Hydrogen-alcohol-ammonia collaborative demand prediction method, system and equipment based on macroscopic-microscopic double-layer framework and industrial chain conduction, and storage medium

Abstract

The invention discloses a hydrogen-alcohol-ammonia collaborative demand prediction method, system and device based on a macroscopic-microscopic double-layer framework and industrial chain conduction and a storage medium, and relates to the field of energy and chemical industry data analysis and artificial intelligence prediction, and the method comprises the steps: collecting and integrating multi-dimensional heterogeneous data reflecting the upstream and downstream relationship of a hydrogen-alcohol-ammonia industrial chain; constructing a feature data set, constructing a hydrogen demand double-layer prediction model, and obtaining a reference hydrogen demand prediction amount; establishing a hydrogen resource dynamic allocation model based on market and policy factors, and calculating the amount of hydrogen allocated to downstream alcohol and ammonia production links according to the reference hydrogen demand prediction amount; constructing an industrial chain conduction model, and deducing final demanded quantities of alcohol and ammonia according to the distributed hydrogen quantity in combination with a stoichiometric relationship and production efficiency; according to the method, the hydrogen alcohol ammonia demand prediction precision and consistency can be remarkably improved, the defect of independent prediction is overcome, the market dynamic response capability is enhanced, and more scientific and reliable data support is provided for decision making.

Description

Technical Field

[0001] This invention relates to the field of energy and chemical industry data analysis and artificial intelligence prediction, and in particular to a method, system, equipment and storage medium for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission. Background Technology

[0002] In recent years, the strategic importance of new energy carriers and green chemical raw materials, represented by hydrogen, alcohols (methanol / ethanol), and ammonia, has become increasingly prominent. Hydrogen, alcohols, and ammonia are closely coupled in the industrial chain: hydrogen is the core raw material for the synthesis of ammonia and green methanol, while all three play important roles in transportation, industry, and agriculture. Accurately predicting the market demand for hydrogen, alcohols, and ammonia is crucial for ensuring energy security, optimizing industrial layout, and guiding capital investment.

[0003] Currently, demand forecasting methods for such chemical products mainly suffer from the following problems: First, independent forecasting ignores coupling. Traditional methods typically treat hydrogen, alcohols, and ammonia as isolated products for forecasting, neglecting their rigid transformation relationships and supply-demand linkages within the industrial chain. The inherent flaw of this forecasting approach is that it cannot guarantee logical consistency and material balance between upstream and downstream products in the forecast results, making it difficult to accurately reflect the demand transmission relationships within the industrial chain determined by rigid chemical processes. Second, the models are simplistic and lack adaptability. Many methods employ traditional econometric models or simple time series models, making it difficult to capture the complex impacts of nonlinear, high-impact factors such as policy changes, technological innovations, and drastic energy price fluctuations on the market. These models, relying on extrapolation from historical data, exhibit significant forecast lag and insufficient accuracy when facing structural market changes because they cannot effectively internalize and quantify the impact of key exogenous variables. Third, the perspective is limited and lacks hierarchy. Existing research often focuses only on long-term trend judgments at the macro level or short-term sales forecasts at the micro level, failing to effectively combine macro strategies with micro-level operational dynamics. This disconnect in research perspectives results in long-term strategic planning lacking support from micro-market realities, while short-term operational management lacks guidance from macro-strategic directions, making it difficult for forecast results to simultaneously meet the dual needs of strategic planning and operational scheduling. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention provides a method, system, device and storage medium for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission.

[0005] Therefore, the technical problem solved by this invention is the low prediction accuracy and poor decision support capability in the demand forecasting of hydrogen alcohols and ammonia due to independent models, neglect of industrial chain coupling, and lack of macro-micro integration.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, this invention provides a method for predicting the synergistic demand for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, including: Collect and integrate multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain, and construct a feature dataset by dividing it according to geographical and time dimensions; A two-layer hydrogen demand prediction model is constructed based on the feature dataset to obtain the baseline hydrogen demand prediction. Establish a dynamic hydrogen resource allocation model based on market and policy factors, and calculate the amount of hydrogen allocated to downstream alcohol and ammonia production based on the baseline hydrogen demand forecast. A supply chain transmission model was constructed. Based on the allocated hydrogen quantity, combined with stoichiometry and production efficiency, the final demand for alcohols and ammonia was derived. An evaluation index system was established to verify the prediction results.

[0007] As a preferred approach to a synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, the following is provided: The collection and integration of multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain, divided by geographical and temporal dimensions, constructs a feature dataset including: Based on the collected multidimensional heterogeneous data, at least one cross-domain coupling feature is created to quantify the coupling relationship of the industrial chain. The creation logic of the cross-domain coupling feature is to perform quantitative calculation by combining the correlation relationship of different types of data, which includes at least an index for measuring the relative competitiveness of green hydrogen at the economic level and an index for measuring the support of regional policies for the industrial chain.

[0008] As a preferred approach to a synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, the following is provided: The collection and integration of multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain, divided by geographical and temporal dimensions, to construct a feature dataset also includes: The index measuring the competitiveness of the green hydrogen economy is calculated by comparing the sum of cost data related to traditional energy with data related to renewable energy and hydrogen production efficiency; the index measuring the strength of regional policy support is calculated by weighting and summing the quantitative scores of various support policies in the region.

[0009] As a preferred approach to a synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, the following is provided: The method for constructing a two-layer hydrogen demand prediction model based on a feature dataset to obtain a baseline hydrogen demand prediction includes: A two-level hydrogen demand forecasting model is constructed, which includes both macro-trend forecasting and micro-dynamic forecasting. The macro-trend forecasting is carried out using the system dynamics method, and a causal feedback loop model containing multiple subsystems is established. Each subsystem affects hydrogen demand based on its different internal influencing factors and their interrelated paths. Micro-dynamic prediction is performed using a data-driven regression model; A baseline hydrogen demand forecast was obtained using a two-layer hydrogen demand forecasting model.

[0010] The beneficial effects of this preferred technical solution are as follows: By employing a macro-micro dual-level prediction model, combining system dynamics methods for macro-trend prediction and data-driven regression models for micro-dynamic prediction, hydrogen demand can be analyzed and predicted from different levels and perspectives. The causal feedback loop model of system dynamics can comprehensively consider the interactions between multiple subsystems and their influencing factors, grasping the macro-trend of hydrogen demand; the data-driven regression model can capture dynamic changes at the micro level, improving the accuracy and flexibility of prediction.

[0011] As a preferred approach to a synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, the following is provided: The construction of the industrial chain transmission model, based on the allocated hydrogen quantity and combining stoichiometry and production efficiency, derives the final demand for alcohols and ammonia, and establishes an evaluation index system to verify the prediction results, including: A mathematical model for hydrogen-alcohol-ammonia conversion was established. Based on the amount of hydrogen allocated to the production of alcohol and ammonia, and combined with the unit product hydrogen consumption coefficient determined by the chemical reaction equation and the production process efficiency coefficient, the output of alcohol and ammonia was derived.

[0012] The beneficial effects of this preferred technical solution are as follows: by establishing a mathematical model for hydrogen-alcohol-ammonia conversion, and utilizing stoichiometry and production efficiency coefficients, the production of alcohols and ammonia can be accurately derived based on the allocated amount of hydrogen, thus realizing the effective transmission from hydrogen allocation to alcohol and ammonia production, and providing a scientific and reasonable calculation basis for demand forecasting in the industrial chain.

[0013] As a preferred approach to a synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, the following is provided: The construction of the industrial chain transmission model, which derives the final demand for alcohols and ammonia based on the allocated hydrogen quantity, combined with stoichiometry and production efficiency, and establishes an evaluation index system to verify the prediction results, also includes: Establish an evaluation index system to measure the accuracy of predictions by comparing the differences between model predictions and actual values; By associating the prediction results with spatiotemporal labels, a multi-dimensional, multi-scenario demand prediction database is constructed. Each prediction value in the database is determined by a multivariate function, which combines product labels, time labels, spatial labels, and scenario labels to clarify the specific prediction situation.

[0014] The beneficial effects of this preferred technical solution are as follows: Establishing an evaluation index system can quantitatively assess the accuracy of the prediction results, which helps to identify and improve problems in the prediction model in a timely manner. Linking the prediction results with spatiotemporal labels to construct a multi-dimensional, multi-scenario demand prediction database can provide comprehensive and detailed information support for demand prediction under different products, times, regions, and scenarios, facilitating multi-dimensional analysis and decision-making.

[0015] As a preferred approach to a synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, the following is provided: The construction of the industrial chain transmission model, which derives the final demand for alcohols and ammonia based on the allocated hydrogen quantity, combined with stoichiometry and production efficiency, and establishes an evaluation index system to verify the prediction results, also includes: Scenario analysis is conducted based on the forecast results, and the impact of changes in different policies or market variables on the future demand for hydrogen alcohols and ammonia is quantitatively simulated by calculating the demand elasticity coefficient.

[0016] Secondly, this invention provides a synergistic demand forecasting system for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, comprising: The data acquisition and integration module and the feature dataset construction module are used to collect and integrate multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain, and construct feature datasets by dividing them according to geographical and time dimensions. The hydrogen demand two-layer forecasting module is used to build a hydrogen demand two-layer forecasting model based on the feature dataset and obtain the baseline hydrogen demand forecast. The hydrogen resource dynamic allocation module is used to establish a hydrogen resource dynamic allocation model based on market and policy factors, and calculate the amount of hydrogen allocated to downstream alcohol and ammonia production based on the baseline hydrogen demand forecast. The supply chain transmission and result verification module is used to construct a supply chain transmission model. Based on the allocated hydrogen quantity, combined with stoichiometry and production efficiency, it derives the final demand for alcohols and ammonia, and establishes an evaluation index system to verify the prediction results.

[0017] Thirdly, the present invention provides a computer device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the hydrogen alcohol and ammonia synergistic demand forecasting method based on a macro-micro dual-layer framework and industrial chain transmission are implemented.

[0018] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of a method for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission.

[0019] The beneficial effects of this invention are as follows: Addressing the problems of independent forecasting models, neglect of supply chain coupling, and failure to effectively combine macro and micro perspectives in existing methods for predicting demand for alcohols, methanols, and ammonia, this invention constructs a macro-micro dual-layer forecasting framework and introduces a supply chain transmission mechanism, significantly improving the internal logical consistency of the forecast results and their responsiveness to market dynamics. Compared with traditional independent forecasting models and single econometric models, the method of this invention has significant advantages in forecasting accuracy and decision support capabilities. It can better reflect the inherent coupling relationship of the alcohol, methanol, and ammonia supply chain, nonlinear market dynamics, and multi-level decision-making needs. It is not only applicable to the analysis of alcohol, methanol, and ammonia demand in a single region, but can also be extended to the analysis of other related energy and chemical products in different countries and market environments, demonstrating good adaptability and scalability. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is an overall flowchart of a method for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, provided by the present invention. Detailed Implementation

[0022] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0023] Example 1, referring to Figure 1This is the first embodiment of the present invention, which provides a method for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, including: S1: Collect and integrate multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain, divide it by geographical and time dimensions, and construct a feature dataset; S2: Construct a two-layer hydrogen demand prediction model based on the feature dataset to obtain the baseline hydrogen demand forecast; S3: Establish a dynamic hydrogen resource allocation model based on market and policy factors, and calculate the amount of hydrogen allocated to downstream alcohol and ammonia production based on the baseline hydrogen demand forecast. S4: Construct a supply chain transmission model, derive the final demand for alcohols and ammonia based on the allocated hydrogen quantity, combined with stoichiometry and production efficiency, and establish an evaluation index system to verify the prediction results.

[0024] It should be noted that through steps S1-S4, by systematically integrating macro and micro forecasts, dynamic resource allocation, and industrial chain transmission mechanisms, coordinated and high-precision forecasts of the demand for hydrogen, alcohols, and ammonia are achieved, significantly improving the logical consistency, market responsiveness, and decision support value of the forecast results.

[0025] Example 2, refer to Figure 1 As an embodiment of the present invention, based on the previous embodiment, a method for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission is provided, including: In this embodiment, the multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain collected and integrated in step S1 above, and the feature dataset constructed by dividing it according to geographical and time dimensions, includes: Collect multidimensional heterogeneous data, including at least: macroeconomic data, industrial chain data, policy and regulatory data, and historical production and sales data; Based on the collected multidimensional heterogeneous data, at least one cross-domain coupling feature is created to quantify the coupling relationship in the industrial chain; the cross-domain coupling feature includes at least the Green Hydrogen Competitiveness Index (GHCI), calculated as follows: In the formula, P gas Price of natural gas per unit of calorific value; P carbon The carbon trading price per unit of carbon emissions; P re For renewable energy feed-in tariffs; or elec The efficiency of hydrogen production in the electrolyzer.

[0026] Cross-domain coupling characteristics also include the Regional Policy Support Index (RPSI), calculated as follows: In the formula, S k For the first k The quantitative score of each support policy; w k This represents the weighting coefficient for this policy. N This represents the total number of relevant policies within the region.

[0027] In another possible implementation, when creating cross-domain coupling features for quantifying the coupling relationships within a supply chain, complex network theory can be used as a basis. This involves treating each link in the supply chain as a node in a network, and the relationships between links as edges, thus constructing a supply chain network model. The coupling relationships within the supply chain can be quantified by calculating network topology indicators such as node degree, clustering coefficient, and shortest path length. Furthermore, combining data on logistics, information flow, and capital flow within the supply chain further enriches the connotation of the cross-domain coupling features.

[0028] In another possible implementation, when creating cross-domain coupling features for quantifying supply chain coupling relationships, multivariate statistical methods such as principal component analysis and factor analysis can be used to reduce the dimensionality of the collected multidimensional heterogeneous data and extract key features. The extracted features are then combined and transformed to create cross-domain coupling features that reflect supply chain coupling relationships. The effectiveness and stability of these cross-domain coupling features are then verified by analyzing supply chain data from different periods and regions.

[0029] In another possible implementation, when collecting and integrating multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain, it is also possible to cooperate with authoritative data providers in the industry to obtain detailed data covering all aspects of the hydrogen alcohol and ammonia industry chain, including production output and cost data of production enterprises, logistics data of transportation, and market sales data. At the same time, unstructured data, such as expert opinions and market dynamics, can be collected from relevant industry websites, forums, news media, and other channels using web crawling technology, and then cleaned and integrated.

[0030] In another possible implementation, when collecting and integrating multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain, a data collection platform can be established to encourage companies in the industry chain to actively upload data, providing incentives such as priority in data sharing and industry reports. Publicly available data released by government departments, such as energy policies and statistical data, should be collected and organized regularly. Valuable information should be extracted from massive amounts of data and integrated using data mining techniques.

[0031] In this embodiment, step S2 above, which involves constructing a two-layer hydrogen demand prediction model based on a feature dataset to obtain a baseline hydrogen demand prediction, includes: The two-layer prediction model for hydrogen demand includes: macro-trend prediction through a system dynamics model, and micro-dynamic prediction through a data-driven regression model.

[0032] Specifically, macroeconomic trend forecasting employs system dynamics methods to establish causal feedback loop models that include at least multiple subsystems such as transportation, chemical production, and port logistics. The main influencing factors of the transportation subsystem are as follows: 1. Population + Transportation industry output value → Bus ownership → Number of hydrogen fuel cell vehicles → Hydrogen demand for transportation; 2. Freight volume + transportation industry output value → heavy truck ownership → hydrogen fuel cell vehicle ownership → hydrogen demand for transportation; 3. Policy support + comprehensive benefits + number of buses + number of heavy trucks → number of hydrogen fuel cell vehicles; 4. Unit hydrogen consumption + driving mileage → annual hydrogen consumption → hydrogen demand for transportation.

[0033] The main influencing factors of the port logistics subsystem are as follows: 1. Port cargo throughput + container traffic volume → port machinery inventory → demand for hydrogen-powered forklifts → demand for hydrogen in ports; 2. Green port policy + environmental protection requirements → clean energy substitution rate → hydrogen fuel cell application → port hydrogen demand; 3. Logistics park construction + warehouse area → + hydrogen fuel cell logistics vehicles → + demand for hydrogen in logistics; 4. Overall benefits → + Proportion of hydrogen energy equipment → + Hydrogen demand in ports.

[0034] The main influencing factors of the ammonia synthesis production subsystem are as follows: 1. Crop planting area → + Crop yield → + Ammonia synthesis demand → + Crop yield per unit area → + Crop output; 2. GDP → + Agricultural output value → + Crop yield per unit area → + Crop output; 3. Last year's demand for synthetic ammonia → + crop yield per unit area → + crop output; 4. Fertilizer policy factors → Ammonia synthesis demand → Crop yield per unit area → Crop output.

[0035] The main influencing factors of the methanol synthesis production subsystem are as follows: 1. Ethylene production → + Methanol synthesis production → + Methanol synthesis hydrogen demand; 2. Pesticide production → + Synthetic methanol production → + Hydrogen demand for synthetic methanol; 3. Crop planting area → Grain output → Pesticide output → Methanol synthesis output → Methanol synthesis hydrogen demand; 4. Methanol production → pesticide production → grain production → agricultural output value.

[0036] The main influencing factors of the petrochemical refining subsystem are as follows: 1. Regional refining capacity utilization rate + crude oil processing volume → + refined oil production → + hydrorefining demand → + petrochemical hydrogen demand; 2. Regional chemical production load + chemical product demand → + deep processing capacity → + hydrocracking demand → + petrochemical hydrogen demand; 3. Industrial park construction + industrial chain extension → fine chemical industry development → industrial hydrogen demand → petrochemical hydrogen demand.

[0037] In another possible implementation, fuzzy mathematics and grey system theory can be introduced when forecasting macroeconomic trends to handle uncertainties and fuzzy information in the model. For example, for factors that are difficult to quantify precisely, such as the degree of policy impact and market expectations, fuzzy language can be used to describe them, and methods such as fuzzy inference and grey prediction can be applied. Simultaneously, Monte Carlo simulation techniques can be combined to run multiple simulations of the model, analyzing the uncertainties and risks of macroeconomic trends.

[0038] In another possible implementation, when conducting macroeconomic trend forecasting, system dynamics software (such as Vensim, Stella, etc.) can be used to construct macroeconomic trend forecasting models and combined with Geographic Information Systems (GIS). The geographical distribution information of the industrial chain is integrated into the model to analyze the development trends and mutual influences of the industrial chain in different regions. Through the spatial analysis functions of GIS, the spatial distribution characteristics of macroeconomic trends can be intuitively displayed, providing decision support for regional energy planning and policy formulation.

[0039] Furthermore, the data-driven regression model is the Support Vector Regression (SVR) model, and the regression function is expressed as: In the formula, f ( x ) is the prediction function; x i The input feature vector; b For deviation; α* i and α i It is a Lagrange multiplier.

[0040] The Support Vector Regression (SVR) model uses a composite kernel function. K comp ( x i, x j This composite kernel function consists of a polynomial kernel function used to capture global data trends. K poly With the Gaussian radial basis kernel function used to capture local nonlinear relationships K rbf The weighted composition is calculated using the following formula: In the formula, λ is a weighting coefficient between [0,1], which is used to assess the ability of the adaptive equilibrium model to fit the global trend and local nonlinear relationship.

[0041] The hyperparameters of the Support Vector Regression (SVR) model are determined adaptively using an improved particle swarm optimization (PSO) algorithm with adaptive inertia weights. To address the difficulty in selecting SVR hyperparameters, this invention proposes a linearly decreasing adaptive inertia weight strategy, improving the standard PSO algorithm. The inertia weight w(t) is dynamically adjusted with the iteration number t, and its calculation formula is as follows: In the formula, w ( t ) is the first t The inertia weight of the next iteration; w max and w min These are the maximum and minimum values ​​of the inertia weight, respectively. T max This represents the maximum number of iterations.

[0042] In this embodiment, the hydrogen resource dynamic allocation model established in step S3 above, based on market and policy factors, calculates the amount of hydrogen allocated to downstream alcohol and ammonia production stages according to the baseline hydrogen demand forecast, including: To establish a dynamic allocation model for hydrogen resources, dynamic allocation coefficients were constructed. β i,y The calculation formula is as follows: In the formula, β i,y For products i exist y Hydrogen allocation coefficient for the year; w i For products i The priority weight of industrial policies; P i,y For products i Price index; D i,y For products iThe market demand index.

[0043] In this embodiment, step S4 above involves constructing a supply chain transmission model. Based on the allocated hydrogen quantity, combined with stoichiometry and production efficiency, the final demand for alcohols and ammonia is derived. An evaluation index system is then established to verify the prediction results, including: A mathematical model for the hydrogen-ethanol-ammonia conversion is established, and the conversion formula is as follows: In the formula: Q amm,y , Q alc,y The first y Annual production of synthetic ammonia and methanol; H amm,y , H alc,y The amount of hydrogen allocated to the corresponding production; α amm , α alc The hydrogen consumption coefficient per unit product is determined based on the chemical reaction equation. or amm , or alc This is the efficiency coefficient of the production process.

[0044] An evaluation index system is established, including mean absolute error (MAE) and root mean square error (RMSE), and its calculation formula is as follows: In the formula, y i For the true value, These are the model's predicted values. n This represents the number of samples.

[0045] By associating the prediction results with spatiotemporal labels, a multi-dimensional, multi-scenario demand prediction database is constructed, in which any predicted value in the database... Q forecast Each is determined by a multivariate function, and its structure is expressed as: In the formula, Q forecast A specific demand forecast value stored in the database; F model For a complete prediction model; i For product labels, specify the predicted object as hydrogen, alcohol, or ammonia; tSpecify the time point or range for the forecast, such as year, quarter, or month, for the time label; r Specify the predicted geographic area for the spatial label; S k As a scenario label, it represents a pre-defined macro-scenario. k The scenario is defined by a specific value vector of a set of key input parameters.

[0046] Scenario analysis is conducted based on the forecast results, and the demand elasticity coefficient is calculated. e Q,x To quantitatively simulate the impact of changes in different policies or market variables x on the future demand Q for hydrogenated methanol and ammonia, the calculation formula is as follows: In the formula, e Q,x For demand Q For variables x The elastic modulus; Q new For in variable x The value of becomes x new Under the new scenario, the new demand predicted by the model; Q base To determine the variables in the baseline scenario x The value is x base At that time, the baseline demand predicted by the model; x new For variables x Values ​​under the new scenario (e.g., adjusted carbon tax price); x base For variables x Initial values ​​under the baseline scenario.

[0047] The forecast results will be used to guide downstream industries in capacity planning, capital investment, raw material procurement, and government energy strategy formulation.

[0048] In another possible implementation, when constructing the supply chain transmission model, an agent-based modeling approach can be adopted. This involves abstracting each entity in the supply chain (such as manufacturing companies, transportation companies, and sales companies) as an agent, with each agent possessing its own behavioral rules and decision-making mechanisms. The supply chain transmission model is constructed by simulating the interactions and information transmission between these agents. Furthermore, considering the uncertainties in the supply chain, random variables and probability distributions are introduced to improve the model's accuracy and reliability.

[0049] In another possible implementation, the supply chain transmission model can be constructed using a combination of system dynamics and machine learning. First, a system dynamics model is used to establish the macroscopic structure and causal relationships of the supply chain. Then, machine learning algorithms are used to optimize and adjust the model's parameters. Through training and validation with historical data, the model's performance is continuously improved, enabling it to better reflect the actual operation of the supply chain.

[0050] Example 3 illustrates a schematic scheme of a synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission. It should be noted that the technical solution of a synergistic demand forecasting system for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission is based on the same concept as the aforementioned synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission. Details not described in detail in the synergistic demand forecasting system for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission in this example can be found in the description of the aforementioned synergistic demand forecasting method for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission.

[0051] This embodiment also provides a synergistic demand forecasting system for hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, including: The data acquisition and integration module and the feature dataset construction module are used to collect and integrate multidimensional heterogeneous data reflecting the upstream and downstream relationships of the hydrogen alcohol and ammonia industry chain, and construct feature datasets by dividing them according to geographical and time dimensions. The hydrogen demand two-layer forecasting module is used to build a hydrogen demand two-layer forecasting model based on the feature dataset and obtain the baseline hydrogen demand forecast. The hydrogen resource dynamic allocation module is used to establish a hydrogen resource dynamic allocation model based on market and policy factors, and calculate the amount of hydrogen allocated to downstream alcohol and ammonia production based on the baseline hydrogen demand forecast. The supply chain transmission and result verification module is used to construct a supply chain transmission model. Based on the allocated hydrogen quantity, combined with stoichiometry and production efficiency, it derives the final demand for alcohols and ammonia, and establishes an evaluation index system to verify the prediction results.

[0052] This embodiment also provides an electronic device applicable to a method for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, including: The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement a method for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, as proposed in the above embodiments.

[0053] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a method for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission, as proposed in the above embodiment.

[0054] The storage medium proposed in this embodiment belongs to the same inventive concept as the method for predicting the synergistic demand of hydrogen alcohols and ammonia based on a macro-micro dual-layer framework and industrial chain transmission proposed in the above embodiment. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0055] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A hydrogen-alcohol-ammonia collaborative demand forecasting method based on a macro-micro double-layer framework and industry chain transmission, characterized in that, The method comprises the following steps: Collect and integrate multi-dimensional heterogeneous data reflecting the upstream and downstream relationship of the hydrogen-alcohol-ammonia industry chain, divide according to the geographical and time dimensions, and construct a feature data set; Based on the feature data set, a hydrogen demand double-layer prediction model is constructed to obtain a benchmark hydrogen demand prediction value; A hydrogen resource dynamic allocation model based on market and policy factors is established, and according to the benchmark hydrogen demand prediction value, the hydrogen amount allocated to the downstream alcohol and ammonia production links is calculated; An industry chain transmission model is constructed, and according to the allocated hydrogen amount, combined with the stoichiometric relationship and production efficiency, the final demand amount of alcohol and ammonia is deduced, and an evaluation index system is established to verify the prediction result.

2. The hydrogen-alcohol-ammonia collaborative demand forecasting method based on macro-micro double-layer framework and industry chain transmission according to claim 1, characterized in that, The collection and integration of multi-dimensional heterogeneous data reflecting the upstream and downstream relationship of the hydrogen-alcohol-ammonia industry chain, divided according to the geographical and time dimensions, and the construction of the feature data set comprise: Based on the collected multi-dimensional heterogeneous data, at least one cross-domain coupling feature for quantifying the coupling relationship of the industry chain is created, and the creation logic of the cross-domain coupling feature is to quantitatively calculate the association relationship of different types of data, which at least includes an index for measuring the relative competitiveness of green hydrogen in the economic field and an index for measuring the support intensity of regional policies to the industry chain.

3. The hydrogen-alcohol-ammonia collaborative demand forecasting method based on macro-micro double-layer framework and industry chain transmission according to claim 2, characterized in that, The collection and integration of multi-dimensional heterogeneous data reflecting the upstream and downstream relationship of the hydrogen-alcohol-ammonia industry chain, divided according to the geographical and time dimensions, and the construction of the feature data set further comprise: The index for measuring the economic competitiveness of green hydrogen is calculated by comparing the sum of cost data related to traditional energy with renewable energy and hydrogen production efficiency related data; and the index for measuring the support intensity of regional policies is obtained by weighted sum of the quantitative scores of various support policies in the region.

4. The hydrogen-alcohol-ammonia collaborative demand forecasting method based on macro-micro double-layer framework and industry chain transmission according to claim 3, characterized in that, The construction of the hydrogen demand double-layer prediction model based on the feature data set to obtain the benchmark hydrogen demand prediction value comprises: A hydrogen demand double-layer prediction model including macro trend prediction and micro dynamic prediction is constructed, system dynamics method is used for macro trend prediction, a causal feedback loop model including multiple subsystems is established, and each subsystem has different influencing factors and their interrelated paths to affect hydrogen demand; A data-driven regression model is used for micro dynamic prediction; The benchmark hydrogen demand prediction value is obtained through the hydrogen demand double-layer prediction model.

5. The hydrogen-alcohol-ammonia collaborative demand forecasting method based on macro-micro double-layer framework and industry chain transmission according to claim 4, characterized in that, The construction of the industry chain transmission model, according to the allocated hydrogen amount, combined with the stoichiometric relationship and production efficiency, to deduce the final demand amount of alcohol and ammonia, and the establishment of an evaluation index system to verify the prediction result comprise: A hydrogen-alcohol-ammonia conversion mathematical model is established, and according to the hydrogen amount allocated to alcohol and ammonia production, combined with the hydrogen consumption coefficient per unit product determined according to the chemical reaction equation and the production process efficiency coefficient, the production of alcohol and ammonia is deduced.

6. The hydrogen-alcohol-ammonia synergistic demand prediction method based on macro-micro double-layer framework and industry chain conduction according to claim 5, characterized in that, The construction of the industry chain transmission model, according to the allocated hydrogen amount, combined with the stoichiometric relationship and production efficiency, to deduce the final demand amount of alcohol and ammonia, and the establishment of an evaluation index system to verify the prediction result further comprise: An evaluation index system is established to measure the accuracy of the prediction by comparing the difference between the model prediction value and the true value; The prediction result is associated with the space-time label, a multi-dimensional and multi-scenario demand prediction database is constructed, each prediction value in the database is determined by a multivariate function, and the multivariate function combines product labels, time labels, space labels and scenario labels to specify specific prediction cases.

7. The hydrogen-alcohol-ammonia synergistic demand prediction method based on macro-micro double-layer framework and industry chain conduction according to claim 6, characterized in that, The construction of the industry chain transmission model, according to the allocated hydrogen amount, combined with the stoichiometric relationship and production efficiency, deduces the final demand amount of alcohol and ammonia, and establishes an evaluation index system to verify the prediction result, and the method further comprises: Based on the prediction result, scenario analysis is carried out, and the influence of changes of different policies or market variables on the future demand amount of hydrogen alcohol and ammonia is quantitatively simulated by calculating the demand elasticity coefficient.

8. A hydrogen-alcohol-ammonia collaborative demand forecasting system based on a macro-micro double-layer framework and an industry chain conduction, applying the method of any one of claims 1-7, characterized in that, It comprises: A data collection and integration and feature data set construction module is used to collect and integrate multi-dimensional heterogeneous data reflecting the upstream and downstream relationship of the hydrogen alcohol and ammonia industry chain, and to construct a feature data set according to the geographical and time dimensions; A hydrogen demand double-layer prediction module is used to construct a hydrogen demand double-layer prediction model based on the feature data set to obtain a benchmark hydrogen demand prediction amount; A hydrogen resource dynamic allocation module is used to establish a hydrogen resource dynamic allocation model based on market and policy factors, and to calculate the hydrogen amount allocated to the downstream alcohol and ammonia production link according to the benchmark hydrogen demand prediction amount; An industry chain transmission and result verification module is used to construct an industry chain transmission model, to deduce the final demand amount of alcohol and ammonia according to the allocated hydrogen amount, combined with the stoichiometric relationship and production efficiency, and to establish an evaluation index system to verify the prediction result.

9. An electronic device, comprising: It comprises: A memory and a processor; The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions, and the computer executable instructions are executed by the processor to realize the steps of the method in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It has computer executable instructions, and the computer executable instructions are executed by the processor to realize the steps of the method in any one of claims 1 to 7.

Images