A distributed data optimization method and device for a wind-solar-hydrogen-ammonia alcohol production process

Abstract

This application provides a distributed data optimization method and apparatus for the production process of hydrogen, ammonia, and methanol from wind and solar power, belonging to the field of new energy comprehensive utilization technology. The method involves: collecting data from wind and solar power generation, energy storage systems, hydrogen / ammonia / methanol production process units, and end-use energy systems, and inputting this data into a digital twin model. The optimization targets are green electricity utilization rate, product output efficiency, and energy storage loss rate. Constraints include energy balance, material balance, equipment operation, and safety. A collaborative optimization model is constructed, combining real-time collected data. Edge computing devices perform local optimization solutions for each process unit, while a cloud server performs global optimization solutions and coordination, outputting the optimal operating parameter settings for each process unit and distributing them. Actual operating data is then collected, and the parameters of the collaborative optimization model are dynamically adjusted based on the deviation between the actual operating data and the expected targets. This application improves system energy efficiency, optimizes energy utilization, enhances operational stability, and adapts to dynamic demands.

Description

Technical Field

[0001] This application belongs to the field of new energy comprehensive utilization technology, specifically relating to a distributed data optimization method and device for the wind-solar-hydrogen-ammonia-ethanol production process. Background Technology

[0002] The integrated wind-solar-hydrogen-ammonia-ethanol system converts wind and solar power into green hydrogen, which is then used for ammonia production via hydrogen-nitrogen synthesis and alcohol production via hydrogen-carbon coupling. This enables cross-scenario energy storage and utilization, becoming an important direction for non-electrical utilization of new energy sources, especially suitable for distributed energy scenarios such as highway service areas. However, existing integrated wind-solar-hydrogen-ammonia-ethanol systems have the following problems: First, dynamic matching between wind and solar power generation and the production process is difficult. Wind and solar power generation is highly volatile and random, and traditional centralized control modes struggle to achieve real-time adaptation between hydrogen, ammonia, and alcohol production loads and wind and solar power, resulting in low green electricity utilization and significant wind and solar curtailment. Second, there is insufficient coordinated optimization of energy and material flows. The production processes of hydrogen, ammonia, and alcohol involve multiple energy and material conversion stages. Existing systems often employ independent control strategies, lacking unified data modeling and coordinated optimization mechanisms, leading to low system efficiency. Third, data utilization of distributed production units is inadequate. Operational data, equipment status data, and energy storage data from each production unit (hydrogen, ammonia, and alcohol production) are collected and processed independently, failing to form a data-driven global optimization capability, making it difficult to achieve refined scheduling and real-time response of the production process. Finally, there is a lack of dedicated optimization methods suitable for scenarios such as highway service areas. Existing methods are mostly designed for large chemical industrial parks or centralized energy systems, failing to consider the characteristics of highway service area scenarios such as compact space, large load fluctuations, and high safety requirements, making direct application difficult.

[0003] Therefore, there is an urgent need for a method that can adapt to wind and solar fluctuations, achieve distributed collaborative optimization of the hydrogen ammonia production process, and improve the overall energy efficiency and operational stability of the system. Summary of the Invention

[0004] In a first aspect, embodiments of this application provide a distributed data optimization method for the production process of hydroammonium phosphate from wind and solar power, comprising the following steps: S1. Collect and preprocess data from wind and solar power generation, energy storage systems, hydrogen production process units, ammonia production process units, alcohol production process units, and end-use energy systems. S2. Input the preprocessed data into the pre-set digital twin model of the wind-solar-hydrogen-ammonia-ethanol integrated system, and based on the digital twin model, construct and update the distributed collaborative optimization model with green electricity utilization rate, product output efficiency and energy storage loss rate as optimization objectives, and energy balance, material balance, equipment operation and safety as constraints. S3. Based on the distributed collaborative optimization model and the preprocessed data input in real time, the edge computing devices deployed locally in each process unit perform the local optimization solution for the corresponding process unit, and the cloud server performs the global optimization solution and coordination, outputting the optimal operating parameter settings for each process unit. S4. Send the optimal operating parameter settings to the corresponding process units for execution, collect actual operating data, and dynamically adjust the parameters of the distributed collaborative optimization model based on the deviation between the actual operating data and the expected target; S5. Visualize the optimization process data, solution results, and operating status of the integrated wind-solar-hydrogen-ammonia-ethanol system.

[0005] Furthermore, the specific steps of step S1 are as follows: S11. Collect real-time data of the wind and solar power generation system. The real-time data of the wind and solar power generation system includes the output power of the photovoltaic array, the output power of the wind turbine, and the ultra-short-term power prediction value for a future set time period. S12. Collect the status data of the energy storage system. The status data of the energy storage system includes the state of charge and charge / discharge power of electrochemical energy storage, as well as the hydrogen storage tank capacity and hydrogen inflow / outflow of hydrogen energy storage. S13. Collect real-time operating data of each process unit. The real-time operating data of each process unit includes the electrolysis current, voltage, electrolyte temperature and hydrogen production rate of the hydrogen production process unit, the reactor temperature, pressure, catalyst activity coefficient and ammonia production of the ammonia production process unit, and the CO2 capture, synthesis reaction temperature, pressure and methanol production of the methanol production process unit. S14. Collect demand data from end-use energy systems, including the immediate demand for hydrogen, ammonia fuel, and methanol fuel for transportation. S15. Preprocess the collected real-time data, status data, real-time running data, and demand data. The preprocessing includes outlier detection and removal, missing value imputation, and data standardization to generate a unified preprocessed running data stream.

[0006] Furthermore, the specific steps of step S2 are as follows: S21. Map the preprocessed running data stream to the digital twin model of the integrated wind-solar-hydrogen-ammonia-ethanol system in real time, and update the state parameters of each virtual entity component in the digital twin model; S22. Based on the real-time state reflected by the updated digital twin model, a multi-objective function F of the distributed collaborative optimization model is defined, wherein the multi-objective function F is expressed as:

[0007] in, To improve the utilization rate of green electricity, To improve overall product output efficiency, Energy storage loss rate, , , These are dynamic weighting coefficients; S23. Load optimization objectives and constraints into the digital twin model; The constraints include at least the following: Real-time energy balance constraints between wind and solar power generation, energy storage charging and discharging power and energy consumption of each process unit; Input and output material balance constraints of hydrogen, nitrogen, and carbon elements in each production stage; Temperature and pressure safety boundary constraints for the operation of each process unit equipment.

[0008] Furthermore, the specific steps of step S3 are as follows: S31. The local edge computing devices of each process unit, based on the preprocessed data received in real time and the local constraints obtained from the digital twin model, solve the local optimization sub-problem with the goal of optimizing their own process efficiency, and generate preliminary operating parameters. S32. The cloud server receives preliminary operating parameters and related status data uploaded by all edge computing devices; S33. The cloud server, guided by the global objective function of the distributed collaborative optimization model, coordinates the initial operating parameters of each process unit and solves the overall optimal solution of the wind-solar-hydrogen-ammonia-methanol integrated system through a global optimization algorithm. S34. The cloud server will send the optimal operating parameter settings of each process unit obtained from the solution to the corresponding edge computing devices.

[0009] Furthermore, the global optimization algorithm in step S33 adopts a distributed model predictive control algorithm, and the specific steps are as follows: S331. The cloud server is based on a digital twin model to construct a global centralized prediction model that couples the dynamic relationships between wind and solar power generation, energy storage systems and various process units; S332. In each coordination cycle, using the real-time status data gathered from all edge computing devices as the initial value and the global objective function of the distributed collaborative optimization model as the performance index, within the first prediction time domain preset based on the wind and solar power generation fluctuation cycle and the dynamic response time of the ammonia / methanol production process, solve the constrained optimization problem to obtain the future control sequence that makes the wind-solar-hydrogen-ammonia-methanol integrated system optimal as a whole. S333. The control quantities of each unit at the first moment in the future control sequence are sent as the optimal operating parameter settings to the corresponding edge computing devices for cross-unit collaboration.

[0010] Furthermore, after step S34 in step S3, the following steps are also included: S35. If the fluctuation range of wind and solar power generation exceeds the preset threshold, or if there is a sudden change in end-user demand, the cloud server will trigger an emergency recalculation process of the distributed collaborative optimization model and prioritize ensuring that safety constraints are met.

[0011] Furthermore, the specific steps of step S4 are as follows: S41. The edge computing device converts the received optimal operating parameter settings into specific control commands and sends them to the actuators of the corresponding hydrogen production process unit, ammonia production process unit, or alcohol production process unit. S42. Collect the actual operating data of each process unit after executing control commands in real time, and calculate the deviation between the actual operating data and the optimal operating parameter setting value; S43. Based on the aforementioned deviation, a model predictive control algorithm is used to dynamically adjust the local model parameters or dynamic weight coefficients of the corresponding process unit in the distributed collaborative optimization model. , , This is to achieve closed-loop feedback optimization.

[0012] Furthermore, in step S43, the execution process of the model predictive control algorithm includes the following steps: S431. The edge computing device constructs a local simplified prediction model based on the sub-model corresponding to this process unit in the digital twin model; S432. In each control cycle, with the actual operating data collected by this process unit and the set value received from the cloud server as input, and with the goal of quickly and accurately tracking the set value, an optimization problem with local constraints is solved in a second prediction time domain shorter than the first prediction time domain to obtain the optimal control command for this process unit. S433. The optimal control command obtained from the solution is sent to the actuator of this process unit, and the optimization is continuously updated based on the new data after execution to achieve closed-loop elimination of set value deviation.

[0013] Furthermore, the specific steps of step S5 are as follows: S51. In the visualization interface, the real-time running status of each virtual component in the digital twin model, the key indicator curves of the preprocessed running data stream, and the solution process and results of the distributed collaborative optimization model are dynamically displayed. S52. Show the comprehensive energy efficiency indicators of the integrated wind-solar-hydrogen-ammonia-methanol system. The comprehensive energy efficiency indicators include real-time green electricity consumption rate, hydrogen / ammonia / methanol production rate and equivalent carbon emission reduction. S53. Through the human-computer interaction interface, respond to the operator's request to view the basis for optimization decisions, and allow the operator to confirm or manually fine-tune the optimal operating parameter settings.

[0014] Secondly, embodiments of this application also provide a distributed data optimization device for the production process of hydroammonium from wind and solar power, comprising: The data acquisition and preprocessing module is used to collect and preprocess data from wind and solar power generation, energy storage systems, hydrogen production process units, ammonia production process units, alcohol production process units, and end-use energy systems. The digital twin and optimization modeling module is used to input pre-processed data into a pre-set digital twin model of the wind-solar-hydrogen-ammonia-ethanol integrated system, and based on the digital twin model, with green electricity utilization rate, product output efficiency and energy storage loss rate as optimization objectives, and with energy balance, material balance, equipment operation and safety as constraints, to construct and update a distributed collaborative optimization model. The distributed collaborative optimization solution module includes: Edge computing devices deployed locally in each process unit perform local optimization solutions for the corresponding process unit based on a distributed collaborative optimization model and preprocessed data input in real time. The cloud server performs global optimization and coordination, and outputs the optimal operating parameter settings for each process unit. The dynamic feedback and execution control module is used to send the optimal operating parameter settings to the corresponding process units for execution, and to collect actual operating data. Based on the deviation between the actual operating data and the expected target, the parameters of the distributed collaborative optimization model are dynamically adjusted. The visualization and human-computer interaction module is used to visualize and display the optimization process data, solution results, and operating status of the integrated wind-solar-hydrogen-ammonia-methanol system.

[0015] As can be seen from the above technical solutions, this application has the following advantages: The distributed data optimization method and apparatus for the wind-solar-hydrogen-ammonia-methanol production process provided in this application achieves dynamic matching between wind and solar power generation and the hydrogen-ammonia-methanol production process through distributed collaborative optimization, improving the utilization rate of green electricity, reducing wind and solar curtailment, and optimizing the coordination of energy flow and material flow, thereby improving the overall energy efficiency of the system. It can respond in real time to the fluctuations in wind and solar power generation and changes in end-user demand, quickly adjusting the operating parameters of each process unit to adapt to the characteristics of large load fluctuations in distributed energy scenarios such as highway service areas. It integrates the operating data, equipment status data, and energy storage data of each production unit, constructing a distributed collaborative optimization model based on digital twins, forming a data-driven global optimization capability, and improving the level of refined scheduling of the production process. During the optimization process, the safety boundary constraints of equipment operation are always taken as an important condition to ensure that each process unit operates within a safe range, and safety constraints are prioritized in emergency situations. By visually displaying the optimization process data, solution results, and system operating status, it provides operators with intuitive decision-making basis and allows them to confirm or manually fine-tune the optimal operating parameter settings, improving the operability and management efficiency of the system. Attached Figure Description

[0016] To more clearly illustrate the technical solution of this application, the accompanying drawings used in the description will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the distributed data optimization method for the production process of hydroammonium phosphate (HAP) of the present invention.

[0018] Figure 2 This is a schematic diagram of the distributed data optimization device for the wind-solar hydrogen ammonia production process of the present invention. Detailed Implementation

[0019] The various embodiments of this disclosure will be described more fully in the detailed steps of the distributed data optimization method for the production process of hydroammonium from wind and solar power. This disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of this disclosure to the specific embodiments disclosed herein, but rather this disclosure should be understood to cover all adjustments, equivalents, and / or alternatives falling within the spirit and scope of the various embodiments of this disclosure.

[0020] This embodiment provides a distributed data optimization method for the production process of hydrogen ammonia and methanol from wind and solar power, which improves system energy efficiency, optimizes energy utilization, enhances operational stability, and adapts to dynamic demands.

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1 The diagram shows a flowchart of a distributed data optimization method for the production process of hydrogen ammonia and methanol using wind and solar power, in a specific embodiment. The method includes the following steps: S1. Collect and preprocess data from wind and solar power generation, energy storage systems, hydrogen production process units, ammonia production process units, alcohol production process units, and end-use energy systems. It should be noted that by collecting data from wind and solar power generation, energy storage systems, various process units, and end-user energy systems, a comprehensive understanding of the system's operational information is obtained, providing a data foundation for optimization. Preprocessing operations such as outlier detection and removal, missing value imputation, and data standardization are performed on the collected data to improve data quality and consistency, ensuring the accuracy of optimization modeling. S2. Input the preprocessed data into the pre-set digital twin model of the wind-solar-hydrogen-ammonia-ethanol integrated system, and based on the digital twin model, construct and update the distributed collaborative optimization model with green electricity utilization rate, product output efficiency and energy storage loss rate as optimization objectives, and energy balance, material balance, equipment operation and safety as constraints. It should be noted that by inputting the preprocessed data into the digital twin model, a virtual mapping of the integrated wind-solar-hydrogen-ammonia-ethanol system was achieved, providing an intuitive and accurate description of the system state for optimization modeling. With green electricity utilization rate, product output efficiency, and energy storage loss rate as optimization objectives, and energy balance, material balance, equipment operation and safety as constraints, a distributed collaborative optimization model was constructed, clarifying the direction and boundaries of optimization, making the optimization process targeted and feasible. S3. Based on the distributed collaborative optimization model and the preprocessed data input in real time, the edge computing devices deployed locally in each process unit perform the local optimization solution for the corresponding process unit, and the cloud server performs the global optimization solution and coordination, outputting the optimal operating parameter settings for each process unit. It should be noted that the local edge computing devices of each process unit perform local optimization and solve the problem based on their own data and constraints, generating preliminary operating parameters. This fully leverages the advantages of local computing and improves optimization efficiency. The cloud server, guided by the global objective function, coordinates the preliminary operating parameters of each process unit, solves the overall optimal solution of the system, and distributes the optimal operating parameter settings to each edge computing device. This achieves synergy between local and global optimization, ensuring optimal overall system performance. S4. Send the optimal operating parameter settings to the corresponding process units for execution, collect actual operating data, and dynamically adjust the parameters of the distributed collaborative optimization model based on the deviation between the actual operating data and the expected target; It should be noted that the optimal operating parameter settings are sent to the corresponding process units for execution, and actual operating data is collected to provide real-time feedback information for the dynamic adjustment of the system. Based on the deviation between the actual operating data and the expected target, the parameters of the distributed collaborative optimization model are dynamically adjusted, realizing closed-loop feedback optimization, improving the system's adaptability and adjustment capability to actual operating deviations, and ensuring the continuity and stability of the optimization effect. S5. Visualize the optimization process data, solution results, and operating status of the integrated wind-solar-hydrogen-ammonia-ethanol system; It should be noted that the visualization of optimization process data, solution results, and the operating status of the integrated wind-solar-hydrogen-ammonia-ethanol system allows operators to intuitively understand the system's operation and optimization effects, facilitating monitoring and management. Through a human-machine interface, the system responds to operators' requests to view the basis for optimization decisions and allows them to confirm or manually fine-tune the optimal operating parameter settings, improving the system's operability and flexibility and enhancing operators' control over the system.

[0023] This embodiment improves the overall energy efficiency of the wind-solar-hydrogen-ammonia-methanol production system through distributed data optimization, realizes dynamic matching and refined scheduling of energy, and enhances system flexibility and operational stability.

[0024] Furthermore, as a refinement and extension of the specific implementation methods of the above embodiments, in order to fully illustrate the specific implementation process of this embodiment, another distributed data optimization method for the production process of wind-solar-hydrogen-ammonia-methanol is provided, taking a certain highway service area integrated wind-solar-hydrogen-ammonia-methanol system as the application object. The terms and definitions in this embodiment are as follows: Standard state: refers to the gaseous state at 0℃ and 1 atmosphere (101325Pa); Hydrogen conversion standard: The volume (Nm³) and mass (kg) conversions of hydrogen mentioned in this article are all based on the hydrogen density of 0.08988 kg / Nm³ under standard conditions; Standard coal equivalent coefficients: hydrogen 12 kgce / kg, ammonia 5 kgce / kg, methanol 8 kgce / kg, referencing industry-standard energy conversion rates; The system is configured as follows: Wind and solar power generation system: 10MW photovoltaic array + 5MW wind turbine generator, with the ultra-short-term power forecast period set at 30 minutes; Energy storage system: 20MWh electrochemical energy storage device + 1000kg-class high-pressure hydrogen energy storage system; Process units: 500 Nm³ / h water electrolysis hydrogen production unit, 300 kg / h hydrogen-nitrogen synthesis ammonia production unit, 200 kg / h CO2 capture-hydrogen-carbon coupling alcohol production unit; End-user energy demand: hydrogen refueling for hydrogen-powered vehicles, power supply from ammonia-fueled generator sets, and heating from methanol-fueled boilers within the service area, with a daily fluctuation range of ±30%; The method includes the following steps: S1. Collect and preprocess data from wind and solar power generation, energy storage systems, hydrogen production process units, ammonia production process units, alcohol production process units, and end-use energy systems; the specific steps of step S1 are as follows: S11. Collect real-time data of the wind and solar power generation system. The real-time data of the wind and solar power generation system includes the output power of the photovoltaic array, the output power of the wind turbine, and the ultra-short-term power prediction value for a future set time period (e.g., 5-60 minutes). For example, data is collected in real time through photovoltaic inverters and wind turbine controllers. The output power of the photovoltaic array (range 0-10MW) and the output power of the wind turbine (range 0-5MW) are collected once every 1 second. Combined with numerical weather forecasts and historical data, an LSTM neural network is used to predict the ultra-short-term power in the next 30 minutes, and the prediction error is controlled within ±8%. S12. Collect the status data of the energy storage system. The status data of the energy storage system includes the state of charge and charge / discharge power of electrochemical energy storage, as well as the hydrogen storage tank capacity and hydrogen inflow / outflow of hydrogen energy storage. For example, the state of charge (SOC, range 0-100%) and charge / discharge power (-20MW~+20MW, negative for charging, positive for discharging) of electrochemical energy storage are collected through the battery management system (BMS); the hydrogen storage tank capacity (0-1000kg, equivalent volume 0-11121Nm³ under standard conditions) and hydrogen inflow / outflow (-500Nm³ / h~+500Nm³ / h, equivalent mass flow rate -44.94kg / h~+44.94kg / h) are collected through the hydrogen energy storage monitoring system. The conversion standard is: hydrogen density 0.08988kg / Nm³ under standard conditions (0℃, 101325Pa); the data acquisition cycle is 2 seconds. S13. Collect real-time operating data of each process unit. The real-time operating data of each process unit includes the electrolysis current, voltage, electrolyte temperature and hydrogen production rate of the hydrogen production process unit, the reactor temperature, pressure, catalyst activity coefficient and ammonia production of the ammonia production process unit, and the CO2 capture, synthesis reaction temperature, pressure and methanol production of the methanol production process unit. For example, the hydrogen production process unit collects electrolysis current (0-5000A), voltage (0-800V), electrolyte temperature (60-85℃), and hydrogen production rate (0-500Nm³ / h, equivalent mass rate under standard conditions 0-44.94kg / h, conversion standard: hydrogen density under standard conditions 0.08988kg / Nm³) through the electrolyzer control system, with a collection cycle of 1 second; The ammonia production process unit collects temperature (350-450℃), pressure (10-15MPa), catalyst activity coefficient (0.8-1.0), and ammonia yield (0-300kg / h) through a reactor monitoring system, with a collection cycle of 2 seconds; The methanol production unit collects the CO2 collection rate (0-250 kg / h) through a CO2 capture device, and collects the reaction temperature (220-280℃), pressure (5-10 MPa), and methanol production (0-200 kg / h) through a synthesis reactor, with a collection cycle of 2 seconds; S14. Collect demand data from end-use energy systems, including the immediate demand for hydrogen, ammonia fuel, and methanol fuel for transportation. For example, the service area energy management platform collects real-time data on the immediate demand for hydrogen (0-300 Nm³ / h), ammonia fuel (0-200 kg / h), and methanol fuel (0-150 kg / h) for transportation, with a collection cycle of 5 seconds. S15. Preprocess the collected real-time data, status data, real-time running data and demand data. The preprocessing includes outlier detection and removal, missing value imputation and data standardization to generate a unified preprocessed running data stream. For example, outlier detection and removal: The 3σ criterion is used to identify outlier data that exceeds reasonable ranges such as photovoltaic power (0-10MW) and electrolysis temperature (60-85℃). For example, outlier values ​​of 12MW photovoltaic power are directly removed. Missing value imputation: For data with short-term missing values ​​(≤5 seconds), linear interpolation is used to impute them; for data with long-term missing values ​​(>5 seconds), the historical average of the same time period is used to replace them. Data standardization: Map all data to the [0,1] interval. For example, the hydrogen production rate of 500 Nm³ / h is standardized to 1, and 0 Nm³ / h is standardized to 0, generating a real-time running data stream in a unified format. S2. Input the preprocessed data into the pre-set digital twin model of the integrated wind-solar-hydrogen-ammonia-methanol system, and based on the digital twin model, construct and update a distributed collaborative optimization model with green electricity utilization rate, product output efficiency, and energy storage loss rate as optimization objectives, and energy balance, material balance, equipment operation, and safety as constraints; the specific steps of step S2 are as follows: S21. Map the preprocessed running data stream to the digital twin model of the integrated wind-solar-hydrogen-ammonia-ethanol system in real time, and update the state parameters of each virtual entity component in the digital twin model; For example, the pre-processed real-time data stream is transmitted to the digital twin platform via industrial Ethernet. The platform uses Unity3D to construct an integrated virtual scene of wind, solar, hydrogen, ammonia, and methanol, which maps the real-time status of physical entities such as the power generation status of photovoltaic panels, the operating parameters of electrolyzers, and the liquid level of hydrogen storage tanks. The update cycle is consistent with the data acquisition cycle (1-5 seconds), ensuring that the deviation between the virtual model and the physical system status is ≤2%. S22. Based on the real-time state reflected by the updated digital twin model, a multi-objective function F of the distributed collaborative optimization model is defined, wherein the multi-objective function F is expressed as:

[0025] in, To improve the utilization rate of green electricity, To improve overall product output efficiency, Energy storage loss rate, , , These are dynamic weighting coefficients; For example, based on the operational needs of the service area, dynamic weighting coefficients are set as follows: α=0.4 (weight corresponding to green electricity utilization rate), β=0.3 (weight corresponding to comprehensive product output efficiency), and γ=0.3 (weight corresponding to energy storage loss rate). A multi-objective function is then constructed using the formula:

[0026] Among them: green electricity utilization rate The calculation method is: actual total green electricity consumption / total wind and solar power generation, with a target value ≥ 90% (i.e., ≥0.9); Overall product output efficiency The calculation method is as follows: Target value ≥ 0.85; Among them, 12, 5, and 8 are the unit mass conversion coefficients (kgce / kg) of hydrogen, ammonia, and methanol, respectively, which conform to the industry's general conversion standard. Energy storage loss rate The calculation method is: energy loss of energy storage system / total energy input of energy storage, with a target value ≤ 5% (i.e., Laverage ≤ 0.05). S23. Load optimization objectives and constraints into the digital twin model; The constraints include at least the following: Real-time energy balance constraints between wind and solar power generation, energy storage charging and discharging power and energy consumption of each process unit; Input and output material balance constraints of hydrogen, nitrogen, and carbon elements in each production stage; Temperature and pressure safety boundary constraints for the operation of each process unit equipment; For example, the energy balance constraint is: wind and solar power generation power + energy storage discharge power = energy storage charging power + hydrogen production energy consumption + ammonia production energy consumption + alcohol production energy consumption, with an allowable deviation of ≤ ±5%; Material balance constraints: Hydrogen production = Hydrogen consumption for ammonia production + Hydrogen consumption for alcohol production + Change in hydrogen storage + Hydrogen consumption at the end; Nitrogen consumption for ammonia production = Nitrogen supply from air separator (Nitrogen-to-hydrogen ratio 2.8-3.2:1); Carbon consumption for alcohol production = CO2 capture (carbon-to-hydrogen ratio 1:3.8-4.2). Safety boundary constraints: Electrolyzer temperature ≤ 85℃, ammonia reactor pressure ≤ 15MPa, hydrogen storage tank pressure ≤ 35MPa, electrochemical energy storage SOC ≥ 10% and ≤ 90%; S3. Based on the distributed collaborative optimization model and the preprocessed data input in real time, the edge computing devices deployed locally in each process unit perform the local optimization solution for the corresponding process unit, and the cloud server performs the global optimization solution and coordination, outputting the optimal operating parameter settings for each process unit. The specific steps of step S3 are as follows: S31. The local edge computing devices of each process unit, based on the preprocessed data received in real time and the local constraints obtained from the digital twin model, solve the local optimization sub-problem with the goal of optimizing their own process efficiency, and generate preliminary operating parameters. For example, edge computing gateways (configured with Intel Core i7 processors and 8GB of memory) are deployed in the hydrogen production, ammonia production, and alcohol production process units respectively. Each edge device solves a local optimization sub-problem based on locally preprocessed data and constraints. Hydrogen production edge equipment: With the goal of achieving the optimal electrolysis efficiency (≥75%), the electrolysis current and voltage setpoints are solved. For example, when green electricity is sufficient, the electrolysis current is set to 4500A, the voltage to 750V, and the hydrogen production rate to 480Nm³ / h (equivalent mass 43.14kg / h, conversion standard: hydrogen density under standard conditions 0.08988kg / Nm³). Ammonia production peripheral equipment: With the goal of achieving the optimal ammonia synthesis efficiency (≥92%), solve for the reactor temperature and pressure setpoints, for example, set the temperature to 420℃, the pressure to 13MPa, and the ammonia production to 280kg / h; Methanol production edge equipment: With the goal of achieving the optimal methanol synthesis efficiency (≥88%), the set values ​​of reaction temperature, pressure and CO2 capture rate are solved. For example, the set temperature is 250℃, the pressure is 8MPa, the CO2 capture rate is 220kg / h, and the methanol production rate is 180kg / h. S32. The cloud server receives preliminary operating parameters and related status data uploaded by all edge computing devices; For example, each edge computing device uploads preliminary operating parameters and device status data to the cloud server every 10 seconds (configured with 2 Intel Xeon Gold 6330 processors, 128GB memory, and Redis for data caching). S33. The cloud server, guided by the global objective function of the distributed collaborative optimization model, coordinates the initial operating parameters of each process unit and solves the overall optimal solution of the wind-solar-hydrogen-ammonia-methanol integrated system through a global optimization algorithm. In step S33, the global optimization algorithm adopts a distributed model predictive control algorithm, and the specific steps are as follows: S331. The cloud server is based on a digital twin model to construct a global centralized prediction model that couples the dynamic relationships between wind and solar power generation, energy storage systems and various process units; S332. In each coordination cycle, using the real-time status data gathered from all edge computing devices as the initial value and the global objective function of the distributed collaborative optimization model as the performance index, within the first prediction time domain preset based on the wind and solar power generation fluctuation cycle and the dynamic response time of the ammonia / methanol production process, solve the constrained optimization problem to obtain the future control sequence that makes the wind-solar-hydrogen-ammonia-methanol integrated system optimal as a whole. S333. The control quantities of each unit at the first moment in the future control sequence are sent as the optimal operating parameter settings to the corresponding edge computing devices to enable cross-unit collaboration; For example, a distributed model predictive control algorithm is used: Construct a global prediction model: Based on a digital twin model, establish a coupled dynamic model of wind and solar power generation, energy storage, hydrogen production, ammonia production, and alcohol production, with a model error ≤3%; Set the first prediction time domain: Combining the fluctuation cycle of wind and solar power generation (30 minutes) and the dynamic response time of ammonia / methanol production process (15 minutes), the first prediction time domain is set to 30 minutes, with each 5 minutes as a coordination cycle; Find the global optimal solution: With the goal of maximizing F within 30 minutes, solve the control sequence for the next 30 minutes under constraints. For example, in a certain coordination period, due to a sudden increase in end-user hydrogen demand, adjust the hydrogen production rate to 500 Nm³ / h and reduce the ammonia production rate to 250 kg / h, ensuring a green electricity utilization rate of 92% and an energy storage loss rate of 4.2%. As another implementation of the global optimization algorithm, cloud servers can also use an improved particle swarm optimization algorithm to solve the problem; SS331. The cloud server uses the initial set of running parameters uploaded by each edge computing device as the initial position of the particle swarm, and the global objective function value as the fitness value; SS332. In each iteration, based on the individual historical best position of each particle i... The global historical best position of the particle swarm Update the velocity and position of each particle using the following formula:

[0027]

[0028] in, Let be the velocity of particle i at time t. The particle position (representing a possible combination of global operating parameters). This represents the historical best position of particle i. This represents the global historical best position of the particle swarm. For inertial weights, , As a learning factor, , A random number within the interval [0,1]; SS333. After updating the particle position, positions that violate energy balance, matter balance, or safety boundary constraints are corrected to within the feasible domain boundary. SS334. When the number of iterations reaches the preset maximum value, or the global historical best position of the particle swarm is reached. When the improvement of the corresponding fitness value in consecutive iterations is less than a threshold, the iteration stops, and the global historical best position of the particle swarm is recorded. The corresponding position is decoded into the optimal operating parameter setting value for each process unit; S34. The cloud server will send the optimal operating parameter settings of each process unit obtained from the solution to the corresponding edge computing devices; For example, the cloud server sends the optimal operating parameter settings for each process unit to the corresponding edge device with a sending delay of ≤2 seconds; Following step S34 in step S3, the following steps are also included: S35. If the fluctuation range of wind and solar power generation exceeds the preset threshold, or if the terminal demand changes abruptly, the cloud server will trigger the emergency recalculation process of the distributed collaborative optimization model and prioritize ensuring that the safety constraints are met. For example, when the power fluctuation of wind and solar power exceeds ±15% (e.g., the photovoltaic power drops suddenly from 8MW to 6MW), or when there is a sudden change in end-user demand (e.g., the demand for hydrogen fuel cell vehicles increases from 200Nm³ / h to 300Nm³ / h due to concentrated hydrogen refueling), the cloud immediately triggers an emergency recalculation to prioritize ensuring equipment safety constraints (e.g., the temperature of the electrolyzer does not exceed the limit and the pressure of the hydrogen storage tank is within the limit). The recalculation takes ≤3 seconds. S4. Send the optimal operating parameter settings to the corresponding process units for execution, collect actual operating data, and dynamically adjust the parameters of the distributed collaborative optimization model based on the deviation between the actual operating data and the expected target; The specific steps of step S4 are as follows: S41. The edge computing device converts the received optimal operating parameter settings into specific control commands and sends them to the actuators of the corresponding hydrogen production process unit, ammonia production process unit, or alcohol production process unit. For example, the edge computing device converts the optimal parameter settings into actuator instructions. For instance, the hydrogen production edge device sends a control signal of 4800A current and 760V voltage to the electrolyzer power supply; the ammonia production edge device sends a set value of 430℃ to the reactor temperature controller, and the actuator response time is ≤1 second. S42. Collect the actual operating data of each process unit after executing control commands in real time, and calculate the deviation between the actual operating data and the optimal operating parameter setting value; For example, the actual operating data of the process unit is collected in real time, and the deviation from the set value is calculated. For example, the actual hydrogen production rate is 470 Nm³ / h, and the deviation from the set value of 480 Nm³ / h is -10 Nm³ / h; the actual ammonia production pressure is 12.8 MPa, and the deviation from the set value of 13 MPa is -0.2 MPa. S43. Based on the aforementioned deviation, a model predictive control algorithm is used to dynamically adjust the local model parameters or dynamic weight coefficients of the corresponding process unit in the distributed collaborative optimization model. , , To achieve closed-loop feedback optimization; In step S43, the execution process of the model predictive control algorithm includes the following steps: S431. The edge computing device constructs a local simplified prediction model based on the sub-model corresponding to this process unit in the digital twin model; S432. In each control cycle, with the actual operating data collected by this process unit and the set value received from the cloud server as input, and with the goal of quickly and accurately tracking the set value, an optimization problem with local constraints is solved in a second prediction time domain shorter than the first prediction time domain to obtain the optimal control command for this process unit. S433. The optimal control command obtained by the solution is sent to the actuator of this process unit, and the optimization is continuously updated based on the new data after execution to achieve closed-loop elimination of set value deviation; For example, a model predictive control algorithm is used: Constructing a simplified local prediction model: Based on a digital twin model, a simplified model of electrolysis current-hydrogen production rate is established for hydrogen production edge devices, with a model response time ≤ 0.5 seconds; Set a second prediction time domain: Set the second prediction time domain to 5 minutes (30 minutes shorter than the first prediction time domain), and solve the local optimization problem once every 1 minute; Rolling updates and optimizations: For the deviation of -10 Nm³ / h in hydrogen production rate, the electrolysis current was adjusted to 4850 A, so that the actual hydrogen production rate gradually approaches 480 Nm³ / h, and the deviation elimination time is ≤3 minutes. As another way to implement dynamic adjustment algorithms, edge computing devices can also use proportional-integral-derivative control algorithms for local closed-loop control. The execution process of the proportional-integral-derivative control algorithm includes the following steps: SS431. Edge computing devices calculate actual operating data. (e.g., actual hydrogen production rate) and optimal operating parameter settings Real-time deviation between (e.g., setting hydrogen production rate) ; SS432. Calculate the control quantity based on the proportional, integral, and derivative terms. The calculation formula is:

[0029] in, , , These are the proportional, integral, and derivative coefficients, which are pre-tuned or adjusted online according to the characteristics of the process unit. SS433. The calculated control quantity The commands are converted into specific actuator instructions (such as adjusting the current setpoint of the electrolytic cell power supply) and sent to the corresponding process unit to achieve rapid correction of deviations. S5. Visualize the optimization process data, solution results, and operational status of the integrated wind-solar-hydrogen-ammonia-methanol system; the specific steps of step S5 are as follows: S51. In the visualization interface, the real-time running status of each virtual component in the digital twin model, the key indicator curves of the preprocessed running data stream, and the solution process and results of the distributed collaborative optimization model are dynamically displayed. For example, the visualization interface developed using WebGL technology in the service area control center dynamically displays the operating status of virtual components such as photovoltaic arrays, electrolyzers, and hydrogen storage tanks in the digital twin model, and plots trend charts of key indicators such as green electricity power curves and hydrogen production rate curves, with an update cycle of 1 second; S52. Show the comprehensive energy efficiency indicators of the integrated wind-solar-hydrogen-ammonia-methanol system. The comprehensive energy efficiency indicators include real-time green electricity consumption rate, hydrogen / ammonia / methanol production rate and equivalent carbon emission reduction. For example, the comprehensive energy efficiency indicators are displayed in real time: green electricity consumption rate 92.5%, hydrogen production rate 490 Nm³ / h (equivalent mass 44.04 kg / h, conversion standard: hydrogen density under standard conditions 0.08988 kg / Nm³), ammonia production rate 275 kg / h, methanol production rate 178 kg / h, and equivalent carbon emission reduction 1200 kg / day; S53. Through the human-machine interface, respond to operators' requests to view the basis for optimization decisions, and allow operators to confirm or manually fine-tune the optimal operating parameter settings; For example, operators can view the optimization decision basis (such as the calculation process of prioritizing increasing hydrogen production load when green electricity is sufficient) through the interface, and can manually fine-tune the optimal parameters within a range of ±5%, such as adjusting the temperature of the ammonia production reactor from 430℃ to 425℃. The execution effect is fed back in real time after the adjustment command is issued.

[0030] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0031] like Figure 2 As shown, the following is an embodiment of the distributed data optimization device for the production process of wind-solar-hydrogen amine alcohol provided in this disclosure. This device and the distributed data optimization method for the production process of wind-solar-hydrogen amine alcohol in the above embodiments belong to the same inventive concept. For details not described in detail in the embodiments of the distributed data optimization device for the production process of wind-solar-hydrogen amine alcohol, please refer to the embodiments of the distributed data optimization method for the production process of wind-solar-hydrogen amine alcohol described above.

[0032] The device includes: The data acquisition and preprocessing module is used to collect and preprocess data from wind and solar power generation, energy storage systems, hydrogen production process units, ammonia production process units, alcohol production process units, and end-use energy systems. The digital twin and optimization modeling module is used to input pre-processed data into a pre-set digital twin model of the wind-solar-hydrogen-ammonia-ethanol integrated system, and based on the digital twin model, with green electricity utilization rate, product output efficiency and energy storage loss rate as optimization objectives, and with energy balance, material balance, equipment operation and safety as constraints, to construct and update a distributed collaborative optimization model. The distributed collaborative optimization solution module includes: Edge computing devices deployed locally in each process unit perform local optimization solutions for the corresponding process unit based on a distributed collaborative optimization model and preprocessed data input in real time. The cloud server performs global optimization and coordination, and outputs the optimal operating parameter settings for each process unit. The dynamic feedback and execution control module is used to send the optimal operating parameter settings to the corresponding process units for execution, and to collect actual operating data. Based on the deviation between the actual operating data and the expected target, the parameters of the distributed collaborative optimization model are dynamically adjusted. The visualization and human-computer interaction module is used to visualize and display the optimization process data, solution results, and operating status of the integrated wind-solar-hydrogen-ammonia-methanol system.

[0033] This embodiment achieves distributed collaborative optimization of the wind-solar hydrogen ammonia-methanol production process through the interactive collaboration of data acquisition and preprocessing modules, digital twin and optimization modeling modules, distributed collaborative optimization solution modules, dynamic feedback and execution control modules, and visualization and human-computer interaction modules, thereby improving the utilization rate of green electricity and enhancing the system's flexibility and stability.

[0034] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A distributed data optimization method for a wind-solar-hydrogen-ammonia- alcohol production process, characterized in that, Includes the following steps: S1. Collect and preprocess data from wind and solar power generation, energy storage systems, hydrogen production process units, ammonia production process units, alcohol production process units, and end-use energy systems. S2. Input the preprocessed data into the pre-set digital twin model of the wind-solar-hydrogen-ammonia-ethanol integrated system, and based on the digital twin model, construct and update the distributed collaborative optimization model with green electricity utilization rate, product output efficiency and energy storage loss rate as optimization objectives, and energy balance, material balance, equipment operation and safety as constraints. S3. Based on the distributed collaborative optimization model and the preprocessed data input in real time, the edge computing devices deployed locally in each process unit perform the local optimization solution for the corresponding process unit, and the cloud server performs the global optimization solution and coordination, outputting the optimal operating parameter settings for each process unit. S4. Send the optimal operating parameter settings to the corresponding process units for execution, collect actual operating data, and dynamically adjust the parameters of the distributed collaborative optimization model based on the deviation between the actual operating data and the expected target; S5. Visualize the optimization process data, solution results, and operating status of the integrated wind-solar-hydrogen-ammonia-ethanol system.

2. The distributed data optimization method for the wind-solar-hydrogen-ammonia-methanol production process according to claim 1, characterized in that, The specific steps of step S1 are as follows: S11. Collect real-time data of the wind and solar power generation system. The real-time data of the wind and solar power generation system includes the output power of the photovoltaic array, the output power of the wind turbine, and the ultra-short-term power prediction value for a future set time period. S12. Collect the status data of the energy storage system. The status data of the energy storage system includes the state of charge and charge / discharge power of electrochemical energy storage, as well as the hydrogen storage tank capacity and hydrogen inflow / outflow of hydrogen energy storage. S13. Collect real-time operating data of each process unit. The real-time operating data of each process unit includes the electrolysis current, voltage, electrolyte temperature and hydrogen production rate of the hydrogen production process unit, the reactor temperature, pressure, catalyst activity coefficient and ammonia production of the ammonia production process unit, and the CO2 capture, synthesis reaction temperature, pressure and methanol production of the methanol production process unit. S14. Collect demand data from end-use energy systems, including the immediate demand for hydrogen, ammonia fuel, and methanol fuel for transportation. S15. Preprocess the collected real-time data, status data, real-time running data, and demand data. The preprocessing includes outlier detection and removal, missing value imputation, and data standardization to generate a unified preprocessed running data stream.

3. The distributed data optimization method for the wind-solar-hydrogen-ammonia-methanol production process according to claim 2, characterized in that, The specific steps of step S2 are as follows: S21. Map the preprocessed running data stream to the digital twin model of the integrated wind-solar-hydrogen-ammonia-ethanol system in real time, and update the state parameters of each virtual entity component in the digital twin model; S22. Based on the real-time state reflected by the updated digital twin model, a multi-objective function F of the distributed collaborative optimization model is defined, wherein the multi-objective function F is expressed as: in, To improve the utilization rate of green electricity, To improve overall product output efficiency, Energy storage loss rate, , , These are dynamic weighting coefficients; S23. Load optimization objectives and constraints into the digital twin model; The constraints include at least the following: Real-time energy balance constraints between wind and solar power generation, energy storage charging and discharging power and energy consumption of each process unit; Input and output material balance constraints of hydrogen, nitrogen, and carbon elements in each production stage; Temperature and pressure safety boundary constraints for the operation of each process unit equipment.

4. The distributed data optimization method for the wind-solar-hydrogen-ammonia-methanol production process according to claim 1, characterized in that, The specific steps of step S3 are as follows: S31. The local edge computing devices of each process unit, based on the preprocessed data received in real time and the local constraints obtained from the digital twin model, solve the local optimization sub-problem with the goal of optimizing their own process efficiency, and generate preliminary operating parameters. S32. The cloud server receives preliminary operating parameters and related status data uploaded by all edge computing devices; S33. The cloud server, guided by the global objective function of the distributed collaborative optimization model, coordinates the initial operating parameters of each process unit and solves the overall optimal solution of the wind-solar-hydrogen-ammonia-methanol integrated system through a global optimization algorithm. S34. The cloud server will send the optimal operating parameter settings of each process unit obtained from the solution to the corresponding edge computing devices.

5. The distributed data optimization method for the wind-solar-hydrogen-ammonia-methanol production process according to claim 4, characterized in that, In step S33, the global optimization algorithm adopts a distributed model predictive control algorithm, and the specific steps are as follows: S331. The cloud server is based on a digital twin model to construct a global centralized prediction model that couples the dynamic relationships between wind and solar power generation, energy storage systems and various process units; S332. In each coordination cycle, using the real-time status data gathered from all edge computing devices as the initial value and the global objective function of the distributed collaborative optimization model as the performance index, within the first prediction time domain preset based on the wind and solar power generation fluctuation cycle and the dynamic response time of the ammonia / methanol production process, solve the constrained optimization problem to obtain the future control sequence that makes the wind-solar-hydrogen-ammonia-methanol integrated system optimal as a whole. S333. The control quantities of each unit at the first moment in the future control sequence are sent as the optimal operating parameter settings to the corresponding edge computing devices for cross-unit collaboration.

6. The distributed data optimization method for the wind-solar-hydrogen-ammonia-methanol production process according to claim 4, characterized in that, Following step S34 in step S3, the following steps are also included: S35. If the fluctuation range of wind and solar power generation exceeds the preset threshold, or if there is a sudden change in end-user demand, the cloud server will trigger an emergency recalculation process of the distributed collaborative optimization model and prioritize ensuring that safety constraints are met.

7. The distributed data optimization method for the wind-solar-hydrogen-ammonia-methanol production process according to claim 5, characterized in that, The specific steps of step S4 are as follows: S41. The edge computing device converts the received optimal operating parameter settings into specific control commands and sends them to the actuators of the corresponding hydrogen production process unit, ammonia production process unit, or alcohol production process unit. S42. Collect the actual operating data of each process unit after executing control commands in real time, and calculate the deviation between the actual operating data and the optimal operating parameter setting value; S43. Based on the aforementioned deviation, a model predictive control algorithm is used to dynamically adjust the local model parameters or dynamic weight coefficients of the corresponding process unit in the distributed collaborative optimization model. , , This is to achieve closed-loop feedback optimization.

8. The distributed data optimization method for the wind-solar-hydrogen-ammonia-methanol production process according to claim 7, characterized in that, In step S43, the execution process of the model predictive control algorithm includes the following steps: S431. The edge computing device constructs a local simplified prediction model based on the sub-model corresponding to this process unit in the digital twin model; S432. In each control cycle, with the actual operating data collected by this process unit and the set value received from the cloud server as input, and with the goal of quickly and accurately tracking the set value, an optimization problem with local constraints is solved in a second prediction time domain shorter than the first prediction time domain to obtain the optimal control command for this process unit. S433. The optimal control command obtained from the solution is sent to the actuator of this process unit, and the optimization is continuously updated based on the new data after execution to achieve closed-loop elimination of set value deviation.

9. The distributed data optimization method for the wind-solar-hydrogen-ammonia-methanol production process according to claim 2, characterized in that, The specific steps of step S5 are as follows: S51. In the visualization interface, the real-time running status of each virtual component in the digital twin model, the key indicator curves of the preprocessed running data stream, and the solution process and results of the distributed collaborative optimization model are dynamically displayed. S52. Show the comprehensive energy efficiency indicators of the integrated wind-solar-hydrogen-ammonia-methanol system. The comprehensive energy efficiency indicators include real-time green electricity consumption rate, hydrogen / ammonia / methanol production rate and equivalent carbon emission reduction. S53. Through the human-computer interaction interface, respond to the operator's request to view the basis for optimization decisions, and allow the operator to confirm or manually fine-tune the optimal operating parameter settings.

10. A distributed data optimization device for the production process of hydrogen ammonia and methanol from wind and solar power, characterized in that, include: The data acquisition and preprocessing module is used to collect and preprocess data from wind and solar power generation, energy storage systems, hydrogen production process units, ammonia production process units, alcohol production process units, and end-use energy systems. The digital twin and optimization modeling module is used to input pre-processed data into a pre-set digital twin model of the wind-solar-hydrogen-ammonia-ethanol integrated system, and based on the digital twin model, with green electricity utilization rate, product output efficiency and energy storage loss rate as optimization objectives, and with energy balance, material balance, equipment operation and safety as constraints, to construct and update a distributed collaborative optimization model. The distributed collaborative optimization solution module includes: Edge computing devices deployed locally in each process unit perform local optimization solutions for the corresponding process unit based on a distributed collaborative optimization model and preprocessed data input in real time. The cloud server performs global optimization and coordination, and outputs the optimal operating parameter settings for each process unit. The dynamic feedback and execution control module is used to send the optimal operating parameter settings to the corresponding process units for execution, and to collect actual operating data. Based on the deviation between the actual operating data and the expected target, the parameters of the distributed collaborative optimization model are dynamically adjusted. The visualization and human-computer interaction module is used to visualize and display the optimization process data, solution results, and operating status of the integrated wind-solar-hydrogen-ammonia-methanol system.

Images