A range extended hydrogen fuel power ship propulsion system

Abstract

The application discloses a range-extending hydrogen fuel power ship propulsion system and relates to the technical field of ship propulsion.The range-extending hydrogen fuel power ship propulsion system comprises a collection module, which collects hydrogen energy ship power parameters, removes noise interference through a Kalman filtering algorithm, and generates a standardized data packet; a quantum annealing optimization module, which maps the hydrogen energy ship power parameters to an Ising model after receiving the standardized data packet, generates a control instruction according to a hydrogen fuel consumption rate, a device service life attenuation coefficient and a navigation task priority weight; and a proton membrane regulation module, which inputs the control instruction into a simulation model of a proton exchange membrane, dynamically adjusts hydrogen production rate and fuel cell power, and outputs real-time hydrogen storage tank pressure data.The range-extending hydrogen fuel power ship propulsion system adopts an efficient hydrogen energy closed-loop monitoring unit and a three-level safety protocol, thereby not only improving response speed and accuracy, but also significantly enhancing stability and safety, and ensuring that the ship can reliably operate under complex working conditions and potential safety risks are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of ship propulsion technology, in particular to a range-extended hydrogen fuel power ship propulsion system. BACKGROUND

[0002] With the global emphasis on environmental protection and sustainable development, hydrogen energy as a clean energy in the field of transportation has gradually attracted widespread attention. In particular, in the ship propulsion system, hydrogen fuel cells have become an important development direction for future green ships due to their zero emissions and high efficiency. In recent years, by combining the advantages of hydrogen fuel cells and lithium batteries, not only the endurance and power performance of the ship have been improved, but also the dependence on traditional fossil fuels has been reduced.

[0003] However, the existing technology still faces many challenges. On the one hand, the existing range-extended hydrogen fuel power ship propulsion system usually relies on fixed control logic to adjust the working state of hydrogen fuel cells and lithium batteries. This static control method cannot adapt to dynamic changes in sailing conditions, thereby affecting the overall energy efficiency. On the other hand, for the pressure monitoring of hydrogen storage tanks and the regulation of electrolytic current, although the traditional method can achieve basic regulation function, its response speed and accuracy are still insufficient when facing complex working conditions. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a range-extended hydrogen fuel power ship propulsion system to solve the problems of static control not adapting to dynamic changes and traditional methods responding slowly.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] The present application provides a range-extended hydrogen fuel power ship propulsion system, which comprises a collection module that collects hydrogen energy ship power parameters, removes noise interference through Kalman filtering algorithm, and generates standardized data packets;

[0008] A quantum annealing optimization module receives the standardized data packets, maps the hydrogen energy ship power parameters to the Ising model, and generates control instructions according to the hydrogen fuel consumption rate, equipment life decay coefficient and sailing task priority weight;

[0009] A proton membrane regulation module inputs the control instructions into the simulation model of the proton exchange membrane, dynamically adjusts the hydrogen production rate and fuel cell power, and outputs real-time hydrogen storage tank pressure data;

[0010] A safety monitoring module, based on real-time hydrogen storage tank pressure data, dynamically adjusts the electrolysis current and fuel cell power ratio of the hydrogen energy closed-loop monitoring unit, synchronously monitors the hydrogen purity and oxygen residual amount, and triggers a three-level safety protocol, outputs the electrolysis waste heat temperature, cooling liquid flow rate and safety status code;

[0011] A waste heat recovery and self-checking module, based on the parameters of electrolysis waste heat temperature and cooling liquid flow rate, the waste heat recovery converter increases the coolant temperature and reduces the cooling energy consumption through phase change materials, and the quantum chip performs equipment health self-checking;

[0012] A maintenance feedback module, based on the self-checking results, generates a maintenance recommendation report and uploads it to the shore-based data center.

[0013] As a preferred scheme of the range-extending hydrogen fuel power ship propulsion system of the application, wherein: the hydrogen energy ship power parameters include: lithium battery pack remaining power, hydrogen storage tank internal pressure, real-time power demand of propeller, fuel cell stack temperature, seawater flow rate, ship draft depth, hydrogen fuel consumption rate, equipment life attenuation coefficient and sailing task priority weight.

[0014] As a preferred scheme of the range-extending hydrogen fuel power ship propulsion system of the application, wherein: the standardization data packet acquisition step is as follows,

[0015] After the collected hydrogen energy ship power parameters are denoised by Kalman filtering and combined with the isolated forest algorithm to remove outliers, Min-Max normalization and One-Hot encoding are used to generate a standardized data packet.

[0016] As a preferred scheme of the range-extending hydrogen fuel power ship propulsion system of the application, wherein: the step of mapping the hydrogen energy ship power parameters to the Ising model is as follows,

[0017] The hydrogen fuel consumption rate is encoded as the antiferromagnetic coupling strength, the equipment life attenuation coefficient is mapped as the local magnetic field bias, and the sailing task priority is converted into the constraint energy term;

[0018] Based on the antiferromagnetic coupling strength, the local magnetic field bias and the constraint energy term, a multi-objective energy function of the Ising model is constructed in the quantum simulation algorithm, and the hydrogen energy ship power parameters are mapped to the Ising model.

[0019] As a preferred scheme of the range-extending hydrogen fuel power ship propulsion system of the application, wherein: the control instruction acquisition step is as follows,

[0020] After mapping the hydrogen energy ship power parameters to the Ising model, the anti-ferromagnetic coupling matrix, the local magnetic field vector and the constraint energy item are dynamically synthesized through the matrix operation framework of TensorFlow, and a multi-objective optimization Hamiltonian model is constructed by combining an adaptive weight distribution algorithm;

[0021] Based on the multi-objective optimization Hamiltonian model, the spin state parameters are obtained by iterative solving under the TensorFlow calculation framework through a simulated annealing algorithm, and distributed computing is used.

[0022] According to the spin state parameters, control instructions for fuel cell power and hydrogen production rate are dynamically generated.

[0023] As a preferred scheme of the range-extending hydrogen fuel power ship propulsion system of the application, wherein the real-time hydrogen storage tank pressure data is obtained by the following steps,

[0024] Based on the control instructions, the hydrogen production rate set value and the fuel cell power output are extracted and dynamically loaded into the simulation model of the proton exchange membrane, and the difference between the hydrogen generation and consumption rates is calculated synchronously through the mass conservation equation;

[0025] Based on the difference between the hydrogen generation and consumption rates, the hydrogen production rate set value is corrected in real time through a PID control algorithm and a fluid mechanics algorithm, and the fuel cell power output is dynamically adjusted in combination with an adaptive feedback adjustment mechanism of an equivalent circuit;

[0026] Based on the corrected hydrogen production rate set value and the adjusted fuel cell power output, the Runge-Kutta numerical integration is used to solve the hydrogen storage tank pressure change curve to obtain the real-time hydrogen storage tank pressure data.

[0027] As a preferred scheme of the range-extending hydrogen fuel power ship propulsion system of the application, wherein the hydrogen energy closed-loop monitoring unit dynamically adjusts the electrolysis current and fuel cell power ratio, and the steps are as follows,

[0028] Based on the real-time hydrogen storage tank pressure data, the Kalman filter is combined with the long short-term memory network to predict the pressure trend and generate a safety pressure threshold;

[0029] The hydrogen energy closed-loop monitoring unit calculates the deviation percentage of the real-time hydrogen storage tank pressure and the safety pressure threshold, divides the deviation level, matches different adjustment intensity weights according to the deviation level, and dynamically adjusts the electrolysis current and fuel cell power ratio.

[0030] As a preferred scheme of the range-extending hydrogen fuel power ship propulsion system of the application, wherein the three-level safety protocol specifically includes,

[0031] Based on the hydrogen storage tank pressure, hydrogen purity, oxygen residual amount and electrolysis waste heat temperature, multi-dimensional threshold values are defined.

[0032] The multi-dimensional threshold is classified into a first-level early warning, a second-level production limit and a third-level emergency shutdown.

[0033] As a preferred scheme of the range-extended hydrogen fuel power ship propulsion system, the steps of the waste heat recovery converter for improving the coolant temperature and reducing the cooling energy consumption through the phase change material are as follows,

[0034] Based on the real-time data of the electrolysis waste heat temperature and the cooling liquid flow, a phase change material thermal behavior simulation model is constructed through finite element analysis.

[0035] Based on the phase change material thermal behavior simulation model, a virtual PID controller is used to adjust the virtual coolant flow parameter and generate an optimal flow control strategy, and the optimal flow control strategy is used to guide the cooling pump to adjust the actual coolant temperature.

[0036] As a preferred scheme of the range-extended hydrogen fuel power ship propulsion system, the steps of the quantum chip for performing the equipment health self-check are as follows,

[0037] The quantum chip uses a quantum annealing algorithm to analyze the hydrogen energy ship power parameters, and uses a deep neural network to predict the failure mode, and dynamically adjusts the detection strategy based on reinforcement learning, so as to realize the equipment health self-check.

[0038] The present application has the following advantages:

[0039] By real-time acquisition and processing of the power parameters of the hydrogen energy ship, and using advanced Kalman filtering algorithm and quantum annealing optimization technology to generate accurate control instructions, the output power of the hydrogen fuel cell and the lithium battery pack can be dynamically adjusted according to the real-time sailing conditions, so as to significantly improve the overall energy efficiency of the ship. By using an efficient hydrogen energy closed-loop monitoring unit and a three-level safety protocol, the electrolysis current and the fuel cell power ratio are dynamically adjusted based on the real-time hydrogen storage tank pressure data, and different adjustment intensity weights are matched according to the deviation level, which not only improves the response speed and accuracy, but also significantly enhances the stability and safety, ensuring that the ship can operate reliably under complex working conditions and reducing potential safety risks. BRIEF DESCRIPTION OF DRAWINGS

[0040] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0041] Fig. 1 The figure is a schematic diagram of the range-extended hydrogen fuel power ship propulsion system in the present application.

[0042] Fig. 2 Flow chart for the acquisition step of the control instruction in the present application.

[0043] Fig. 3 Flow chart for the quantum annealing optimization module in the present application.

[0044] Fig. 4 Flow chart for the acquisition of the control instruction by the standardized data packet in the present application. DETAILED DESCRIPTION

[0045] In order to make the above objectives, features and advantages of the present application more apparent, specific embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0046] In the following description, a large number of specific details are set forth in order to facilitate a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the spirit of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0047] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0048] Reference Figs. 1-4 For the first embodiment of the present application, the embodiment provides a range-extended hydrogen fuel power ship propulsion system, comprising the following steps:

[0049] The acquisition module generates a standardized data packet by eliminating noise interference through Kalman filtering algorithm on the collected hydrogen energy ship power parameters.

[0050] Specifically, the following steps are included,

[0051] In the acquisition process, advanced sensor networks are used, and the sensor networks are distributed in various key parts of the ship, including lithium battery pack, hydrogen storage tank, fuel cell stack, propeller, seawater environment monitoring point, ship draft monitoring point, hydrogen fuel consumption rate monitoring point, engine, compressor and navigation task control hub.

[0052] In specific operation, the current integration method and open circuit voltage method are used to measure the remaining power of lithium battery pack, the piezoresistive pressure sensor is used to monitor the internal pressure of hydrogen storage tank, the power sensor is used to detect the real-time power demand of thruster, the thermocouple and infrared temperature sensor are used to measure the temperature of fuel cell stack, the Doppler sonar sensor is used to measure the seawater flow rate, the ultrasonic sensor or pressure sensor is used to evaluate the draft of the ship, the flowmeter or mass sensor is used to monitor the hydrogen fuel consumption rate, the multi-source data such as vibration sensor and temperature sensor are used to estimate the equipment life attenuation coefficient, and the navigation task priority weight is determined according to the task planning center;

[0053] The collected ship power parameters are transmitted to the central processor to form the original data set. Due to the influence of environmental noise and sensor failure, the original ship power parameters usually contain interference and outliers, so further processing is needed to improve the data quality. In order to eliminate noise interference and remove outliers, an adaptive Kalman filtering algorithm is used for noise reduction processing of ship power parameters. Kalman filtering is a recursive optimal estimation method, which can effectively eliminate noise interference in dynamic environment. In specific operation, the Kalman filter is initialized, and the predicted value of ship power parameters at the current time is obtained. Based on the real-time observation of ship power parameters, the predicted value of ship power parameters is updated and iterated constantly, so as to obtain more accurate predicted value of ship power parameters.

[0054] After Kalman filtering noise reduction, outliers are identified and removed by isolated forest algorithm. Isolated forest algorithm quickly locates outliers by randomly dividing ship power parameter set and removes them from ship power parameter set, so as to ensure the reliability of data. In specific operation, isolated forest algorithm constructs multiple isolated trees, calculates the path length of each sample point of ship power parameter in isolated tree, and removes the sample point with shorter path as outlier. This unsupervised learning method not only can efficiently identify outliers, but also can maintain the overall integrity of data set, ensuring the accuracy of subsequent analysis.

[0055] After noise reduction and outlier removal, standardization processing is carried out for subsequent analysis and processing. First of all, Min-Max normalization method is used to map the value range of each ship power parameter to [0, 1] interval to eliminate the influence of different dimensions. In this way, all ship power parameters are converted to a unified scale, making them comparable. For classification variables such as navigation task priority weight, One-Hot encoding method is used. One-Hot encoding converts each classification value into an independent binary vector. For example, navigation task priority can be represented as different binary vectors, each vector corresponds to a priority level. This encoding method converts classification variables into numerical vectors, so that they can be directly used by machine learning algorithms.

[0056] The pre-processed ship power parameters are integrated through data fusion and structured packaging technology to generate standardized data packages. These standardized data packages not only eliminate noise interference and abnormal values, but also unify the data format, facilitating subsequent analysis and providing a solid foundation for subsequent energy management and safety monitoring.

[0057] The quantum annealing optimization module receives the standardized data package and maps the hydrogen energy ship power parameters to the Ising model. It also considers the hydrogen fuel consumption rate, equipment life decay coefficient, and sailing task priority weight to generate control instructions.

[0058] Specifically, the method comprises the following steps:

[0059] When the standardized data package is transmitted to the quantum chip, the quantum chip performs parsing and encoding processing. To adapt to the requirements of the quantum computing framework, a specific encoding method is used to convert different types of parameters in the standardized data package into physical quantities such as anti-ferromagnetic coupling strength, local magnetic field bias, and constraint energy item. The specific operation is as follows:

[0060] First, the interaction between each ship power parameter is identified and quantified through correlation analysis, and a nonlinear coupling coefficient matrix is generated through partial least squares regression. The nonlinear coupling coefficient matrix is then normalized by the quantum state density function, and energy scale is scaled through Ising parameter linear transformation to map the nonlinear coupling coefficient matrix to anti-ferromagnetic coupling strength.

[0061] Secondly, for the influence of external environmental factors such as sea water temperature and flow rate on the propulsion process, the influence mode of external environmental factors on ship propulsion performance is obtained through wavelet multi-scale analysis by environmental modeling analysis method, and the influence mode is parameterized and encoded into local magnetic field bias through local field parameterization.

[0062] In addition, based on the standardized data package, dynamic weight allocation is performed using particle swarm optimization through priority algorithm to obtain the constraint energy item, which is used to meet the hard requirements of sailing task priority;

[0063] The core of the encoding method is to abstract the complex multi-objective optimization problem into the physical quantity relationship in the Ising model, so as to facilitate subsequent quantum simulation and optimization solution. In this process, an adaptive weight allocation algorithm is combined to dynamically adjust the weight for different optimization objectives, ensuring that each objective can be reasonably balanced in global optimization.

[0064] Based on the encoded antiferromagnetic coupling strength, local magnetic field bias and constraint energy term, in the quantum simulation algorithm, the matrix operation framework of TensorFlow is used to dynamically synthesize the antiferromagnetic coupling matrix and the local magnetic field vector, and the constraint energy term is designed combined with the adaptive weight distribution algorithm, a multi-objective energy function of the Ising model is constructed through multi-objective optimization modeling technology, and at the same time, by integrating the interaction between dynamic parameters, external environmental factors and task priority requirements into a unified multi-objective energy function expression, the hydrogen energy ship power parameter is mapped into the Ising model;

[0065] After completing the mapping of the ship power parameters, enter the multi-objective optimization phase, in this phase, based on the custom layer of TensorFlow, the antiferromagnetic coupling matrix, the local magnetic field vector and the constraint energy term are fused, and through the dynamic weight distribution mechanism of the integrated LSTM network, the constraint strength is adaptively adjusted according to the real-time working condition data of the ship, a multi-objective optimization Hamiltonian model is constructed, which considers multiple optimization objectives, and the antiferromagnetic coupling matrix, the local magnetic field vector and the constraint energy term are integrated into the total energy function. In order to solve this total energy function, the simulated annealing algorithm is adopted, which is a global optimization method and is particularly suitable for complex multi-objective optimization problems. In the specific operation, the parameters of the simulated annealing algorithm are initialized, including the initial temperature and the cooling rate. Then in each iteration, a new solution is randomly generated according to the current temperature and the Hamiltonian model, and its energy value is evaluated. If the energy value of the new solution is lower, the solution is accepted, otherwise, whether to accept the new solution is determined according to the current temperature and the energy difference to avoid falling into a local optimal solution. Through the simulated annealing algorithm, the global optimal solution can be gradually approached in the complex solution space. At the same time, make full use of distributed computing technology to distribute the calculation tasks of the simulated annealing algorithm to multiple nodes for parallel execution, further shorten the solving time and improve the calculation efficiency;

[0066] After multiple iterations, the simulated annealing algorithm finally determines the spin state parameters, which are the embodiment of the multi-objective optimization results, specifically reflecting the optimal configuration of fuel cell power and hydrogen production rate under the constraints of meeting energy demand and device life management, etc. Based on the spin state parameters, control instructions for fuel cell power and hydrogen production rate are dynamically generated. In specific operations, through the attention mechanism feature extraction technology, the spin state parameters are analyzed by sliding window segmentation sampling, and the optimization results related to fuel cell power and hydrogen production rate are extracted by gradient boosting regression technology, and specific control instructions are generated according to these results, including fuel cell power adjustment instructions, hydrogen production rate setting instructions and device collaborative linkage instructions. These control instructions directly guide the working state of fuel cells and hydrogen production equipment to ensure that they operate under optimal conditions. In order to further improve stability and reliability, a closed-loop feedback mechanism is designed to monitor the actual operating state of fuel cells and hydrogen production equipment in real time and dynamically adjust the control instructions according to the feedback information to ensure continuous optimization.

[0067] Through the above operations, the combination of various advanced technical means not only improves the response speed and accuracy, but also significantly enhances the stability and safety.

[0068] The proton membrane regulation module inputs the control instructions into the simulation model of the proton exchange membrane, dynamically adjusts the hydrogen production rate and fuel cell power, and outputs real-time hydrogen storage tank pressure data.

[0069] Specifically, the following steps are included,

[0070] Based on the received control instructions, the initial hydrogen production rate setting value and fuel cell power output are extracted. The hydrogen production rate setting value is the target value of the hydrogen production equipment per unit time determined by the multi-objective optimization algorithm. The initial hydrogen production rate setting value and fuel cell power output are dynamically loaded into the simulation model of the proton exchange membrane. The simulation model of the proton exchange membrane is a highly accurate mathematical model that can simulate the working state of the proton exchange membrane, including the process of hydrogen production by electrolysis of water and the electrochemical reaction of the fuel cell. Through the mass conservation equation, the difference between hydrogen generation and consumption rates is calculated simultaneously. The mass conservation equation ensures the balance between input and output, that is, the amount of hydrogen generated must be equal to the amount of hydrogen consumed plus the amount of remaining hydrogen in the hydrogen storage tank. This step provides basic data for subsequent real-time correction. The specific mathematical formula is:

[0071] ; wherein, t represents the sampling time, t represents the hydrogen generation and consumption rate difference at the t sampling time, t represents the initial hydrogen production rate setting value, t represents the hydrogen generation and consumption rate difference at the PID controller output value at a sampling time, representing the real-time power output value of the fuel cell at a sampling time, representing the real-time power output value of the fuel cell at a sampling time, representing the energy conversion efficiency of the fuel cell, representing the energy value released when a unit mass of hydrogen is completely burned;

[0072] Based on the difference between the hydrogen generation and consumption rates, a PID control algorithm and a fluid mechanics algorithm are used for real-time correction. The PID control algorithm is a classic closed-loop control algorithm used to adjust the initial hydrogen production rate set value. Its principle is to process the difference through proportional (P), integral (I), and derivative (D) parts to achieve accurate control. In specific operations, first, the expected value is defined based on the dynamic balance between the hydrogen production rate set value and the fuel cell power output, as well as the change curve of the hydrogen storage tank pressure. When the difference between the hydrogen generation and consumption rates deviates from the expected value, the PID controller dynamically adjusts the initial hydrogen production rate set value according to the current difference, cumulative difference, and difference rate to reduce errors and return the hydrogen content to a stable state.

[0073] At the same time, the fuel cell power output is dynamically adjusted in combination with the adaptive feedback adjustment mechanism of the equivalent circuit. The equivalent circuit dynamically optimizes the fuel cell power output by constructing a dynamic hybrid model of the fuel cell, using online parameter identification technology combined with the maximum power point tracking algorithm. The dynamic hybrid model is a simplified representation of the internal electrochemical process of the fuel cell, which considers electrical characteristics such as resistance, capacitance, and inductance. By monitoring the actual output voltage and current of the fuel cell, the resistance, capacitance, and inductance parameters in the equivalent circuit can be adjusted in real time to optimize the power output of the fuel cell. This adaptive feedback adjustment mechanism not only improves the response speed but also enhances the robustness and stability.

[0074] Based on the corrected hydrogen production rate set value and the adjusted fuel cell power output, the Runge-Kutta numerical integration method is used to solve the hydrogen storage tank pressure change curve to obtain real-time hydrogen storage tank pressure data. This process starts with establishing a hydrogen storage tank pressure change model. According to the ideal gas law or actual gas state equation, the dynamic relationship between hydrogen flow and pressure is established through mass conservation differentiation. Combined with finite volume discretization and nonlinear correction technology, a mathematical model is constructed to describe the state change of the gas in the hydrogen storage tank. This hydrogen storage tank pressure change model considers various factors affecting the hydrogen storage tank pressure, such as temperature, volume, and mole number, ensuring accurate reflection of the actual conditions inside the hydrogen storage tank. Next, a series of key parameters are initialized, including the initial pressure and temperature of the hydrogen storage tank. These initial key parameters are crucial for subsequent accurate calculations as they provide the starting point for the entire simulation process.

[0075] After the key parameters are initialized, the fourth-order Runge-Kutta method (RK4) is used to solve the numerical solution based on the hydrogen storage tank pressure change model. In the specific operation, the current hydrogen pressure state quantity in the hydrogen storage tank within the time step is predicted by four times of local approximation, and four intermediate points are obtained based on the Taylor expansion truncation error control. The slopes of the four intermediate points are calculated, and the hydrogen pressure state quantity in the hydrogen storage tank at the next time step is obtained by weighted average based on the slopes. Through the adaptive step size control technology, the change curve of the hydrogen storage tank pressure with time is finally generated to obtain the hydrogen storage tank pressure data. These hydrogen storage tank pressure data not only reflect the actual pressure change of hydrogen in the hydrogen storage tank, but also provide an important basis for subsequent energy management and safety monitoring. For example, in the case of high load, the hydrogen production rate and fuel cell power output can be dynamically adjusted according to the real-time hydrogen storage tank pressure data to ensure the sufficiency and stability of hydrogen supply. In addition, through the adaptive feedback adjustment mechanism, potential problems can be found and corrected in time to avoid energy waste or safety hazards caused by insufficient or excessive hydrogen.

[0076] Through the above operation, the simulation model of the proton exchange membrane provides a solid technical guarantee for the efficient operation of the hydrogen energy ship.

[0077] The safety monitoring module dynamically adjusts the electrolysis current and fuel cell power ratio based on real-time hydrogen storage tank pressure data, synchronously monitors hydrogen purity and oxygen residual amount, and triggers a three-level safety protocol to output electrolysis waste heat temperature, cooling fluid flow and safety status code.

[0078] Specifically, the following steps are included,

[0079] Based on real-time hydrogen storage tank pressure data, Kalman filtering technology is used to remove noise interference to ensure the accuracy and stability of the data. The noise-reduced hydrogen storage tank pressure data is input into a long short-term memory network (LSTM) for trend prediction. LSTM is a deep learning model, and its construction process is as follows: based on the input structure of the time sequence sliding window, a three-layer gated network is built through the Keras framework, with 64 units in the input layer, 32 units in the hidden layer, and 1 unit in the output layer. The combination of Sigmoid and Tanh activation functions is used to realize the dynamic adjustment of the forget gate, input gate and output gate. The long short-term memory network model is constructed. Next, training is performed using historical hydrogen storage tank pressure time series data for supervised learning with mean square error as the loss function and Adam optimizer for gradient descent. Dropout layer and early stopping mechanism are introduced to prevent overfitting. Finally, the trained long short-term memory network model is obtained.

[0080] The long short-term memory network model after training is particularly suitable for processing time series data and can capture long-term dependencies. Through learning from historical hydrogen storage tank pressure data, the LSTM can predict the pressure trend in the future, including possible pressure peaks or downward trends. Based on the pressure trend prediction results, combined with the ship's operating conditions and safety requirements, a dynamic safety pressure threshold is generated, which not only considers the current pressure state but also integrates future trend predictions, providing a solid foundation for subsequent operations.

[0081] The hydrogen energy closed-loop monitoring unit calculates the difference between the real-time hydrogen storage tank pressure and the safety pressure threshold based on the deviation percentage, and divides it into different deviation levels, such as slight deviation, moderate deviation, and severe deviation. Each level corresponds to a different adjustment intensity weight, which is used to guide subsequent dynamic adjustment operations. In specific operations, first, calculate the deviation percentage of the real-time hydrogen storage tank pressure and the safety pressure threshold, then divide the deviation into different levels according to the pre-set standard, for example, if the deviation is less than 5%, it is slight deviation, if the deviation is between 5% and 10%, it is moderate deviation, and if the deviation exceeds 10%, it is severe deviation. Each deviation level corresponds to a specific adjustment intensity weight. Based on the adjustment intensity weight, the electrolysis current and fuel cell power ratio are dynamically adjusted, for example, in the case of slight deviation, increase the electrolysis current to accelerate the hydrogen production rate and limit the fuel cell power to reduce hydrogen consumption. In the case of severe deviation, reduce the electrolysis current to slow down the hydrogen production rate, and reduce rather than increase the fuel cell power output to reduce hydrogen consumption. This hierarchical adjustment strategy ensures flexibility and robustness under different operating conditions, thereby maintaining the stable operation of the ship.

[0082] After dynamically adjusting the electrolysis current and fuel cell power ratio, the electrolysis waste heat temperature, coolant flow rate, hydrogen purity, and oxygen residual amount are monitored simultaneously to ensure the safe operation of the equipment. The electrolysis waste heat temperature is monitored in real time by the temperature sensor installed on the electrolysis equipment and transmitted to the control center. The control center determines whether the electrolysis equipment is within the safe operating range based on the electrolysis waste heat temperature data and starts the cooling pump as needed. The adjusted coolant flow rate is monitored in real time by the flowmeter and fed back to the control center. The hydrogen purity and oxygen residual amount are mainly obtained through gas sensors. The hydrogen purity sensor is used to detect the impurity content in the generated hydrogen in real time to ensure that the hydrogen purity meets the operating requirements of the fuel cell. The oxygen residual amount sensor is used to monitor the amount of oxygen that may be left over during the electrolysis process to prevent safety hazards caused by oxygen mixing with hydrogen. The monitoring data of electrolysis waste heat temperature, coolant flow rate, hydrogen purity, and oxygen residual amount not only helps to understand the working status of the electrolysis equipment in real time, but also supports other control decisions, such as adjusting the electrolysis current or fuel cell power output.

[0083] To further enhance safety, the hydrogen energy closed-loop monitoring unit defines multi-dimensional thresholds through hydrogen tank pressure, hydrogen purity, oxygen residual amount, and electrolysis waste heat temperature, and classifies them into three levels of safety protocols: level one warning, level two production limit, and level three emergency shutdown.

[0084] Based on the three-level safety protocol, by analyzing the safety range of hydrogen tank pressure, hydrogen purity, oxygen residual amount, and electrolysis waste heat temperature, the first threshold, the second threshold, and the third threshold are defined. When one of the parameters of hydrogen tank pressure, hydrogen purity, oxygen residual amount, and electrolysis waste heat temperature exceeds the first threshold, a level one warning is issued, and the operator is prompted to pay attention. At this time, some preventive measures are automatically taken, such as fine-tuning the electrolysis current or slightly reducing the fuel cell power output, to avoid the pressure continuing to rise.

[0085] If one of the parameters of hydrogen tank pressure, hydrogen purity, oxygen residual amount, and electrolysis waste heat temperature continues to deteriorate and reaches the second threshold, it enters the second level of production limit state. At this time, the power output of the fuel cell is limited, and the backup energy such as lithium battery pack is started to reduce the dependence on hydrogen. At the same time, auxiliary data such as electrolysis waste heat temperature and cooling liquid flow rate are output for the operator to refer to. For example, when the hydrogen tank pressure continues to rise and reaches the second threshold, the power output of the fuel cell is automatically reduced, and the lithium battery pack is started as a temporary power source to ensure basic operation without increasing hydrogen consumption. In addition, real-time monitoring of electrolysis waste heat temperature and cooling liquid flow rate ensures the normal operation of the equipment and prevents damage due to overheating or other problems.

[0086] When one of the parameters of hydrogen tank pressure, hydrogen purity, oxygen residual amount, and electrolysis waste heat temperature reaches the third threshold, the emergency shutdown protocol is triggered immediately. At this time, all high-risk equipment such as fuel cells and electrolysis equipment will be forced to shut down, and a safety status code will be output to ensure the overall safety of the ship. For example, when the hydrogen tank pressure rises sharply and exceeds the third threshold, all related equipment is immediately stopped, and a safety status code is output to inform the operator to take further safety measures. The emergency shutdown mechanism is the last line of defense to quickly cut off the source of danger in extreme cases and protect the safety of the ship and personnel.

[0087] Through the above operations, comprehensive monitoring of hydrogen tank pressure, hydrogen purity, oxygen residual amount, and electrolysis waste heat temperature is achieved, ensuring efficient operation and safety under various working conditions and avoiding potential safety hazards.

[0088] The waste heat recovery and self-checking module is based on the parameters of electrolysis waste heat temperature and cooling liquid flow rate. The waste heat recovery converter increases the coolant temperature and reduces the cooling energy consumption through phase change materials, and the quantum chip performs equipment health self-checking.

[0089] The specific steps include the following:

[0090] To achieve efficient waste heat recovery, based on the parameters of electrolytic waste heat temperature and cooling liquid flow, a phase change material thermal behavior simulation model is constructed using finite element analysis (FEA), which is a numerical simulation method widely used in engineering, especially in heat conduction and fluid mechanics. The parameters of electrolytic waste heat temperature and cooling liquid flow are input into the finite element analysis tool to construct a detailed phase change material thermal behavior simulation model. This model takes into account various thermal physical parameters of the phase change material, such as melting point, latent heat, etc., as well as the dynamic characteristics of the coolant flow, ensuring accurate prediction of the performance of the phase change material under different working conditions. Through the phase change material thermal behavior simulation model, the heat transfer during the cooling process can be accurately simulated, providing a basis for subsequent optimization;

[0091] Based on the phase change material thermal behavior simulation model, a virtual PID controller is used to simulate different operating conditions by adjusting the virtual coolant flow parameter, and the optimal flow control strategy is found. This not only improves the temperature of the coolant and reduces external energy consumption, but also ensures the efficient operation of the ship. According to the generated optimal flow control strategy, the cooling pump adjusts the actual coolant temperature, for example, when the electrolytic waste heat temperature rises, the cooling liquid flow is automatically increased to carry away more heat, and when the temperature returns to normal, the cooling liquid flow is reduced to save energy. This intelligent dynamic adjustment mechanism not only improves the energy efficiency of the cooling process, but also prolongs the service life of the equipment;

[0092] At the same time, the quantum chip performs a device health self-check function. First, it uses quantum annealing algorithm to analyze the power parameters of hydrogen energy ships, efficiently finding the global optimal solution to achieve the best configuration of ship power parameters, ensuring that the equipment can operate in the best state. In addition, deep neural network (DNN) is used to predict potential failure modes. Deep neural network is a powerful machine learning model that can learn complex patterns and rules from a large number of historical power parameters of hydrogen energy ships, and predict possible failure modes such as overheating and pressure abnormalities based on real-time collected power parameters of hydrogen energy ships. To ensure the accuracy of the deep neural network model, a large number of historical power parameters of hydrogen energy ships are used for training, including normal operating state and various fault conditions. The predicted results of the failure modes provide important reference for subsequent maintenance and repair, such as when the deep neural network model predicts that a key component is about to fail, the control center will issue an early warning and suggest appropriate preventive measures;

[0093] To further improve the accuracy and efficiency of the equipment health self-check, reinforcement learning technology is adopted to dynamically adjust the detection strategy. Reinforcement learning is a machine learning method that learns the best action strategy through trial and error. By designing a reinforcement learning framework, the detection strategy is continuously optimized based on real-time feedback from the equipment health self-check. For example, if an abnormal reading is found from a sensor during the detection process, the detection frequency or range is automatically increased to more comprehensively assess the equipment status. In this way, potential problems can be quickly responded to and timely measures can be taken to avoid downtime or accidents caused by failures.

[0094] Through the above operations, the cooling process is optimized, which can reduce external energy consumption while improving overall energy efficiency. At the same time, the equipment health self-check operation can timely discover and solve potential problems, avoiding downtime or accidents caused by failures.

[0095] The maintenance feedback module generates a maintenance recommendation report based on the self-check results and uploads it to the shore-based data center.

[0096] Specifically, the following steps are included,

[0097] When the quantum chip completes the equipment health self-check, a detailed maintenance recommendation report is generated based on the self-check results. First, all relevant sensor data is collected and analyzed, including electrolytic waste heat temperature, cooling fluid flow rate, hydrogen purity, and oxygen residual amount. Sensor data is the basis for evaluating the health of the equipment, and any abnormal value may indicate a potential problem. For example, if the pressure value detected by a sensor consistently exceeds the normal range, the control hub will record this abnormal situation and include it in the maintenance recommendation report.

[0098] Potential failure modes predicted by the deep neural network are specifically noted in the maintenance recommendation report, and appropriate preventive measures are proposed. In addition, information from the expert knowledge base is combined, which contains a wealth of best practices and common problem solutions related to equipment maintenance. The control hub automatically matches the self-check results with the entries in the knowledge base to extract relevant maintenance recommendations. For example, if a sensor reading is abnormal, the control hub will search the expert knowledge base for a similar situation and add the handling method to the maintenance recommendation report. This combination of machine learning and expert knowledge makes the maintenance recommendation report both scientifically sound and practically operable.

[0099] After generating the maintenance recommendation report, it is first encrypted and transmitted to the shore-based data center using a secure and reliable communication protocol such as HTTPS or MQTT. This approach not only ensures security but also effectively prevents data in the maintenance recommendation report from being tampered with or lost during transmission. The shore-based data center receives and saves the maintenance recommendation reports from the ship, facilitating subsequent queries and analysis.

[0100] After receiving the maintenance recommendation report, the shore-based data center will further process and analyze it to find potential trends and patterns. For example, by comparing data from multiple ships' historical maintenance recommendation reports, the average service life of certain components can be found, providing a reference for future maintenance planning. In addition, the shore-based data center can use big data analysis tools such as Hadoop or Spark to quickly process and mine massive amounts of data to find valuable information hidden behind the data in the maintenance recommendation report. For example, by analyzing equipment failure rates in different sea areas and seasons, the team can develop more reasonable maintenance strategies to improve overall operational efficiency.

[0101] Through the above operations, the maintenance recommendation report is generated, significantly improving the reliability and operation efficiency of the ship's power propulsion. Through analysis, the shore-based data center identifies potential trends and patterns, which not only reduces equipment failure rates and downtime, but also improves overall operational efficiency.

[0102] In summary, the present application can generate accurate control instructions by real-time acquisition and processing of power parameters of hydrogen energy ships, and using advanced Kalman filtering algorithm and quantum annealing optimization technology. The output power of hydrogen fuel cells and lithium batteries can be dynamically adjusted according to real-time sailing conditions, significantly improving the overall energy efficiency of the ship. By using an efficient hydrogen energy closed-loop monitoring unit and a three-level safety protocol, the electrolysis current and fuel cell power ratio are dynamically adjusted based on real-time hydrogen storage tank pressure data, and different adjustment intensity weights are matched according to the deviation level, which not only improves response speed and accuracy, but also significantly enhances stability and safety, ensuring reliable operation of the ship under complex working conditions and reducing potential safety risks.

[0103] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and not to limit it. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, which should be covered by the scope of the claims of the present application.

Claims

1. A range extended hydrogen fuel powered marine vessel propulsion system characterized by: The application relates to a hydrogen energy ship power parameter optimization system based on quantum annealing optimization. The acquisition module eliminates noise interference of the collected hydrogen energy ship power parameters through a Kalman filtering algorithm to generate a standardized data package; The quantum annealing optimization module receives the standardized data package, maps the hydrogen energy ship power parameters into an Ising model, and generates control instructions according to the hydrogen fuel consumption rate, the equipment life attenuation coefficient and the sailing task priority weight. The proton membrane regulation module inputs the control instructions into a simulation model of a proton exchange membrane, dynamically adjusts the hydrogen production rate and the fuel cell power, and outputs real-time hydrogen storage tank pressure data. The safety monitoring module dynamically adjusts the electrolysis current and the fuel cell power ratio based on the real-time hydrogen storage tank pressure data, synchronously monitors the hydrogen purity and the oxygen residual amount, triggers a three-level safety protocol, and outputs the electrolysis waste heat temperature, the cooling liquid flow and the safety state code. The waste heat recovery and self-checking module improves the coolant temperature and reduces the cooling energy consumption through a phase change material based on the parameters of the electrolysis waste heat temperature and the cooling liquid flow, and the quantum chip performs the equipment health self-checking. The maintenance feedback module generates a maintenance suggestion report based on the self-checking result and uploads the report to a shore-based data center.

2. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 1, characterized by: The hydrogen energy ship power parameters include the residual capacity of a lithium battery pack, the internal pressure of a hydrogen storage tank, the real-time power demand of a propeller, the temperature of a fuel cell stack, the seawater flow rate, the draft of the ship, the hydrogen fuel consumption rate, the equipment life attenuation coefficient and the sailing task priority weight.

3. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 1, characterized by: The standardized data package is obtained through the following steps, The collected hydrogen energy ship power parameters are denoised through Kalman filtering, and after the abnormal values are removed by combining an isolated forest algorithm, the Min-Max normalization and One-Hot encoding are used to generate the standardized data package.

4. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 1, characterized by: The hydrogen energy ship power parameters are mapped into the Ising model through the following steps, The hydrogen fuel consumption rate is encoded as the antiferromagnetic coupling strength, the equipment life attenuation coefficient is mapped as the local magnetic field bias, and the sailing task priority is converted into the constraint energy term; Based on the antiferromagnetic coupling strength, the local magnetic field bias and the constraint energy term, a multi-objective energy function of the Ising model is constructed in the quantum simulation algorithm, and the hydrogen energy ship power parameters are mapped into the Ising model.

5. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 4, wherein: The control instructions are obtained through the following steps, After the hydrogen energy ship power parameters are mapped into the Ising model, the antiferromagnetic coupling matrix, the local magnetic field vector and the constraint energy term are dynamically synthesized through the matrix operation framework of TensorFlow, and a multi-objective optimization Hamiltonian model is constructed by combining an adaptive weight distribution algorithm; Based on the multi-objective optimization Hamiltonian model, the spin state parameters are obtained by iteratively solving the model in the TensorFlow calculation framework through the simulated annealing algorithm and by using distributed calculation; According to the spin state parameters, the control instructions of the fuel cell power and the hydrogen production rate are dynamically generated.

6. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 1, characterized by: The real-time hydrogen storage tank pressure data are obtained through the following steps, Based on the control instructions, the hydrogen production rate set value and the fuel cell power output are extracted and dynamically loaded into the simulation model of the proton exchange membrane, and the hydrogen generation and consumption rate difference is synchronously calculated through the mass conservation equation; Based on the difference between hydrogen generation and consumption rate, through PID control algorithm and fluid mechanics algorithm, the hydrogen production rate set value is corrected in real time, and the fuel cell power output is dynamically adjusted combined with the adaptive feedback adjustment mechanism of equivalent circuit; Based on the corrected hydrogen production rate set value and the adjusted fuel cell power output, the hydrogen storage tank pressure change curve is solved by Runge-Kutta numerical integration, and real-time hydrogen storage tank pressure data is obtained.

7. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 6, characterized by: The hydrogen energy closed-loop monitoring unit dynamically adjusts the electrolysis current and fuel cell power ratio, and the steps are as follows, Based on the real-time hydrogen storage tank pressure data, the pressure trend is predicted by Kalman filtering combined with long short-term memory network to generate a safety pressure threshold; The hydrogen energy closed-loop monitoring unit calculates the deviation percentage of real-time hydrogen storage tank pressure and safety pressure threshold, and divides the deviation level, matches different adjustment intensity weights according to the deviation level, and dynamically adjusts the electrolysis current and fuel cell power ratio.

8. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 1, characterized by: The three-level safety protocol specifically includes, Based on the hydrogen storage tank pressure, hydrogen purity, oxygen residual amount and electrolysis waste heat temperature, multi-dimensional threshold values are defined; The multi-dimensional threshold values are classified into first-level warning, second-level production limit and third-level emergency shutdown.

9. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 1, characterized by: The waste heat recovery converter improves the coolant temperature and reduces the cooling energy consumption through the phase change material, and the steps are as follows, Based on the real-time data of electrolysis waste heat temperature and cooling liquid flow, a phase change material thermal behavior simulation model is constructed by finite element analysis; Based on the phase change material thermal behavior simulation model, a virtual PID controller is used to adjust the virtual coolant flow parameter and generate an optimal flow control strategy, and at the same time, the optimal flow control strategy is used to guide the cooling pump to adjust the actual coolant temperature.

10. The extended-range hydrogen fuel-powered marine vessel propulsion system as defined in claim 1, characterized by: The steps of the quantum chip executing equipment health self-checking are as follows, The quantum chip uses quantum annealing algorithm to analyze hydrogen energy ship power parameters, and predicts fault modes through deep neural network, and dynamically adjusts detection strategy based on reinforcement learning, to realize equipment health self-checking.

Images