

8. The energy optimization management method for air-sea cross-domain communication gateways in coupled environments according to claim 7, characterized in that, In step S2, the process of constructing a nonlinear motion model based on fused data and determining the environmental state through threshold comparison and trend prediction includes: A nonlinear motion model is constructed based on fused data to calculate vibration intensity data, tilt rate of change data, and trend parameters. When the data does not exceed the first threshold and the trend is stable, the normal mode is executed. When the data exceeds the first threshold but does not exceed the second threshold, or when the trend fluctuates and increases, the moderate sea state mode is executed. When the data exceeds the second threshold, the extreme event mode is executed.

9. The energy optimization management method for air-sea cross-domain communication gateways in coupled environments according to claim 8, characterized in that, The daily mode includes: using a single lithium battery pack for power supply, shutting down non-core circuits through optocoupler isolation relays, compressing and transmitting data using Huffman coding, using an LSTM power prediction model to predict low-energy consumption periods, and adjusting the sampling rate of the inertial measurement unit to 10-20Hz. The medium sea state mode includes: using a main lithium battery pack and an auxiliary energy storage unit for power supply, retaining redundancy in the core communication link, using improved Huffman coding to compress data, and using a reinforcement learning algorithm to dynamically adjust the sampling rate to 20-50Hz based on energy consumption feedback, while simultaneously optimizing the communication transmission power. The extreme event mode includes: activating the TVS surge suppression circuit and redundant lithium battery pack, cutting off non-core functions, fixing the sampling rate to 50Hz, and prioritizing power supply to the data acquisition of the inertial measurement unit and the Beidou communication module.

10. The energy optimization management method for air-sea cross-domain communication gateways in a coupled environment according to claim 7, characterized in that, Step S4, which involves dynamically switching the module's communication mode and adjusting the transmission power using the optimal communication link selection algorithm, includes: Every 1 second, key indicators such as signal strength, bit error rate, latency, and energy consumption level of each communication method are collected, normalized, and weighted. Communication methods with scores higher than the threshold are entered into the candidate pool, and the communication method with the highest score in the candidate pool is selected as the primary communication method, and the second highest score is selected as the alternative communication method. Based on historical data and real-time environmental parameters, the weight allocation is updated through a hybrid model of Bayesian network and reinforcement learning. The reinforcement learning algorithm dynamically adjusts the transmission power of each communication method according to feedback on communication quality and energy consumption. If the score of the primary communication method falls below the threshold for three consecutive times, the system will switch to the alternative communication method and record a fault log.

Images