Marine distributed green ammonia production scheduling system and method based on intelligent management

Abstract

The application provides a smart management-based offshore distributed green ammonia production scheduling system and method, and relates to the technical field of offshore renewable energy utilization. The system comprises a distributed ammonia production platform for distributed production of offshore green ammonia, a renewable energy power generation device, water electrolysis hydrogen production equipment, an ammonia synthesis reactor, a green ammonia energy storage device, and an ammonia production platform sensor network; a multifunctional ship node for green ammonia production, transportation and consumption, a green ammonia transportation carrier, an offshore production node expansion unit, a green ammonia fuel consumption terminal, and a ship state sensor; and an intelligent green ammonia management platform integrating a data transmission module, a multi-objective optimization scheduling module, a fault prediction and emergency response module, and a supply chain management module, which is used for communication with the ammonia production platform and the ship node through Internet of Things technology, so as to realize intelligent closed-loop management of green ammonia production, transportation and consumption, improve production scheduling efficiency and energy utilization rate, and reduce transportation costs.

Description

Technical Field

[0001] This invention relates to the field of marine renewable energy utilization technology, and in particular to a marine distributed green ammonia production scheduling system and method based on intelligent management. Background Technology

[0002] As the global goal of carbon neutrality is advanced, conversion technologies for renewable energy sources such as offshore wind power and green ammonia are receiving increasing attention.

[0003] Existing offshore distributed green ammonia production systems generally rely on manual scheduling or semi-automated control, which cannot respond in real time to dynamic changes such as wind power fluctuations and equipment failures, potentially leading to insufficient energy utilization and low production efficiency. Furthermore, in terms of production allocation mechanisms and supply chain integration, this approach lacks intelligent collaborative management of offshore distributed nodes. In addition, traditional land-based green ammonia production requires long-distance transportation, resulting in high transportation costs and low scheduling matching, easily leading to supply and demand imbalances and failing to form an effective closed-loop control system encompassing production, transportation, and consumption.

[0004] Therefore, it is necessary to provide a marine distributed green ammonia production scheduling system and method based on intelligent management to solve the above-mentioned technical problems. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a marine distributed green ammonia production scheduling system and method based on intelligent management. This system solves the problems that traditional manual or semi-automatic scheduling modes make it difficult for operators to respond quickly to emergencies, reducing the stability and security of the system. Furthermore, existing technologies lack effective intelligent decision support, making it difficult to achieve optimal resource allocation and affecting the overall economic benefits of green ammonia production.

[0006] The present invention provides a marine distributed green ammonia production scheduling system based on intelligent management, the production scheduling system comprising:

[0007] The distributed ammonia production platform is used for the distributed production of green ammonia at sea. It is equipped with renewable energy power generation equipment, water electrolysis hydrogen production equipment, ammonia synthesis reactor, green ammonia energy storage device, and a sensor network for the ammonia production platform.

[0008] A multi-functional ship node is used for the production, transportation and consumption of green ammonia at sea. It is equipped with a green ammonia transport carrier, an offshore production node expansion unit, a green ammonia fuel consumption terminal and ship status sensors.

[0009] The intelligent green ammonia management platform integrates a data transmission module, a multi-objective optimization scheduling module, a fault prediction and emergency response module, and a supply chain management module. It is used to communicate with the distributed ammonia production platform and the multi-functional ship node through Internet of Things technology.

[0010] The data transmission module is used to acquire real-time green ammonia production data of the distributed ammonia production platform and ship transportation status data of the multi-functional ship node, and encrypt and transmit the real-time green ammonia production data and ship transportation status data to the multi-objective optimization scheduling module, the fault prediction and emergency response module, and the supply chain management module based on a blockchain network. The multi-objective optimization scheduling module is used to update the dynamic production scheduling strategy through a dynamic allocation algorithm. The fault prediction and emergency response module is used to construct a comprehensive fault prediction model based on an LSTM neural network to predict anomalies in the distributed ammonia production platform and trigger a comprehensive emergency response mechanism in advance. The supply chain management module is used to generate target replenishment instructions using an inventory optimization algorithm, adjust the dynamic production scheduling strategy in real time based on the target replenishment instructions, and send them to the distributed ammonia production platform and the multi-functional ship node.

[0011] Preferably, the ammonia production platform sensor network collects energy supply data of the renewable energy power generation device, hydrogen synthesis data of the water electrolysis hydrogen production device, ammonia synthesis data of the ammonia synthesis reactor, and green ammonia storage data of the green ammonia energy storage device at preset collection time intervals, and summarizes the energy supply data, hydrogen synthesis data, ammonia synthesis data, and green ammonia storage data to obtain the real-time green ammonia production data;

[0012] The ship status sensor is used to collect the ship transportation status data of the multifunctional ship node;

[0013] Based on Internet of Things (IoT) technology, the collected real-time green ammonia production data and ship transportation status data are uploaded to the data acquisition module.

[0014] Preferably, after receiving the real-time green ammonia production data and the ship transportation status data, the multi-objective optimization scheduling module extracts the renewable energy power generation from the energy supply data; the hydrogen production and consumption power, hydrogen synthesis power, and maximum hydrogen synthesis capacity from the hydrogen synthesis data; the ammonia synthesis and consumption power, ammonia synthesis power, and maximum ammonia synthesis capacity from the ammonia synthesis data; the liquid ammonia inventory from the green ammonia storage data; and the total number of transport vessels, ship transport distance, ship load factor, actual ammonia load, and maximum ship load capacity from the ship transportation status data. It then calculates the sum of the hydrogen production and consumption power and the ammonia synthesis and consumption power to obtain the total consumption power, and acquires the preset scheduling cycle and the actual total green ammonia demand.

[0015] A dynamic production scheduling model is constructed based on the aforementioned dynamic allocation algorithm. This model includes a multi-objective optimization scheduling function and corresponding constraints. The multi-objective optimization scheduling function aims to minimize energy utilization, transportation costs, and inventory fluctuations. The constraints include equipment capacity limits, ship load limits, and market demand constraints. The corresponding calculation formulas are as follows:

[0016] In the formula, Indicates energy utilization rate; Indicates transportation costs; Indicates inventory fluctuation; This indicates the total amount of electricity consumed; M represents the amount of electricity generated from renewable energy sources; M represents the total number of transport vessels. This represents the shipping distance of the i-th ship. represents the load factor of the i-th ship; N represents the total number of distributed ammonia production platforms; This represents the liquid ammonia inventory of the j-th distributed ammonia production platform; This represents the average liquid ammonia inventory across all distributed ammonia production platforms; max indicates the maximum value operation; min indicates the minimum value operation. This represents the hydrogen synthesis power of the j-th distributed ammonia production platform; This represents the maximum hydrogen synthesis capacity of the j-th distributed ammonia production platform; This represents the ammonia synthesis power of the j-th distributed ammonia production platform; This represents the maximum ammonia synthesis capacity of the j-th distributed ammonia production platform; This represents the actual ammonia load of the i-th ship; This represents the maximum deadweight of the i-th vessel; Indicates the preset scheduling period; This indicates the total actual demand for green ammonia.

[0017] Preferably, an improved genetic algorithm is used to solve the dynamic production scheduling model at a preset update time interval, and the dynamic production scheduling strategy is updated based on the solution results, specifically including the ammonia production plan of the distributed ammonia production platform and the ship transportation path of the multifunctional ship node.

[0018] Preferably, the fault prediction and emergency response module is used to construct a comprehensive fault prediction model based on an LSTM neural network, to predict anomalies in the distributed ammonia production platform, and to trigger a comprehensive emergency response mechanism in advance, specifically including:

[0019] Feature extraction is performed on the real-time green ammonia production data to obtain the corresponding real-time green ammonia production features;

[0020] The real-time green ammonia production characteristics are input into the forget gate of the comprehensive fault prediction model to determine the real-time green ammonia production characteristics that need to be discarded from the memory unit. The corresponding calculation formula is as follows:

[0021] In the formula, The output of the forget gate at the current time t is used to determine the real-time green ammonia production features that need to be discarded from the memory unit; Indicates the activation function; Indicates the weight of the forget gate; This represents the hidden state at the previous time step t-1; This represents the real-time green ammonia production characteristics input at the current time t; Indicates hidden state and real-time green ammonia production characteristics The vector formed; Indicates the bias of the forget gate;

[0022] Based on the input gate of the comprehensive fault prediction model, the real-time green ammonia production characteristics that need to be stored in the memory unit are determined, and the corresponding calculation formula is as follows:

[0023] In the formula, The output of the input gate at the current time t is used to determine the real-time green ammonia production characteristics that need to be stored in the memory unit; Indicates the weights of the input gates; Indicates the bias of the input gate;

[0024] Candidate green ammonia production characteristics are determined based on the candidate memory units of the comprehensive fault prediction model, and the corresponding calculation formula is as follows:

[0025] In the formula, This represents the candidate memory unit at the current time t, used to determine the candidate green ammonia production characteristics; Represents the hyperbolic tangent function; Indicates the weight of the candidate memory unit; Indicates the bias of candidate memory units;

[0026] The calculation formula for updating the memory unit is as follows:

[0027] In the formula, This represents the memory unit at the current time t; This represents the memory unit from the previous time step t-1;

[0028] Based on the output gate of the comprehensive fault prediction model, the candidate green ammonia production characteristics that need to be output to the hidden state are determined, and the corresponding calculation formula is as follows:

[0029] In the formula, This represents the output of the output gate at the current time t, used to determine the candidate green ammonia production features that need to be output to the hidden state. Indicates the weight of the output gate; Indicates the bias of the output gate;

[0030] The hidden state at the current moment is determined based on the memory unit and the output of the output gate, and the corresponding calculation formula is as follows:

[0031] In the formula, This represents the hidden state at the current time t.

[0032] Preferably, the hidden state at the current moment is mapped to the comprehensive fault probability through a fully connected layer, and the corresponding calculation formula is as follows:

[0033] In the formula, This represents the overall failure probability at the current time t. ; Indicates the weights of the fully connected layer; Indicates the bias of the fully connected layer;

[0034] Obtain a critical fault threshold, and when the overall fault probability is greater than or equal to the critical fault threshold, trigger the overall emergency response mechanism in advance.

[0035] Preferably, the supply chain management module uses the inventory optimization algorithm to generate the target replenishment instruction, specifically including:

[0036] The average demand rate, standard deviation of the demand rate, and safety factor for demand fluctuations are obtained. The round-trip shipping time from the ship transportation status data is extracted, and the inventory optimization algorithm is used to calculate the inventory level triggering replenishment as follows:

[0037] In the formula, CBK represents the inventory level that triggers replenishment; WFSJ represents the average demand rate; WFSJ represents the round-trip shipping time. Indicates the safety factor for demand fluctuations; Indicates the standard deviation of the demand rate;

[0038] The corresponding target replenishment instruction is generated based on the triggered replenishment inventory level.

[0039] A method for scheduling distributed offshore green ammonia production based on intelligent management, the production scheduling method comprising:

[0040] The real-time green ammonia production data from the distributed ammonia production platform and the ship transportation status data from the multi-functional ship node are acquired and uploaded to the data transmission module.

[0041] The real-time green ammonia production data and the ship transportation status data are transmitted through the data transmission module to the multi-objective optimization scheduling module, the fault prediction and emergency response module, and the supply chain management module.

[0042] Based on the multi-objective optimization scheduling module, the dynamic allocation algorithm is used to update the dynamic production scheduling strategy;

[0043] Based on the fault prediction and emergency response module, a comprehensive fault prediction model based on LSTM neural network is constructed to predict anomalies in the distributed ammonia production platform and trigger the comprehensive emergency response mechanism in advance.

[0044] Based on the supply chain management module, the target replenishment instruction is generated using the inventory optimization algorithm. The dynamic production scheduling strategy is adjusted in real time based on the target replenishment instruction and sent to the distributed ammonia production platform and the multi-functional ship node. The intelligent green ammonia management platform includes the data transmission module, the multi-objective optimization scheduling module, the fault prediction and emergency response module, and the supply chain management module.

[0045] Compared with related technologies, the intelligent management-based offshore distributed green ammonia production scheduling system and method provided by this invention have the following beneficial effects:

[0046] This invention can acquire real-time green ammonia production data from a distributed ammonia production platform and ship transportation status data from multi-functional ship nodes, and upload them to a data transmission module. The data transmission module then transmits the real-time green ammonia production data and ship transportation status data to a multi-objective optimization scheduling module, a fault prediction and emergency response module, and a supply chain management module. Based on the multi-objective optimization scheduling module, a dynamic allocation algorithm is used to update the dynamic production scheduling strategy. Based on the fault prediction and emergency response module, a comprehensive fault prediction model based on an LSTM neural network is constructed to predict anomalies in the distributed ammonia production platform and trigger a comprehensive emergency response mechanism in advance. Based on the supply chain management module, an inventory optimization algorithm is used to generate target replenishment instructions, and the dynamic production scheduling strategy is adjusted in real time based on these instructions and sent to the distributed ammonia production platform and multi-functional ship nodes. The intelligent green ammonia management platform includes a data transmission module, a multi-objective optimization scheduling module, a fault prediction and emergency response module, and a supply chain management module, thereby achieving intelligent closed-loop management of the entire process of green ammonia production, transportation, and consumption, significantly improving production scheduling efficiency and energy utilization, and reducing transportation costs.

[0047] This invention improves renewable energy utilization and significantly reduces wind and electricity curtailment losses through a multi-objective optimization scheduling algorithm and real-time data-driven approach. Furthermore, by dynamically coordinating a distributed ammonia production platform with ship nodes, it shortens the green ammonia production cycle, reduces unit capacity costs, and increases inventory turnover, effectively alleviating resource waste caused by supply-demand imbalances in traditional models. This invention accurately predicts abnormal situations on the ammonia production platform using a comprehensive fault prediction model based on LSTM neural networks, combined with a minute-level emergency response mechanism, reducing system downtime and significantly mitigating the risk of production interruptions due to unforeseen events. It further enhances the system's robustness in extreme weather or equipment failure scenarios, ensuring the continuity and safety of offshore operations. The multi-functional ship nodes of this invention overcome the shortcomings of traditional long-distance transportation modes in land-based production through a three-in-one design integrating production, transportation, and consumption, reducing transportation costs. This invention achieves full-chain traceability and dynamic balance of green ammonia from production to consumption through blockchain smart contracts and inventory optimization algorithms, forming a closed-loop offshore energy network and significantly reducing the environmental impact throughout the entire green ammonia production lifecycle. Attached Figure Description

[0048] Figure 1 This is a system block diagram of the intelligent management-based offshore distributed green ammonia production scheduling system of the present invention;

[0049] Figure 2 This is a flowchart of the intelligent management-based offshore distributed green ammonia production scheduling method of the present invention. Detailed Implementation

[0050] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0051] Example 1

[0052] like Figure 1 As shown, a marine distributed green ammonia production scheduling system based on intelligent management includes:

[0053] The distributed ammonia production platform is used for the distributed production of green ammonia at sea. It is equipped with renewable energy power generation equipment, water electrolysis hydrogen production equipment, ammonia synthesis reactor, green ammonia energy storage device, and a sensor network for the ammonia production platform.

[0054] Among them, the distributed ammonia production platform is the basic production node of the system. It is deployed in areas rich in offshore wind power or photovoltaic resources and realizes on-site production and energy conversion of green ammonia through modular design.

[0055] Specifically, the renewable energy power generation device uses offshore wind turbines or photovoltaic arrays to convert renewable energy sources such as wind and solar power into electricity, providing zero-carbon power for hydrogen production through water electrolysis. This device features maximum power point tracking (MPPT) capability, dynamically adjusting power generation efficiency based on environmental parameters to ensure maximum energy capture. The water electrolysis hydrogen production equipment is typically equipped with a proton exchange membrane electrolyzer, which uses renewable electricity to decompose water into high-purity hydrogen and oxygen. The proton exchange membrane electrolyzer is characterized by fast response and high energy efficiency, adapting to the fluctuating characteristics of offshore wind power. The ammonia synthesis reactor is based on the Haber-Bosch process, synthesizing ammonia from hydrogen and nitrogen under high temperature (300-500℃), high pressure (15-30 MPa), and catalyst (such as an iron-based catalyst), with the nitrogen typically derived from air separation. The green ammonia energy storage device uses low-temperature, atmospheric-pressure storage tanks or pressure vessels to store liquid ammonia. The energy storage device is equipped with a thermal insulation system to reduce evaporation losses and support rapid loading and unloading operations at ship nodes. The ammonia production platform's sensor network consists of various types of sensors, such as anemometers, photovoltaic irradiance meters, pressure transmitters, and level gauges, which collect real-time energy supply status, equipment operating parameters, and inventory data through an IoT gateway.

[0056] In practical applications, distributed ammonia production platforms provide power to water electrolysis hydrogen production equipment via renewable energy power generation devices, enabling zero-carbon energy supply for offshore green ammonia production. This reduces reliance on terrestrial fossil fuels and lowers the environmental impact of the production process at its source. Through the coordinated operation of the ammonia synthesis reactor and the green ammonia energy storage device, production capacity can be dynamically adjusted based on fluctuations in renewable energy availability, avoiding resource waste caused by wind and electricity curtailment and improving energy utilization stability. The ammonia production platform's sensor network can collect real-time data on equipment status, energy supply, and inventory, providing high-precision decision-making support for the intelligent management platform and significantly enhancing the adaptability of the production scheduling system to the complex offshore environment.

[0057] A multi-functional ship node is used for the production, transportation and consumption of green ammonia at sea. It is equipped with a green ammonia transport carrier, an offshore production node expansion unit, a green ammonia fuel consumption terminal and ship status sensors.

[0058] The multi-functional ship node overcomes the limitations of traditional ships' single-transport capabilities, enabling the production, transportation, and consumption of green ammonia at sea. The green ammonia transport vehicle is equipped with cryogenic ammonia storage tanks and an intelligent loading and unloading system. It can transport ammonia according to the route instructions of the intelligent management platform. During transportation, the ship's position, load, and liquid ammonia status can be monitored in real time using ship status sensors such as GPS, inertial navigation systems, and cargo hold monitoring sensors. The offshore production node expansion unit can be equipped with modular ammonia production equipment, enabling on-site auxiliary production during periods of wind power surplus to temporarily increase system capacity. This unit supports plug-and-play operation and can be linked with the ship's energy system via a quick interface. The green ammonia fuel consumption terminal can provide the ship with appropriate electrical power. The ship status sensors can monitor and collect the transportation status data of the multi-functional ship node in real time.

[0059] It should be noted that the integrated design of the green ammonia transport carrier and the offshore production node expansion unit can achieve dynamic optimal planning of the green ammonia transport route through intelligent scheduling, reducing ineffective navigation and energy consumption; at the same time, it can utilize excess offshore wind power to achieve temporary production increases and improve the overall capacity utilization rate of the system.

[0060] Furthermore, green ammonia fuel consumption terminals can directly convert green ammonia into ship power, creating a closed-loop mechanism of self-production and self-consumption, eliminating the long-distance carbon footprint and scheduling delays caused by land transportation. Simultaneously, through ship status sensors, transportation and consumption data can be collected and fed back in real time, providing real-time data support for the dynamic optimization of the supply chain.

[0061] The intelligent green ammonia management platform integrates a data transmission module, a multi-objective optimization scheduling module, a fault prediction and emergency response module, and a supply chain management module. It is used to communicate with the distributed ammonia production platform and the multi-functional ship node through Internet of Things technology.

[0062] The data transmission module is used to acquire real-time green ammonia production data of the distributed ammonia production platform and ship transportation status data of the multi-functional ship node, and encrypt and transmit the real-time green ammonia production data and ship transportation status data to the multi-objective optimization scheduling module, the fault prediction and emergency response module, and the supply chain management module based on a blockchain network. The multi-objective optimization scheduling module is used to update the dynamic production scheduling strategy through a dynamic allocation algorithm. The fault prediction and emergency response module is used to construct a comprehensive fault prediction model based on an LSTM neural network to predict anomalies in the distributed ammonia production platform and trigger a comprehensive emergency response mechanism in advance. The supply chain management module is used to generate target replenishment instructions using an inventory optimization algorithm, adjust the dynamic production scheduling strategy in real time based on the target replenishment instructions, and send them to the distributed ammonia production platform and the multi-functional ship node.

[0063] In the intelligent green ammonia management platform, the data transmission module can acquire production data from the distributed ammonia production platform and transportation data from ship nodes in real time through IoT protocols. It also uses consortium blockchain technology to hash and encrypt this data and distribute it to ensure the immutability and security of the data. This solves the information leakage risk that may exist in traditional IoT transmission. Furthermore, it can preprocess real-time data through edge computing nodes, such as outlier filtering and normalization, to improve data quality.

[0064] The multi-objective optimization scheduling module adopts a dynamic allocation algorithm, which integrates multiple objectives such as energy utilization rate, transportation cost, and inventory balance. It achieves global collaborative optimization of distributed nodes through an improved genetic algorithm, and updates the production scheduling strategy in real time. It generates production plans and ship transportation routes for each ammonia production platform, realizing optimal resource allocation under multi-variable constraints. This avoids the lag and locality defects of manual or semi-automatic scheduling, and improves resource allocation efficiency and production flexibility.

[0065] The fault prediction and emergency response module is based on a long short-term memory network (LSTM) to build a fault prediction model for the equipment, enabling the prediction of potential faults in the ammonia production platform. When an anomaly is detected, it automatically triggers a comprehensive emergency response mechanism, such as switching to a backup node and adjusting the transportation route. This can shorten the anomaly handling time, reduce the risk of equipment damage and production interruption losses, ensure the continuous operation of the production scheduling system, and enhance the system's robustness and security.

[0066] The supply chain management module uses an inventory optimization algorithm that combines real-time demand and transportation status to calculate the optimal replenishment quantity and reorder point, generate corresponding replenishment instructions, and dynamically adjust production plans and ship scheduling. This can reduce the risk of inventory backlog and shortages, improve the response speed and operational efficiency of the entire chain, and at the same time, it can trace the flow of green ammonia through smart contract technology to ensure the transparency and reliability of the supply chain.

[0067] In the specific implementation process, the ammonia production platform sensor network collects energy supply data of the renewable energy power generation device, hydrogen synthesis data of the water electrolysis hydrogen production device, ammonia synthesis data of the ammonia synthesis reactor, and green ammonia storage data of the green ammonia energy storage device according to a preset collection time interval. The energy supply data, hydrogen synthesis data, ammonia synthesis data, and green ammonia storage data are then summarized to obtain the real-time green ammonia production data.

[0068] The ship status sensor is used to collect the ship transportation status data of the multifunctional ship node;

[0069] Based on Internet of Things (IoT) technology, the collected real-time green ammonia production data and ship transportation status data are uploaded to the data acquisition module.

[0070] Among them, the ammonia production platform sensor network collects energy supply, hydrogen synthesis, ammonia synthesis and inventory data at preset intervals, such as minutes, to achieve digital mapping of all elements of the production process.

[0071] For example, real-time power data from renewable energy power generation devices can directly reflect fluctuations in wind and solar energy, helping the system dynamically adjust the workload of hydrogen production and ammonia synthesis, avoiding equipment start-up and shutdown losses or capacity waste caused by unstable energy supply. Refined data collection on hydrogen and ammonia synthesis, such as yield and reaction efficiency, provides real-time capacity boundary conditions for multi-objective optimization scheduling algorithms, ensuring that production plans match actual equipment capacity and improving the rationality of resource utilization.

[0072] Ship status sensors can capture transportation status in real time, such as location, load, and speed. Combined with inventory data from the ammonia production platform, this enables the intelligent management platform to calculate transportation demand and route feasibility in real time.

[0073] For example, when a platform's inventory approaches a safety threshold, the system can quickly dispatch nearby ships to perform transportation tasks based on ship location data, shortening the decision delay of traditional manual dispatching, improving the system's response speed to environmental changes, and significantly enhancing the synergy between production and transportation.

[0074] After receiving the real-time green ammonia production data and the ship transportation status data, the multi-objective optimization scheduling module extracts the renewable energy power generation from the energy supply data; the hydrogen production and consumption power, hydrogen synthesis power, and maximum hydrogen synthesis capacity from the hydrogen synthesis data; the ammonia synthesis and consumption power, ammonia synthesis power, and maximum ammonia synthesis capacity from the ammonia synthesis data; the liquid ammonia inventory from the green ammonia storage data; and the total number of transport vessels, ship transport distance, ship load factor, actual ammonia load, and maximum ship load capacity from the ship transportation status data. It then calculates the sum of the hydrogen production and consumption power and the ammonia synthesis and consumption power to obtain the total consumption power, and acquires the preset scheduling cycle and the actual total green ammonia demand.

[0075] A dynamic production scheduling model is constructed based on the aforementioned dynamic allocation algorithm. This model includes a multi-objective optimization scheduling function and corresponding constraints. The multi-objective optimization scheduling function aims to minimize energy utilization, transportation costs, and inventory fluctuations. The constraints include equipment capacity limits, ship load limits, and market demand constraints. The corresponding calculation formulas are as follows:

[0076] In the formula, Indicates energy utilization rate; Indicates transportation costs; Indicates inventory fluctuation; This indicates the total amount of electricity consumed; M represents the amount of electricity generated from renewable energy sources; M represents the total number of transport vessels. This represents the shipping distance of the i-th ship. represents the load factor of the i-th ship; N represents the total number of distributed ammonia production platforms; This represents the liquid ammonia inventory of the j-th distributed ammonia production platform; This represents the average liquid ammonia inventory across all distributed ammonia production platforms; max indicates the maximum value operation; min indicates the minimum value operation. This represents the hydrogen synthesis power of the j-th distributed ammonia production platform; This represents the maximum hydrogen synthesis capacity of the j-th distributed ammonia production platform; This represents the ammonia synthesis power of the j-th distributed ammonia production platform; This represents the maximum ammonia synthesis capacity of the j-th distributed ammonia production platform; This represents the actual ammonia load of the i-th ship; This represents the maximum deadweight of the i-th vessel; Indicates the preset scheduling period; This indicates the total actual demand for green ammonia.

[0077] An improved genetic algorithm is used to solve the dynamic production scheduling model at a preset update time interval, and the dynamic production scheduling strategy is updated based on the solution results. Specifically, this includes the ammonia production plan of the distributed ammonia production platform and the ship transportation path of the multifunctional ship node.

[0078] Understandably, the multi-objective optimization scheduling module can first extract key information from real-time green ammonia production data and ship transportation status data, such as the total power generation of renewable energy and the actual power used for hydrogen production and ammonia synthesis; the current operating power and maximum capacity of hydrogen production and ammonia synthesis equipment; the liquid ammonia inventory of each production platform; and the number of transport ships, sailing distance, carrying capacity, and actual ammonia load, etc.

[0079] Then, by calculating the total electricity consumption for hydrogen production and ammonia synthesis, i.e., the total electricity demand, and combining this with a preset scheduling cycle, such as 15 minutes, and market demand forecasts, a dynamic production scheduling model can be constructed. This model includes three core optimization objectives: maximizing energy utilization to ensure that renewable energy is used for hydrogen production and ammonia synthesis, reducing wasted electricity; minimizing transportation costs by optimizing ship routes and load allocation to reduce energy consumption and time costs during transportation; and minimizing inventory fluctuations by balancing the liquid ammonia inventory of each ammonia production platform to avoid situations where some platforms have excess inventory while others have shortages. Simultaneously, the model incorporates multiple constraints, such as ensuring that equipment capacity does not exceed maximum limits, ship load does not exceed safety thresholds, and total output meets market demand, to ensure the feasibility and safety of the production scheduling strategy.

[0080] Furthermore, the multi-objective optimization scheduling module adopts an improved genetic algorithm, which is an optimization search algorithm that simulates natural selection and genetic mechanisms. Every certain period of time, such as 5 minutes, it automatically solves the dynamic production scheduling model and generates the optimal production plan and transportation path.

[0081] For example, when the renewable energy generation of an ammonia production platform surges, the improved genetic algorithm prioritizes allocating the increased production to that platform and schedules nearby ships to transport its surplus products. When market demand suddenly increases, the algorithm coordinates multiple platforms to adjust their capacity and plans the shortest path transportation to ensure supply and demand balance. This dynamic adjustment mechanism enables the entire production scheduling system to respond to environmental changes in real time, significantly improving resource utilization efficiency and production flexibility.

[0082] The fault prediction and emergency response module is used to construct a comprehensive fault prediction model based on an LSTM neural network, to predict anomalies in the distributed ammonia production platform, and to trigger a comprehensive emergency response mechanism in advance, specifically including:

[0083] Feature extraction is performed on the real-time green ammonia production data to obtain the corresponding real-time green ammonia production features;

[0084] The real-time green ammonia production characteristics are input into the forget gate of the comprehensive fault prediction model to determine the real-time green ammonia production characteristics that need to be discarded from the memory unit. The corresponding calculation formula is as follows:

[0085] In the formula, The output of the forget gate at the current time t is used to determine the real-time green ammonia production features that need to be discarded from the memory unit; Indicates the activation function; Indicates the weight of the forget gate; This represents the hidden state at the previous time step t-1; This represents the real-time green ammonia production characteristics input at the current time t; Indicates hidden state and real-time green ammonia production characteristics The vector formed; Indicates the bias of the forget gate;

[0086] Based on the input gate of the comprehensive fault prediction model, the real-time green ammonia production characteristics that need to be stored in the memory unit are determined, and the corresponding calculation formula is as follows:

[0087] In the formula, The output of the input gate at the current time t is used to determine the real-time green ammonia production characteristics that need to be stored in the memory unit; Indicates the weights of the input gates; Indicates the bias of the input gate;

[0088] Candidate green ammonia production characteristics are determined based on the candidate memory units of the comprehensive fault prediction model, and the corresponding calculation formula is as follows:

[0089] In the formula, This represents the candidate memory unit at the current time t, used to determine the candidate green ammonia production characteristics; Represents the hyperbolic tangent function; Indicates the weight of the candidate memory unit; Indicates the bias of candidate memory units;

[0090] The calculation formula for updating the memory unit is as follows:

[0091] In the formula, This represents the memory unit at the current time t; This represents the memory unit from the previous time step t-1;

[0092] Based on the output gate of the comprehensive fault prediction model, the candidate green ammonia production characteristics that need to be output to the hidden state are determined, and the corresponding calculation formula is as follows:

[0093] In the formula, This represents the output of the output gate at the current time t, used to determine the candidate green ammonia production features that need to be output to the hidden state. Indicates the weight of the output gate; Indicates the bias of the output gate;

[0094] The hidden state at the current moment is determined based on the memory unit and the output of the output gate, and the corresponding calculation formula is as follows:

[0095] In the formula, This represents the hidden state at the current time t.

[0096] The hidden state at the current moment is mapped to the comprehensive failure probability through a fully connected layer, and the corresponding calculation formula is as follows:

[0097] In the formula, This represents the overall failure probability at the current time t. ; Indicates the weights of the fully connected layer; Indicates the bias of the fully connected layer;

[0098] Obtain a critical fault threshold, and when the overall fault probability is greater than or equal to the critical fault threshold, trigger the overall emergency response mechanism in advance.

[0099] Understandably, by leveraging the time-series data processing capabilities of LSTM neural networks, comprehensive fault prediction models can automatically capture the long-term dependencies and subtle fluctuation patterns of equipment operating parameters. For example, the forget gate mechanism can filter out random noise under normal operating conditions, such as short-term wind power fluctuations, while the input gate and candidate memory units focus on persistently abnormal features, such as the electrolytic cell voltage continuously deviating from the threshold, thereby enabling early identification of equipment performance degradation or potential faults.

[0100] Furthermore, this model can construct a dynamic digital twin of the platform's state through iterative updates of multi-layered memory units, combining historical operating patterns with real-time data to predict the platform's overall failure probability. When the overall failure probability exceeds a critical threshold, such as 0.9, the system will trigger an emergency response in advance, such as automatically switching to a backup ammonia production platform, dispatching ships to transfer inventory, and activating energy storage devices to maintain production.

[0101] By employing the above methods, the fault handling mechanism can be shifted from a passive shutdown and maintenance mode to a proactive risk isolation mode, reducing production interruption time caused by sudden failures and avoiding delays and misjudgments due to manual intervention. For example, when the temperature of a reactor on a certain platform rises abnormally, the system will issue an early warning and activate adjacent platforms to increase production, while simultaneously scheduling the nearest vessel to transfer inventory. The entire process requires no manual intervention, significantly improving the system's anti-interference capability and operational stability in complex marine environments, and ensuring the continuity and safety of offshore green ammonia production.

[0102] The supply chain management module uses the inventory optimization algorithm to generate the target replenishment instruction, specifically including:

[0103] The average demand rate, standard deviation of the demand rate, and safety factor for demand fluctuations are obtained. The round-trip shipping time from the ship transportation status data is extracted, and the inventory optimization algorithm is used to calculate the inventory level triggering replenishment as follows:

[0104] In the formula, CBK represents the inventory level that triggers replenishment; WFSJ represents the average demand rate; WFSJ represents the round-trip shipping time. Indicates the safety factor for demand fluctuations; Indicates the standard deviation of the demand rate;

[0105] The corresponding target replenishment instruction is generated based on the triggered replenishment inventory level.

[0106] It should be noted that the supply chain management module can use inventory optimization algorithms to generate target replenishment instructions, enabling dynamic and precise control of green ammonia inventory.

[0107] This process integrates core factors such as demand stability, volatility risk, and transportation timeliness. It determines the basic replenishment quantity by analyzing the average demand rate, quantifies the degree of demand volatility by combining the standard deviation of the demand rate, amplifies the safety redundancy by using the demand volatility safety factor, and then incorporates the round-trip transportation time of ships to calculate the impact of replenishment delays, ultimately forming a scientifically determined replenishment inventory level.

[0108] Furthermore, the target replenishment instruction generated based on the triggered replenishment inventory level can avoid the inventory backlog or shortage problems caused by traditional experience-based replenishment, reduce storage costs and energy waste caused by excess inventory, and ensure supply continuity by replenishing in advance when demand changes suddenly. This data-driven replenishment mechanism significantly improves the inventory turnover rate of the entire chain, enhances the supply chain's adaptability to complex demand scenarios at sea, and ensures efficient coordination among all aspects of green ammonia production, transportation, and consumption.

[0109] Example 2

[0110] like Figure 2 As shown, a method for scheduling distributed green ammonia production at sea based on intelligent management is described, the production scheduling method comprising:

[0111] S1, acquire the real-time green ammonia production data in the distributed ammonia production platform and the ship transportation status data in the multi-functional ship node and upload them to the data transmission module;

[0112] S2, the real-time green ammonia production data and the ship transportation status data are transmitted to the multi-objective optimization scheduling module, the fault prediction and emergency response module and the supply chain management module through the data transmission module;

[0113] S3, Based on the multi-objective optimization scheduling module, the dynamic allocation algorithm is used to update the dynamic production scheduling strategy;

[0114] S4. Based on the fault prediction and emergency response module, construct the comprehensive fault prediction model based on LSTM neural network to predict anomalies in the distributed ammonia production platform and trigger the comprehensive emergency response mechanism in advance.

[0115] S5, based on the supply chain management module, the target replenishment instruction is generated using the inventory optimization algorithm. The dynamic production scheduling strategy is adjusted in real time based on the target replenishment instruction and sent to the distributed ammonia production platform and the multi-functional ship. The intelligent green ammonia management platform includes the data transmission module, the multi-objective optimization scheduling module, the fault prediction and emergency response module, and the supply chain management module.

[0116] Through the above embodiments, this invention, through a marine distributed green ammonia production scheduling system and method based on intelligent management, can acquire real-time green ammonia production data from a distributed ammonia production platform and ship transportation status data from multi-functional ship nodes, and upload them to a data transmission module. The data transmission module transmits the real-time green ammonia production data and ship transportation status data to a multi-objective optimization scheduling module, a fault prediction and emergency response module, and a supply chain management module. Based on the multi-objective optimization scheduling module, a dynamic allocation algorithm is used to update the dynamic production scheduling strategy. Based on the fault prediction and emergency response module, a comprehensive fault prediction model based on an LSTM neural network is constructed to predict anomalies in the distributed ammonia production platform and trigger a comprehensive emergency response mechanism in advance. Based on the supply chain management module, an inventory optimization algorithm is used to generate target replenishment instructions, and the dynamic production scheduling strategy is adjusted in real time based on these instructions and sent to the distributed ammonia production platform and multi-functional ship nodes. The intelligent green ammonia management platform includes a data transmission module, a multi-objective optimization scheduling module, a fault prediction and emergency response module, and a supply chain management module, thereby achieving intelligent closed-loop management of the entire process of green ammonia production, transportation, and consumption, significantly improving production scheduling efficiency and energy utilization, and reducing transportation costs.

[0117] This invention improves renewable energy utilization and significantly reduces wind and electricity curtailment losses through a multi-objective optimization scheduling algorithm and real-time data-driven approach. Furthermore, by dynamically coordinating a distributed ammonia production platform with ship nodes, it shortens the green ammonia production cycle, reduces unit capacity costs, and increases inventory turnover, effectively alleviating resource waste caused by supply-demand imbalances in traditional models. This invention accurately predicts abnormal situations on the ammonia production platform using a comprehensive fault prediction model based on LSTM neural networks, combined with a minute-level emergency response mechanism, reducing system downtime and significantly mitigating the risk of production interruptions due to unforeseen events. It further enhances the system's robustness in extreme weather or equipment failure scenarios, ensuring the continuity and safety of offshore operations. The multi-functional ship nodes of this invention overcome the shortcomings of traditional long-distance transportation modes in land-based production through a three-in-one design integrating production, transportation, and consumption, reducing transportation costs. This invention achieves full-chain traceability and dynamic balance of green ammonia from production to consumption through blockchain smart contracts and inventory optimization algorithms, forming a closed-loop offshore energy network and significantly reducing the environmental impact throughout the entire green ammonia production lifecycle.

[0118] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0119] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0120] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

Claims

1. A marine distributed green ammonia production scheduling system based on intelligent management, characterized in that: The production scheduling system includes: The distributed ammonia production platform is used for the distributed production of green ammonia at sea. It is equipped with renewable energy power generation equipment, water electrolysis hydrogen production equipment, ammonia synthesis reactor, green ammonia energy storage device, and a sensor network for the ammonia production platform. A multi-functional ship node is used for the production, transportation and consumption of green ammonia at sea. It is equipped with a green ammonia transport carrier, an offshore production node expansion unit, a green ammonia fuel consumption terminal and ship status sensors. The intelligent green ammonia management platform integrates a data transmission module, a multi-objective optimization scheduling module, a fault prediction and emergency response module, and a supply chain management module. It is used to communicate with the distributed ammonia production platform and the multi-functional ship node through Internet of Things technology. The data transmission module is used to acquire real-time green ammonia production data of the distributed ammonia production platform and ship transportation status data of the multi-functional ship nodes, and encrypt and transmit the real-time green ammonia production data and ship transportation status data to the multi-objective optimization scheduling module, the fault prediction and emergency response module, and the supply chain management module based on a blockchain network. The multi-objective optimization scheduling module is used to update the dynamic production scheduling strategy through a dynamic allocation algorithm. The fault prediction and emergency response module is used to construct a comprehensive fault prediction model based on an LSTM neural network to predict anomalies in the distributed ammonia production platform and trigger a comprehensive emergency response mechanism in advance. The supply chain management module is used to generate target replenishment instructions using an inventory optimization algorithm, adjust the dynamic production scheduling strategy in real time based on the target replenishment instructions, and send them to the distributed ammonia production platform and the multi-functional ship nodes. The ammonia production platform sensor network collects energy supply data from the renewable energy power generation device, hydrogen synthesis data from the water electrolysis hydrogen production device, ammonia synthesis data from the ammonia synthesis reactor, and green ammonia storage data from the green ammonia energy storage device at preset acquisition time intervals. The energy supply data, hydrogen synthesis data, ammonia synthesis data, and green ammonia storage data are then aggregated to obtain the real-time green ammonia production data. The ship status sensor is used to collect the ship transportation status data of the multifunctional ship node; Based on Internet of Things (IoT) technology, the collected real-time green ammonia production data and ship transportation status data are uploaded to the data transmission module; After receiving the real-time green ammonia production data and the ship transportation status data, the multi-objective optimization scheduling module extracts the renewable energy power generation from the energy supply data; the hydrogen production and consumption power, hydrogen synthesis power, and maximum hydrogen synthesis capacity from the hydrogen synthesis data; the ammonia synthesis and consumption power, ammonia synthesis power, and maximum ammonia synthesis capacity from the ammonia synthesis data; the liquid ammonia inventory from the green ammonia storage data; and the total number of transport vessels, ship transport distance, ship load factor, actual ammonia load, and maximum ship load capacity from the ship transportation status data. It then calculates the sum of the hydrogen production and consumption power and the ammonia synthesis and consumption power to obtain the total consumption power, and acquires the preset scheduling cycle and the actual total green ammonia demand. A dynamic production scheduling model is constructed based on the aforementioned dynamic allocation algorithm. This model includes a multi-objective optimization scheduling function and corresponding constraints. The multi-objective optimization scheduling function aims to maximize energy utilization, minimize transportation costs, and minimize inventory fluctuations. The constraints include equipment capacity limits, ship load limits, and market demand constraints. The corresponding calculation formulas are as follows: In the formula, Indicates energy utilization rate; Indicates transportation costs; Indicates inventory fluctuation; This indicates the total amount of electricity consumed; M represents the amount of electricity generated from renewable energy sources; M represents the total number of transport vessels. This represents the shipping distance of the i-th ship. represents the load factor of the i-th ship; N represents the total number of distributed ammonia production platforms; This represents the liquid ammonia inventory of the j-th distributed ammonia production platform; This represents the average liquid ammonia inventory across all distributed ammonia production platforms; max indicates the maximum value operation; min indicates the minimum value operation. This represents the hydrogen synthesis power of the j-th distributed ammonia production platform; This represents the maximum hydrogen synthesis capacity of the j-th distributed ammonia production platform; This represents the ammonia synthesis power of the j-th distributed ammonia production platform; This represents the maximum ammonia synthesis capacity of the j-th distributed ammonia production platform; This represents the actual ammonia load of the i-th ship; This represents the maximum deadweight of the i-th vessel; Indicates the preset scheduling period; This indicates the total actual demand for green ammonia.

2. The marine distributed green ammonia production scheduling system based on intelligent management according to claim 1, characterized in that, An improved genetic algorithm is used to solve the dynamic production scheduling model at a preset update time interval, and the dynamic production scheduling strategy is updated based on the solution results. Specifically, this includes the ammonia production plan of the distributed ammonia production platform and the ship transportation path of the multifunctional ship node.

3. The marine distributed green ammonia production scheduling system based on intelligent management according to claim 1, characterized in that, The fault prediction and emergency response module is used to construct a comprehensive fault prediction model based on an LSTM neural network, to predict anomalies in the distributed ammonia production platform, and to trigger a comprehensive emergency response mechanism in advance, specifically including: Feature extraction is performed on the real-time green ammonia production data to obtain the corresponding real-time green ammonia production features; The real-time green ammonia production characteristics are input into the forget gate of the comprehensive fault prediction model to determine the real-time green ammonia production characteristics that need to be discarded from the memory unit. The corresponding calculation formula is as follows: In the formula, The output of the forget gate at the current time t is used to determine the real-time green ammonia production features that need to be discarded from the memory unit; Indicates the activation function; Indicates the weight of the forget gate; This represents the hidden state at the previous time step t-1; This represents the real-time green ammonia production characteristics input at the current time t; Indicates hidden state and real-time green ammonia production characteristics The vector formed; Indicates the bias of the forget gate; Based on the input gate of the comprehensive fault prediction model, the real-time green ammonia production characteristics that need to be stored in the memory unit are determined, and the corresponding calculation formula is as follows: In the formula, The output of the input gate at the current time t is used to determine the real-time green ammonia production characteristics that need to be stored in the memory unit; Indicates the weights of the input gates; Indicates the bias of the input gate; Candidate green ammonia production characteristics are determined based on the candidate memory units of the comprehensive fault prediction model, and the corresponding calculation formula is as follows: In the formula, This represents the candidate memory unit at the current time t, used to determine the candidate green ammonia production characteristics; Represents the hyperbolic tangent function; Indicates the weight of the candidate memory unit; Indicates the bias of candidate memory units; The calculation formula for updating the memory unit is as follows: In the formula, This represents the memory unit at the current time t; This represents the memory unit from the previous time step t-1; Based on the output gate of the comprehensive fault prediction model, the candidate green ammonia production characteristics that need to be output to the hidden state are determined, and the corresponding calculation formula is as follows: In the formula, This represents the output of the output gate at the current time t, used to determine the candidate green ammonia production features that need to be output to the hidden state. Indicates the weight of the output gate; Indicates the bias of the output gate; The hidden state at the current moment is determined based on the memory unit and the output of the output gate, and the corresponding calculation formula is as follows: In the formula, This represents the hidden state at the current time t.

4. The marine distributed green ammonia production scheduling system based on intelligent management according to claim 3, characterized in that, The hidden state at the current moment is mapped to the comprehensive failure probability through a fully connected layer, and the corresponding calculation formula is as follows: In the formula, This represents the overall failure probability at the current time t. ; Indicates the weights of the fully connected layer; Indicates the bias of the fully connected layer; Obtain a critical fault threshold, and when the overall fault probability is greater than or equal to the critical fault threshold, trigger the overall emergency response mechanism in advance.

5. The marine distributed green ammonia production scheduling system based on intelligent management according to claim 1, characterized in that, The supply chain management module uses the inventory optimization algorithm to generate the target replenishment instruction, specifically including: The average demand rate, standard deviation of the demand rate, and safety factor for demand fluctuations are obtained. The round-trip shipping time from the ship transportation status data is extracted, and the inventory optimization algorithm is used to calculate the inventory level triggering replenishment as follows: In the formula, CBK represents the inventory level that triggers replenishment; WFSJ represents the average demand rate; WFSJ represents the round-trip shipping time. Indicates the safety factor for demand fluctuations; Indicates the standard deviation of the demand rate; The corresponding target replenishment instruction is generated based on the triggered replenishment inventory level.

6. A method for scheduling marine distributed green ammonia production based on intelligent management, applied to the marine distributed green ammonia production scheduling system based on intelligent management as described in any one of claims 1-5, characterized in that, The production scheduling method includes: The real-time green ammonia production data from the distributed ammonia production platform and the ship transportation status data from the multi-functional ship node are acquired and uploaded to the data transmission module. The real-time green ammonia production data and the ship transportation status data are transmitted through the data transmission module to the multi-objective optimization scheduling module, the fault prediction and emergency response module, and the supply chain management module. Based on the multi-objective optimization scheduling module, the dynamic allocation algorithm is used to update the dynamic production scheduling strategy; Based on the fault prediction and emergency response module, a comprehensive fault prediction model based on LSTM neural network is constructed to predict anomalies in the distributed ammonia production platform and trigger the comprehensive emergency response mechanism in advance. Based on the supply chain management module, the target replenishment instruction is generated using the inventory optimization algorithm. The dynamic production scheduling strategy is adjusted in real time based on the target replenishment instruction and sent to the distributed ammonia production platform and the multi-functional ship node. The intelligent green ammonia management platform includes the data transmission module, the multi-objective optimization scheduling module, the fault prediction and emergency response module, and the supply chain management module.

Images