Main control factor and cooperation mechanism analysis method of multi-field coupling thermochemical energy storage system

Abstract

The invention discloses a main control factor and cooperation mechanism analysis method for a multi-field coupling thermochemical energy storage system, and the method comprises the steps: obtaining the multi-field coupling historical data of a reaction unit in the thermochemical energy storage system, and determining a structured multi-field data matrix; constructing a structured variable set which comprises a target variable and an observation variable set, and forming an observation information expansion matrix; calculating mutual information and specific mutual information based on the target variable and the observation variable set; fine causal decomposition is carried out, specifically, for each target state, cooperative-unique-redundant decomposition is carried out on the specific mutual information based on incremental sorting and recursive distribution of the specific mutual information to obtain a decomposition result, and the decomposition result is decomposed into redundant causal contribution, unique causal contribution, cooperative causal contribution and causal leakage components; based on a decomposition result, integrating causal contributions of the full-state space and carrying out normalization processing; and identifying main control factors and a cooperation mechanism of the system, and generating optimization suggestions. The invention further discloses a system, electronic equipment and a computer readable storage medium.

Description

Technical Field

[0001] This invention relates to the field of thermochemical energy storage system analysis and optimization technology, specifically to a method and system for analyzing the main controlling factors and synergistic mechanisms of multi-field coupled thermochemical energy storage systems. It relates to a technique for identifying the main controlling factors and quantifying the synergistic mechanisms of multi-field coupled thermochemical energy storage systems based on causal component decomposition, applicable to the analysis of the main controlling factors and synergistic enhancement mechanisms of the thermo-fluid-chemical multi-field coupled process in the calcium oxide / calcium hydroxide reaction unit. Background Technology

[0002] Against the backdrop of energy structure transformation and the "dual carbon" goal, efficient and high-density thermal energy storage technology is key to realizing the consumption of renewable energy and the utilization of industrial waste heat. Thermochemical energy storage has become a research hotspot due to its advantages such as high energy density, long-term storage and low heat loss. Among them, the reversible reaction system based on CaO / Ca(OH)2 is regarded as a highly promising technology route due to its abundant raw materials, moderate reaction temperature and high energy density.

[0003] In practical engineering applications, the energy storage / release performance of the CaO / Ca(OH)₂ reaction unit is significantly affected by the coupling of multiple physical fields, including heat, fluidization, and chemistry. The reaction unit involves strong coupling between water vapor transport within porous media, non-uniform temperature field distribution, chemical reaction kinetics, and heat and mass transfer processes, forming a nonlinear dynamic system spanning multiple scales (microscopic pores-mesoscopic clusters-macroscopic units). Currently, performance analysis and design optimization for such systems mainly rely on experimental trial and error and numerical simulations based on simplified assumptions, which present the following prominent bottlenecks:

[0004] 1. The multi-field coupling mechanism is unclear, and synergistic and redundant effects are difficult to quantify: Existing analytical methods mostly focus on the influence of a single physical field or a few variables, lacking a systematic quantification of the synergistic enhancement, unique contributions, or redundant cancellation effects among multiple heat-fluidization fields. For example, it is impossible to determine whether the temperature gradient and water vapor diffusion are synergistically promoting reactions or whether there is information redundancy; it is also difficult to identify which factor among heat transfer, mass transfer, and reaction resistance dominates the energy storage / release rate under specific operating conditions.

[0005] 2. Complex cross-scale correlations, with key factor identification relying on experience: The cross-scale correlation mechanisms from microscopic pore structure to macroscopic energy storage / release performance have not been fully revealed. Traditional methods often use parameter sensitivity analysis or statistical regression to screen factors, failing to distinguish the true contribution and coupling effect of each factor from a causal inference perspective. This results in a lack of focus in the design direction of enhancing the energy storage power density of the reaction unit, making it difficult to achieve "precise enhancement." 3. Lack of dynamic causal analysis tools for nonlinear and unsteady-state processes: The reaction unit involves coupling across multiple time scales during the transient energy storage / release process. Traditional steady-state or quasi-steady-state models struggle to capture the evolution of dominant factors in dynamic processes, and are even less able to quantify the dynamic adjustment of multi-field synergistic mechanisms at different time scales.

[0006] To address the aforementioned challenges, integrating information-theoretic causal inference with multi-field coupled system analysis has become an important approach to overcome bottlenecks. In recent years, causal discovery methods, such as Granger causality and transfer entropy, have been introduced into engineering system analysis, but they still struggle to handle synergistic and redundant effects among multiple variables, and lack robustness to nonlinearity, non-Gaussian noise, and unobserved variables (such as microstructural evolution).

[0007] Therefore, in order to deeply reveal the synergistic enhancement mechanism of the thermal-fluidic-chemical multi-field coupling in the CaO / Ca(OH)2 reaction unit and achieve the leap from "qualitative description" to "quantitative attribution", there is an urgent need for a technology that can systematically decompose and quantify the synergistic, unique and redundant causal contributions of each physical field. This technology should go beyond simple correlation and association analysis to study complex physical nonlinear systems, accurately identify the main controlling factors of performance enhancement, and provide interpretable and operable optimization basis for reaction unit structural design, material selection and operation control. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method and system for analyzing the main control factors and synergistic mechanisms of multi-field coupled thermochemical energy storage systems. Addressing the technical deficiencies in existing multi-field coupled analysis of thermochemical energy storage systems, such as the difficulty in quantifying and identifying the main control factors of performance, the inability to analyze the synergistic and redundant mechanisms between multiple physical fields, and the lack of assessment of the impact of unobserved factors, this invention provides a technique for identifying the main control factors and quantifying the synergistic mechanisms of multi-field coupled thermochemical energy storage systems based on causal component decomposition. By introducing an information-theoretic causal inference framework, the complex dependencies between system performance and multi-physical field variables are precisely decomposed into three basic causal modes: redundancy, uniqueness, and synergy. The impact of unobserved factors is quantified, thereby directly and quantitatively revealing the intrinsic mechanism of performance enhancement from the data. This provides a non-invasive and interpretable analytical tool for solving the key causal identification problem in the optimal design of reaction units.

[0009] This invention provides a method for analyzing the controlling factors and synergistic mechanisms of multi-field coupled thermochemical energy storage systems, achieving causal inference and quantification based on causal component decomposition, including:

[0010] S1, acquire multi-field coupling historical data of reaction units in thermochemical energy storage system, and determine a structured multi-field data matrix based on the multi-field coupling historical data; wherein, the multi-field coupling historical data comes from numerical simulation output or experimental monitoring system, and includes time series data under one or more operating conditions;

[0011] S2, construct a structured variable set, which includes a target variable set and an observation variable set, and form an observation information extension matrix based on the observation variable set, wherein the observation information extension matrix includes the observation variables at the current time and the historical state of the observation variables at the current time;

[0012] S3, calculate mutual information and specific mutual information based on the target variable and the set of observed variables; wherein the specific mutual information is the input of fine causal decomposition;

[0013] S4, perform refined causal decomposition, including: for each target state, perform cooperative-unique-redundant decomposition on the specific mutual information based on incremental sorting and recursive allocation to obtain the decomposition result, decomposing the specific mutual information into redundant causal contribution, unique causal contribution, cooperative causal contribution, and causal leakage component; wherein the redundant causal contribution represents the causal contribution of all possible non-empty subsets of the observation information expansion matrix to the target variable in a certain possible state; the unique causal contribution represents information about a certain possible state that is unique to a single variable and cannot be obtained from any other single variable in the set of observed variables; the cooperative causal contribution represents additional information about a certain possible state that emerges only when multiple variables in the variable combination are jointly observed, and cannot be obtained from any subset of the variable combination; the causal leakage component represents the part of information about the target variable in a certain possible state that is jointly participated in by the entire observation information expansion matrix but cannot be explained, the part of information originating from latent variables not included in the current observation or purely random fluctuations;

[0014] S5, Integrate the causal contributions of the entire state space based on the decomposition results;

[0015] S6, normalize the global causal contribution;

[0016] S7. Based on the normalized global causal contribution identification system's main control factors and collaborative mechanisms, optimization suggestions are generated.

[0017] Preferably, S1 includes:

[0018] S11, acquire multi-field coupling historical data of the reaction unit in the thermochemical energy storage system; the multi-field coupling historical data includes: temperature field information, water vapor concentration field information, flow velocity field information, pore structure parameters, reaction progress and external heat and mass input conditions; wherein the external heat and mass input conditions include boundary heat flux density, boundary vapor pressure, energy storage efficiency and energy release power;

[0019] S12, preprocessing the multi-field coupling historical data, including: for the data output from the numerical simulation, deriving the evolution data of each physical field at discrete spatiotemporal nodes from the established fluid-thermal-chemical multi-field coupling cross-scale model, which can analyze the key processes from the pore scale to the reaction unit scale; for the data from the experimental monitoring system, synchronously acquiring time-series signals through a sensor network deployed within the reaction unit; performing high-dimensional data feature extraction operations on the evolution data of each physical field at discrete spatiotemporal nodes and the time-series signals, and using statistical methods to compress the spatially distributed high-dimensional data of the temperature field, concentration field, and velocity field to extract feature values ​​that can represent the overall state of the physical field;

[0020] S13, determining a structured multi-field data matrix based on the preprocessed multi-field coupled historical data, including: performing time interpolation and spatial alignment preprocessing on the sensor data of each sensor in the sensor network to form the spatiotemporally matched structured multi-field data matrix; and performing independent normalization processing on the variable data of each dimension in the structured multi-field data matrix.

[0021] Preferably, S2 includes:

[0022] S21, Select the core performance indicators as the target variables;

[0023] S22, Select relevant physical field variables to construct the set of observation variables, and form an observation information matrix containing historical information based on the set of observation variables.

[0024] Preferably, S3 includes:

[0025] S31, Based on information theory, calculate the mutual information between the target variable and the set of observed variables;

[0026] S32, based on the mutual information, further calculate the specific mutual information for each possible state of the target variable.

[0027] Preferably, S3 further includes:

[0028] S33, calculate the expectation based on the specific mutual information of all possible states of the target variable to restore the total mutual information.

[0029] Preferably, S4 includes:

[0030] S41, Initialize the remaining variable set, which contains all observation variable indices, including: creating a list containing all observation variable indices, called the remaining variable set, which will be dynamically updated during the allocation process;

[0031] S42, calculate the specific mutual information of each non-empty subset in the observation information extended matrix with respect to the target state;

[0032] S43, each non-empty subset is sorted in ascending order according to the value of specific mutual information, and the information increment between adjacent sorted non-empty subsets is calculated by subtracting the information of the previous subset from the information of the next subset. The information increment is used to characterize the information gain brought about by introducing new variables or combinations of variables on the basis of existing information.

[0033] S44, recursively allocate the information increment, including: processing each non-empty subset and its corresponding information increment in sequence according to the sorting order in S43;

[0034] S45, defining unique contributions, including: after completing the recursive allocation according to S44, in all redundant contribution dictionaries, the entries with a single variable as the key are redefined as the unique contributions corresponding to that variable.

[0035] S46, the causal leakage component is obtained by subtracting the sum of all allocated redundant contributions and cooperative contributions from the total specific mutual information, wherein all allocated redundant contributions include items that have been converted into unique contributions.

[0036] Preferably, S5 includes:

[0037] S51 integrates all possible target states into a full state space;

[0038] S52, for each possible target state in the full state space, perform a probability-weighted average of the corresponding decomposition results to obtain the global causal contribution and global causal leakage that reflect the average causal strength of the multi-field coupled thermochemical energy storage system, wherein the global causal contribution includes global redundancy, uniqueness and cooperative contribution.

[0039] Preferably, S6 includes:

[0040] S61, normalize the global causal contribution to a proportion relative to the total mutual information;

[0041] S62 normalizes causal leakage to a ratio relative to the entropy of the target variable.

[0042] Preferably, S7 includes:

[0043] S71, by comparing the magnitude of the normalized global causal contribution, unique controlling factors, synergistic reinforcement mechanisms, and redundant variable combinations are identified, and the modeling completeness of the thermochemical energy storage system is evaluated. The criteria for identifying unique controlling factors and synergistic mechanisms include: if the normalized unique contribution of a variable is significantly higher than that of other variables, then the variable is inferred to be a unique controlling factor; if the normalized synergistic contribution of a variable combination is significantly greater than zero, then the combination is identified to have a synergistic reinforcement mechanism; if the normalized redundant contribution of a variable combination is significant, then it indicates that the information of these variables overlaps and can be used to guide model simplification; if the normalized causal leakage is high, then it suggests the existence of important unobserved variables and the need for further model improvement or experimental observation.

[0044] S72, based on the quantitative identification conclusions of the normalized global causal components, main control factors and synergistic mechanisms, combined with the physical mechanism of the reaction unit, provides optimization suggestions for the optimization design and operation strategy formulation of thermochemical energy storage reaction units, including: implementing precise control of unique main control factors, designing engineering schemes to promote the joint effect of synergistic variable groups, simplifying monitoring systems or simulation models based on redundancy analysis, and supplementing key variable observations or deepening mechanism research for high causal leakage indication directions.

[0045] The second aspect of the present invention provides a system for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system, for implementing the method of the first aspect, comprising:

[0046] The structured multi-field data matrix construction module (101) is used to acquire the multi-field coupling historical data of the reaction unit in the thermochemical energy storage system, and to determine the structured multi-field data matrix based on the multi-field coupling historical data; wherein, the multi-field coupling historical data comes from the numerical simulation output or the experimental monitoring system, and includes time series data under one or more operating conditions;

[0047] The structured variable set construction module (102) is used to construct a structured variable set, which includes a target variable set and an observation variable set, and forms an observation information extension matrix based on the observation variable set, wherein the observation information extension matrix includes the observation variable at the current time and the historical state of the observation variable at the current time;

[0048] The mutual information and specific mutual information calculation module (103) is used to calculate mutual information and specific mutual information based on the target variable and the set of observed variables; wherein the specific mutual information is the input of fine causal decomposition;

[0049] The refined causal decomposition module (104) is used to perform refined causal decomposition, including: for each target state, performing a cooperative-unique-redundant decomposition on the specific mutual information based on incremental sorting and recursive allocation to obtain the decomposition result, decomposing the specific mutual information into redundant causal contribution, unique causal contribution, cooperative causal contribution, and causal leakage component; wherein the redundant causal contribution represents the causal contribution of all possible non-empty subsets of the observation information expansion matrix to the target variable in a certain possible state; the unique causal contribution represents information about a certain possible state that is unique to a single variable and cannot be obtained from any other single variable in the set of observation variables; the cooperative causal contribution represents additional information about a certain possible state that emerges only when multiple variables in the variable combination are jointly observed, and cannot be obtained from any subset of the variable combination; the causal leakage component represents the part of information about the target variable in a certain possible state that is jointly participated in by the entire observation information expansion matrix but cannot be explained, the part of information originating from latent variables not included in the current observation or purely random fluctuations;

[0050] The full-state-space causal contribution integration module (105) is used to integrate the causal contributions of the full-state space based on the decomposition results.

[0051] The global causal contribution processing module (106) is used to normalize the global causal contribution.

[0052] The main control factor and coordination mechanism analysis module (107) is used to generate optimization suggestions based on the main control factors and coordination mechanisms of the global causal contribution identification system after normalization.

[0053] A third aspect of the present invention provides an electronic device including a processor and a memory, the memory storing a plurality of instructions, the processor being configured to read the instructions and execute the method as described in the first aspect.

[0054] A fourth aspect of the present invention provides a computer-readable storage medium storing a plurality of instructions which can be read by a processor and executed as described in the first aspect.

[0055] The beneficial effects of the method and system of the present invention are as follows:

[0056] The method and system of this invention involve a technique for identifying the controlling factors and quantifying the synergistic mechanisms of multi-field coupled thermochemical energy storage systems based on causal component decomposition. This provides a novel, data-driven causal analysis paradigm for in-depth analysis of the multi-physics coupling mechanisms in complex energy storage systems. This method can not only clearly quantify the unique and synergistic contributions of each physical field to system performance and identify key controlling factors, but also assess the influence of unobserved factors. Therefore, it provides direct and reliable theoretical basis and quantitative guidance for the design, optimization, and control of high-performance thermochemical energy storage reaction units. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the specific embodiments or related technologies of the present invention, the drawings used in the description of the specific embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0058] Figure 1 A flowchart of a method for identifying the main controlling factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition, provided by existing technology.

[0059] Figure 2 This is a flowchart of step S1 of the method for identifying the main controlling factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition according to an embodiment of the present invention.

[0060] Figure 3 This is a flowchart of step S2 of the method for identifying the main controlling factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition according to an embodiment of the present invention.

[0061] Figure 4 This is a flowchart of step S3 of the method for identifying the main controlling factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition according to an embodiment of the present invention.

[0062] Figure 5 This is a flowchart of step S4 of the method for identifying the main controlling factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition according to an embodiment of the present invention.

[0063] Figure 6 This is a flowchart of step S5 of the method for identifying the main controlling factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition according to an embodiment of the present invention.

[0064] Figure 7This is a flowchart of step S6 in the method for identifying the main controlling factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition according to an embodiment of the present invention.

[0065] Figure 8 This is a flowchart of step S7 of the method for identifying the main controlling factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition according to an embodiment of the present invention.

[0066] Figure 9 This is a system architecture diagram for identifying the main controlling factors and quantifying the collaborative mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition, provided according to an embodiment of the present invention.

[0067] Figure 10 This is a structural diagram of an electronic device provided according to an embodiment of the present invention. Detailed Implementation

[0068] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. 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.

[0069] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0070] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0071] Example 1

[0072] Figure 1 This is a flowchart of the method for identifying the main control factors and quantifying the synergistic mechanism of a multi-field coupled thermochemical energy storage system based on causal component decomposition, provided in Embodiment 1 of the present invention.

[0073] like Figure 1 As shown, this embodiment provides a method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system. It achieves causal inference and quantification based on causal component decomposition, including:

[0074] S1, acquire multi-field coupling historical data of reaction units in thermochemical energy storage system, and determine a structured multi-field data matrix based on the multi-field coupling historical data; wherein, the multi-field coupling historical data comes from numerical simulation output or experimental monitoring system, and includes time series data under one or more operating conditions;

[0075] like Figure 2 As shown, in a preferred embodiment, S1 includes:

[0076] S11, acquire multi-field coupling historical data of the reaction unit in the thermochemical energy storage system; the multi-field coupling historical data includes: temperature field information. Water vapor concentration field information Flow field information Pore ​​structure parameters , reaction progress and external heat mass input conditions; wherein the external heat mass input conditions include boundary heat flux density Boundary vapor pressure Energy storage efficiency and energy release power ;

[0077] S12, preprocessing the multi-field coupling historical data, including: for the data output from the numerical simulation, deriving the evolution data of each physical field at discrete spatiotemporal nodes from the established fluid-thermal-chemical multi-field coupling cross-scale model, which can analyze the key processes from the pore scale to the reaction unit scale; for the data from the experimental monitoring system, synchronously acquiring time-series signals through a sensor network deployed within the reaction unit; performing high-dimensional data feature extraction operations on the evolution data of each physical field at discrete spatiotemporal nodes and the time-series signals, and using statistical methods to compress the spatially distributed high-dimensional data of the temperature field, concentration field, and velocity field to extract feature values ​​that can represent the overall state of the physical field;

[0078] S13, determining a structured multi-field data matrix based on the preprocessed multi-field coupled historical data, including: performing time interpolation and spatial alignment preprocessing on the sensor data of each sensor in the sensor network to form the spatiotemporally matched structured multi-field data matrix; and performing independent normalization processing on the variable data of each dimension in the structured multi-field data matrix.

[0079] In this embodiment, S1 aims to construct a high-quality, structured multi-field coupled historical dataset to provide input for subsequent causal inference and quantification. The multi-field coupled historical data originates from numerical simulation output or experimental monitoring systems and includes time-series data under one or more operating conditions.

[0080] In some embodiments of this example, the core physical field variables and performance indicators of interest in the historical data of the reaction unit of the CaO / Ca(OH)2 thermochemical energy storage system include, but are not limited to: temperature field information. Water vapor concentration field information Flow field information Pore ​​structure parameters , reaction progress External heat mass input conditions (such as boundary heat flux density) Boundary vapor pressure Energy storage efficiency or energy release power Performance indicators, etc.

[0081] In some embodiments of this example, the specific implementation methods for data acquisition and preprocessing of the multi-field coupled dataset corresponding to the reaction unit of the energy storage system are as follows: For numerical simulation data, the evolution data of each physical field at discrete spatiotemporal nodes are derived from the established fluid-thermal-chemical multi-field coupled cross-scale model. The model should be able to analyze the key processes from the pore scale to the reaction unit scale. For experimental data, time-series signals are synchronously acquired through a sensor network (such as thermocouples, humidity sensors, flow meters, etc.) arranged in the reactor. Since the sampling frequency and location of each sensor may be different, preprocessing such as time interpolation and spatial alignment is required to form a spatiotemporally matched structured multi-field data matrix.

[0082] In this embodiment, to improve the feasibility and efficiency of subsequent information theory calculations, a high-dimensional data feature extraction operation is performed. This step aims to achieve information dimensionality reduction and feature extraction. Specifically, statistical methods are used to compress high-dimensional data with spatial distributions, such as temperature fields, concentration fields, and flow velocity fields, and extract feature values ​​that can represent the overall state of the physical field, such as: spatial mean (reflecting the overall level), variance (reflecting uniformity), gradient in a specific direction (reflecting the driving force intensity), etc.

[0083] To obtain a structured multi-field data matrix and eliminate the differences in dimensions and orders of magnitude between different physical variables, thus preventing them from biasing information-theoretic calculations, it is necessary to perform independent normalization processing on the data of each dimension of the variable to obtain normalized variable data. This embodiment uses the Z-score normalization method to normalize the variable data. The calculation formula is shown in equation (1) below:

[0084] (1);

[0085] in, Indicates the first The time point, the first The original observations of each physical variable; and These represent the sample mean and sample standard deviation of the variable across all time points, respectively. This process transforms all variables into a distribution with a mean of 0 and a standard deviation of 1, thus creating a normalized historical dataset that can be directly used by subsequent causal decomposition algorithms.

[0086] S2, construct a structured variable set, which includes a target variable set and an observation variable set, and form an observation information extension matrix based on the observation variable set, wherein the observation information extension matrix includes the observation variables at the current time and the historical state of the observation variables at the current time;

[0087] like Figure 3 As shown, in a preferred embodiment, S2 includes:

[0088] S21, Select the core performance indicators as the target variables;

[0089] S22, Select relevant physical field variables to construct the set of observation variables, and form an observation information matrix containing historical information based on the set of observation variables.

[0090] In this embodiment, to perform causal analysis and quantitative decomposition of the performance formation mechanism of the thermochemical energy storage reaction unit, it is necessary to first clarify the "cause" and "effect" in the research object. This step aims to construct a set of structured variables suitable for the analytical framework of this invention.

[0091] A variable that characterizes the core performance of a system and is affected by multi-physics coupling is selected as the research object, called the target variable, and denoted as . The target variable is usually a key performance indicator of the reaction unit in the future, such as the energy storage efficiency of the reaction unit in the energy storage stage. Energy release power during the energy release phase or reaction process In the future state. Its mathematical representation is Equation (2):

[0092] (2);

[0093] in, For the target variable, This represents a forward-looking time increment. This setting reflects the temporal principle of causality—cause must precede or be synchronous with effect—and focuses on understanding how factors influence the future state of the system.

[0094] Corresponding to the target variable, a set of observed variables that may influence it is defined. These variables originate from the multi-field coupling historical data obtained in step S1 and constitute the set of physical quantities that could potentially "cause" the problem. All variables to be examined in time... The observable physical field variables constitute a multidimensional vector. As shown in equation (3):

[0095] (3);

[0096] in, Representing the The observed variables at time... The value of , for example, the spatial average temperature at that moment. Inlet water vapor partial pressure Integral average velocity of the characteristic cross section of the flow field or reaction bed porosity wait, This represents the total number of observed variables to be studied.

[0097] Select a variable to be studied in terms of mechanism decomposition (e.g., energy storage efficiency of the reaction unit). ), denoted as the target variable, where This represents a future observation time window, used to indicate the focus on the state of the target variable at a certain future moment.

[0098] To comprehensively capture the causal chain from "past" to "future" and accommodate instantaneous (same time step) and lagging (historical time step) dependencies, an observation information extension matrix (or historical window vector) is constructed to predict and explain the target variable. The changes. The observation information extended matrix not only includes the observed variables at the current moment, but also encompasses their historical states. The general form can be expressed as shown in equation (4):

[0099]

[0100] in, By appropriately setting the length of the observation information extension matrix (or historical window vector) to account for the maximum historical lag order, the observation information extension matrix can simultaneously characterize instantaneous causality and lagged causality. Instantaneous causality represents the mutual influence between different physical fields at the same moment; lagged causality represents the delayed effect of historical states on future goals. This design allows the analysis method to closely align with the spatiotemporal continuity of information transmission in actual physical processes, laying a comprehensive data foundation for the subsequent precise decomposition of the synergistic, unique, and redundant causal relationships of the energy storage and release mechanism of thermochemical energy storage reaction units.

[0101] S3, calculate mutual information and specific mutual information based on the target variable and the set of observed variables; wherein the specific mutual information is the input of fine causal decomposition;

[0102] like Figure 4 As shown, in a preferred embodiment, S3 includes:

[0103] S31, Based on information theory, calculate the mutual information between the target variable and the set of observed variables;

[0104] S32, based on the mutual information, further calculate the specific mutual information for each possible state of the target variable;

[0105] In this embodiment, after constructing the set of target variables and observed variables, step S3 introduces core metrics of information theory (mutual information and its fine-grained form specific mutual information) to achieve precise quantification of causal relationships. These metrics can go beyond linear correlation analysis, capturing generalized statistical associations between variables, including nonlinear dependencies, and serve as the mathematical basis for subsequent decomposition of causal components such as co-occurrence, uniqueness, and redundancy.

[0106] First, corresponding to step S31, calculate the target variable. With observation information extended matrix Mutual information between Mutual information Quantified through observation What can be obtained about The average reduction in uncertainty, i.e. The information carried The total amount of information. Its calculation formula is shown in the following formula (5):

[0107] (5);

[0108] In the formula, The target variable The information entropy of the sequence formed at all time points, It is an expanded matrix of known observation information. hour Conditional entropy; The target variable Values Simultaneously observe the extended matrix of information Values The joint probability at time; For target variable Values At that time, the target variable The corresponding marginal probabilities; Expanding the observation information matrix Values At that time, the observation information expansion matrix The corresponding marginal probabilities; the mutual information value is always non-negative if and only if the target variable With observation information extended matrix The value is zero when the components are independent, which means it reflects the current observation information expansion matrix of the system. For target variable Without any causal input, the probability distribution in equation (5) The estimation is obtained from the multi-field coupling historical data after the reaction unit preprocessing in step S1. The estimation includes reliably estimating the probability distribution of each dimension variable based on data transformation methods, including kernel density estimation, histogram binning, or k-nearest neighbor-based estimator estimation.

[0109] However, total mutual information is a global average measure that may mask the heterogeneity of causal dependencies across different system states. For example, some observed variables may only have a significant effect on the target variable under specific conditions (such as high temperature or high steam concentration).

[0110] Corresponding to step S32, to reveal the causal mechanism of this state dependence, this embodiment of the invention further introduces the concept of specific mutual information in a fine-grained form. Specific mutual information Focus on target variable In a certain state or value At that time, the observation information expansion matrix The amount of targeted information provided is measured by the amount of information known. The difference between the joint distribution and the distribution under the conditional independence assumption, given this event. Based on the Kullback-Leibler divergence definition, this embodiment will consider specific mutual information. The calculation formula is defined as equation (6):

[0111] (6);

[0112] Equation (6) can be understood as: in the state of the target variable Conditional distribution of observed variables under the given "situation". Compared to its unconditional distribution The degree of "deviation" between the conditional distribution and the original distribution, i.e., the information gain between the conditional distribution and the original distribution, is the amount of information contained in this degree of deviation. Observation information extension matrix under specific conditions The causal contribution of the target variable. This calculation step outputs the state for each target variable. Calculated specific mutual information This provides direct input for subsequent steps to perform fine causal decomposition, thereby enabling the investigation of how the causal roles of various physical field variables in the thermochemical energy storage reaction unit change dynamically under different performance levels or system states.

[0113] In a preferred embodiment, S3 further includes:

[0114] S33, calculate the expectation based on the specific mutual information of all possible states of the target variable to restore the total mutual information.

[0115] In this embodiment, by considering all possible states of the target variable... The total mutual information can be recovered by calculating the expectation of the specific mutual information. The calculation formula is defined as shown in equation (7):

[0116] (7);

[0117] In the formula, For target variable With the entire observation information expansion matrix Mutual information between them The marginal probability corresponding to the target variable. for Specific mutual information under a state.

[0118] S4, perform refined causal decomposition, including: for each target state, perform cooperative-unique-redundant decomposition on the specific mutual information based on incremental sorting and recursive allocation to obtain the decomposition result, decomposing the specific mutual information into redundant causal contribution, unique causal contribution, cooperative causal contribution, and causal leakage component; wherein the redundant causal contribution represents the causal contribution of all possible non-empty subsets of the observation information expansion matrix to the target variable in a certain possible state; the unique causal contribution represents information about a certain possible state that is unique to a single variable and cannot be obtained from any other single variable in the set of observed variables; the cooperative causal contribution represents additional information about a certain possible state that emerges only when multiple variables in the variable combination are jointly observed, and cannot be obtained from any subset of the variable combination; the causal leakage component represents the part of information about the target variable in a certain possible state that is jointly participated in by the entire observation information expansion matrix but cannot be explained, the part of information originating from latent variables not included in the current observation or purely random fluctuations;

[0119] In this embodiment, step S4 aims to use the specific mutual information calculated in step S3. By performing refined subdivision, it surpasses traditional correlation analysis and quantitatively analyzes the extended matrix of observation information. Each variable or combination of variables affects the target state. The causal contribution pattern of events. The core idea of ​​causal component decomposition is that the information about the target provided by multiple variables is not a simple linear superposition; there are both overlapping (redundancy) and joint complementary (synergistic) parts. For thermochemical energy storage systems, this means that understanding the impact of variables such as temperature, concentration, and flow rate on the target variable (such as energy storage efficiency) requires not only knowing their independent contributions, but also determining whether they "provide the same information," "each provide unique information," or "jointly generate new information."

[0120] Specifically, for each specific state of the target variable, the observation information is expanded into a matrix. All possible non-empty subsets (i.e., combinations of variables) , )right The causal contribution can be broken down into the following three basic types:

[0121] 1. Redundant causal contribution : refers to a combination of variables ( In the context of all variables, there is a shared and repetitive relationship regarding each other. Information such as inlet temperature and bed volume average temperature may be highly correlated in a reaction unit, and their information about future energy storage efficiency is largely redundant. Identifying redundant contributions helps simplify models and sensing systems.

[0122] 2. Unique causal contribution Specifically refers to a single variable Unique to, and unobtainable from any other single variable in the set of observed variables, regarding This information identifies the unique causal role of the variable. For example, water vapor partial pressure may provide unique predictive information about reaction completion that temperature or flow rate does not. Identifying unique contributions is key to pinpointing the "controlling factors."

[0123] 3. Co-causal contribution : refers to only when the combination of variables ( Emergence only occurs when multiple variables in a variable combination are observed jointly. Information that cannot be obtained from any of these subsets This provides additional information. It demonstrates the "1+1>2" coupling effect between multiple physics fields. For example, temperature gradients and vapor concentration gradients alone may not predict local reaction hotspots, but their combined spatial distribution patterns can provide crucial information. Quantifying the synergistic contribution is the direct basis for revealing the "multi-field synergistic enhancement mechanism."

[0124] 4. In addition, the decomposition also includes a "causal leakage component". It represents the entire observation information expansion matrix. Even with joint participation, it is impossible to explain what is happening. That part of the information. This part of the information comes from potential variables that are not included in the current observations (such as the microscopic evolution of catalyst activity, dynamic changes in pore structure, etc.) or purely random fluctuations, and is an important indicator for assessing the completeness of the current model / observation.

[0125] The above decomposition must mathematically satisfy the consistency axiom. Specifically, for any combination of variables... The information it provides about General Information All will be decomposed into the sum of redundancy, uniqueness, and collaborative contributions, ensuring that for a given target state... The total information contribution of all observed variables satisfies the following conservation equation, as shown in equation (8):

[0126] (8);

[0127] In equation (8), for The specific mutual information in the state. Combination of variables Provided about Redundant causal contributions; Single variable Provided about Unique causal contribution; Combination of variables The overall joint provision of information The co-causal contribution; The causal leakage component represents the latent variables not included in the current observation system for... causal contribution; This represents the set of all combinations that contain more than one variable. This represents the total number of observed variables to be studied.

[0128] like Figure 5 As shown, in a preferred embodiment, this example employs an algorithm based on incremental sorting and recursive allocation of specific mutual information to implement the above decomposition. This algorithm strictly follows axioms such as nonnegativity and symmetry, and has clear physical interpretability. The flow of this recursive decomposition algorithm, step S4, includes:

[0129] S41, Initialize the remaining variable set, which contains all observation variable indices, including: creating a list containing all observation variable indices, called the remaining variable set, which will be dynamically updated during the allocation process;

[0130] S42, calculate the specific mutual information of each non-empty subset in the observation information extended matrix with respect to the target state;

[0131] S43, each non-empty subset is sorted in ascending order according to the value of specific mutual information, and the information increment between adjacent sorted non-empty subsets is calculated by subtracting the information of the previous subset from the information of the next subset. The information increment is used to characterize the information gain brought about by introducing new variables or combinations of variables on the basis of existing information.

[0132] S44, recursively allocate the information increment, including: processing each non-empty subset and its corresponding information increment sequentially according to the sorting order in S43, including the following processing types:

[0133] (a) If the non-empty subset is a single variable, the current information increment is added to the redundant contribution with the current set of remaining variables as the key, and then the variable is removed from the set of remaining variables. The information provided by the single variable is considered as part of the redundant information shared by all the current remaining variables, including itself.

[0134] (b) If the non-empty subset is a multivariate combination, the current information increment is added to the collaborative contribution term where the non-empty subset itself is the key. The incremental information only appears when all variables are included in the joint observation combination, which is a collaborative effect.

[0135] S45, defining unique contributions, including: after completing the recursive allocation according to S44, in all redundant contribution dictionaries, the entries with a single variable as the key are redefined as the unique contributions corresponding to that variable.

[0136] S46, the causal leakage component is obtained by subtracting the sum of all allocated redundant contributions and cooperative contributions from the total specific mutual information, wherein all allocated redundant contributions include items that have been converted into unique contributions.

[0137] Specific mutual information Decompose into redundant contributions Unique contributions Collaborative contributions and causal leakage The recursive allocation algorithm includes the following core steps:

[0138] The corresponding specific implementation methods include:

[0139] 1. Initialize the set of remaining variables Create a list containing the indices of all observed variables, called the residual variable set, initially set to 0. This set will be dynamically updated during the allocation process.

[0140] 2. Calculate the specific mutual information of all subsets: for the entire observation information expansion matrix Calculate each of its non-empty subsets Regarding the target state Specific mutual information .

[0141] 3. Information sorting and incremental calculation: This involves sorting all the variables obtained in the previous step into subsets... According to its specific mutual information value Sort the subsets in ascending order. Then, calculate the information increment between adjacent sorted subsets. (That is, the information content of the later subset minus the information content of the earlier subset). These increments represent the information gain brought about by introducing new variables (or combinations of variables) on the basis of existing information.

[0142] 4. Recursively allocate information increments: Process each subset sequentially according to the sorting order in step 3. and its corresponding information increment Traverse subsets in sorted order :like If it is a single variable, then the current... Accumulated to the current Redundant terms for the key, and from Remove the variable; if If it is a combination of multiple variables, then the current... Accumulated to In the key of the collaborative term;

[0143] Specifically, the processing types included are as follows:

[0144] (a) If It is a single variable (i.e.) ): Increment the current information Redundant contributions accumulated to the current set of remaining variables as the key Then, the variable is removed from the set of remaining variables. This operation means that the information provided by this single variable is considered part of the redundant information shared by all the remaining variables (including itself).

[0145] (b) If It is a combination of multiple variables (i.e.) ): Increment the current information Accumulated to this combination Collaborative contribution items that are themselves keys This operation means that the incremental information is only available in the joint observation set. This effect only occurs when all variables are included, and it is a synergistic effect.

[0146] 5. Define unique contributions: After completing the above allocation according to step 4, in all redundant contribution dictionaries, the entries whose keys are single variables (i.e., The variable was redefined as the unique contribution corresponding to that variable. This is because when the remaining set of variables contains only the variable itself, the redundant information it allocates is actually information unique to that variable and cannot be provided by other variables.

[0147] 6. Calculate causal leakage: Causal leakage By total specific mutual information Subtracting the sum of all allocated redundant contributions (including those converted to unique contributions) and collaborative contributions yields the result shown in equation (9):

[0148] (9);

[0149] The summation of redundancy contributions includes terms from all multivariate and univariate keys.

[0150] The algorithm used in this embodiment, based on incremental sorting and recursive allocation with specific mutual information, provides a computable and interpretable solution for quantitatively extracting cooperative, unique, and redundant causal patterns from high-dimensional coupled data. The output of this process is specific to each target state. The complete set of decomposition results provides the finest granular data for the next step of overall causal quantification and mechanism interpretation.

[0151] S5, Integrate the causal contributions of the entire state space based on the decomposition results;

[0152] like Figure 6 As shown, in a preferred embodiment, S5 includes:

[0153] S51 integrates all possible target states into a full state space;

[0154] S52, for each possible target state in the full state space, perform a probability-weighted average of the corresponding decomposition results to obtain the global causal contribution and global causal leakage that reflect the average causal strength of the multi-field coupled thermochemical energy storage system, wherein the global causal contribution includes global redundancy, uniqueness and cooperative contribution.

[0155] In this embodiment, the redundancy, uniqueness, cocausal contribution, and causal leakage obtained in step S4 are all for the target variable. In a certain specific state Local quantization at time. Corresponding to step S52, in order to obtain a global index reflecting the overall average causal strength of the system, it is necessary to perform mathematical expectation calculations on each causal component under all possible target states, and obtain the calculation formulas for the overall causal quantification result as shown in equations (10-1) to (10-4):

[0156] (10-1);

[0157] (10-2);

[0158] (10-3);

[0159] (10-4);

[0160] In the formula, The marginal probability corresponding to the target variable. Combination of variables Provided about Redundant causal contributions Combination of variables Provided information about the target variable The global redundant causal contribution was quantified by the combination of variables. The amount of repetitive information provided on average across all system states; Single variable Provided about Unique causal contribution, For a single variable The globally unique causal contribution of the target variable was quantified. The amount of causal information that is unique on average across all system states and cannot be substituted by other variables; Combination of variables The overall joint provision of information Co-causal contribution Combination of variables Provides global co-causal contributions to the target variable; For global causal leakage, the amount of information that cannot be explained by the average observable variables under all system states is quantified, i.e., the average impact of unmodeled factors or random noise.

[0161] This step elevates state-based causal analysis to system-level causal quantification, providing stable and comparable numerical data for subsequent identification of controlling factors and analysis of collaborative mechanisms.

[0162] S6, normalize the global causal contribution;

[0163] like Figure 7 As shown, in a preferred embodiment, S6 includes:

[0164] S61, normalize the global causal contribution to a proportion relative to the total mutual information;

[0165] S62 normalizes causal leakage to a ratio relative to the entropy of the target variable.

[0166] In this embodiment, the global causal component contribution calculated in step S5 has a clear absolute value meaning. However, in order to facilitate horizontal comparison of the relative importance of each contribution and establish a unified evaluation benchmark, normalization processing is required. This step aims to convert the absolute causal contribution into a relative proportion through normalization, thereby clearly and unambiguously identifying the main control factors and core coupling mechanisms affecting system performance.

[0167] Corresponding to step S61, the redundant, unique, and cocausal contributions provided by all observed variables are divided by the target variable. With observation information extended matrix Total mutual information between This mutual information represents the total amount of uncertainty in the target variable that the observed variables can explain. The normalization formula is shown in equation (11):

[0168]

[0169] (11);

[0170] in, This represents the set of all combinations that contain more than one variable. This represents the total number of observed variables to be studied. It directly represents the proportion of this type of causal contribution in "information explained by all observable variables", and its range is [0, 1]; For target variable With observation information extended matrix Total mutual information between them.

[0171] Corresponding to step S62, the causal leakage component is... Divide by the target variable Shannon entropy of itself The entropy of the target variable represents its inherent total uncertainty. The normalization formula is shown in equation (12):

[0172] (12);

[0173] in, Target variable Its own Shannon entropy.

[0174] Normalized causal leakage satisfaction Its physical meaning is: the proportion of system uncertainty that is not explained by the current set of observed variables. The closer the value is to 1, the weaker the explanatory power of the observed variables on the target variable, and the greater the influence of unknown or unmodeled factors.

[0175] S7. Based on the normalized global causal contribution identification system's controlling factors and collaborative mechanisms, generate optimization suggestions.

[0176] Based on the normalized global causal component contribution obtained in step S6, the importance ranking of each factor to the system objective and the essence of their interaction mechanism can be systematically and clearly revealed. This provides direct and reliable data-driven decision support for the optimized design of thermochemical energy storage reaction units, the formulation of operating strategies, and the determination of subsequent in-depth research directions.

[0177] like Figure 8 As shown, in a preferred embodiment, S7 includes:

[0178] S71, by comparing the normalized global causal contributions, unique controlling factors, synergistic reinforcement mechanisms, and redundant variable combinations are identified, and the modeling completeness of the thermochemical energy storage system is evaluated. The criteria for identifying unique controlling factors and synergistic mechanisms include: if the normalized unique contribution of a variable is significantly higher than that of other variables, then the variable is inferred to be a unique controlling factor; if the normalized synergistic contribution of a variable combination is significantly greater than zero, then the combination is identified to have a synergistic reinforcement mechanism; if the normalized redundant contribution of a variable combination is significant, then it indicates that the information of these variables overlaps and can be used to guide model simplification; if the normalized causal leakage is high, then it suggests the existence of important unobserved variables and the need for further model refinement or experimental observation.

[0179] In some embodiments of this example, the main controlling factors and cooperative mechanism identification methods of the thermochemical energy storage reaction unit system are defined as follows:

[0180] (1) Identification of unique controlling factors: comparing each variable one by one Unique contribution to normalization If the normalized unique contribution value of a certain variable is significantly higher than the contribution values ​​of all other variables, then it can be inferred that the variable is the unique controlling factor of the thermochemical energy storage system.

[0181] (2) Identification of co-reinforcement mechanisms: Examine all multivariate combinations Normalized synergistic contribution If one or more combinations Values ​​significantly greater than zero (e.g., exceeding 0.1 or dominating all synergistic contributions) indicate the existence of a "synergistic reinforcement mechanism" among these variables, which can be used to reveal deep coupling relationships between multiphysics fields, such as how specific spatial configurations of temperature and concentration fields jointly determine local reaction hotspots;

[0182] (3) Information redundancy identification and model simplification guidelines: focus on normalized redundancy contribution This is especially true for combinations involving a large number of variables. If the redundancy contribution of a combination is significant, it indicates that the information provided by these variables is highly overlapping. This information can be used to guide the selection of representative variables from the redundant variable group when building predictive models or designing monitoring systems, thereby reducing model complexity or the number of sensors.

[0183] (4) System completeness assessment and unobserved factors: Examining normalized causal leakage If the value is high (e.g., greater than 0.3), it indicates that the current set of observed variables is insufficient to explain the target, suggesting the existence of unmodeled key variables that require further modeling or experimental observation.

[0184] S72, based on the quantitative identification conclusions of the normalized global causal components, main control factors and synergistic mechanisms, combined with the physical mechanism of the reaction unit, provides optimization suggestions for the optimization design and operation strategy formulation of thermochemical energy storage reaction units, including: implementing precise control of unique main control factors, designing engineering schemes to promote the joint effect of synergistic variable groups, simplifying monitoring systems or simulation models based on redundancy analysis, and supplementing key variable observations or deepening mechanism research for high causal leakage indication directions.

[0185] In this embodiment, the optimization suggestions generated for the thermochemical energy storage reaction unit include, but are not limited to, the following optimization suggestions: if temperature and water vapor concentration contribute significantly in synergy, it is recommended to design a gradient thermo-mass coupling supply strategy; if a certain variable makes a unique and prominent contribution, the control precision of that variable can be enhanced in a targeted manner; if causal leakage is high, it is recommended to introduce a microstructure evolution model or supplement experimental observations.

[0186] Example 2

[0187] like Figure 9 As shown, this embodiment provides a system for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system, used to implement the method of Embodiment 1, including:

[0188] The structured multi-field data matrix construction module 101 is used to acquire the multi-field coupling historical data of the reaction unit in the thermochemical energy storage system, and determine the structured multi-field data matrix based on the multi-field coupling historical data; wherein, the multi-field coupling historical data comes from numerical simulation output or experimental monitoring system, and includes time series data under one or more operating conditions;

[0189] The structured variable set construction module 102 is used to construct a structured variable set, which includes a target variable set and an observation variable set, and to form an observation information extension matrix based on the observation variable set, wherein the observation information extension matrix includes the observation variable at the current time and the historical state of the observation variable at the current time;

[0190] The mutual information and specific mutual information calculation module 103 is used to calculate mutual information and specific mutual information based on the target variable and the set of observed variables; wherein the specific mutual information is the input of fine causal decomposition;

[0191] The refined causal decomposition module 104 is used to perform refined causal decomposition, including: for each target state, performing a cooperative-unique-redundant decomposition on the specific mutual information based on incremental sorting and recursive allocation to obtain the decomposition result, decomposing the specific mutual information into redundant causal contributions, unique causal contributions, cooperative causal contributions, and causal leakage components; wherein the redundant causal contributions represent the causal contributions of all possible non-empty subsets of the observation information expansion matrix to the target variable in a certain possible state; the unique causal contributions represent information about a certain possible state that is unique to a single variable and cannot be obtained from any other single variable in the set of observed variables; the cooperative causal contributions represent additional information about a certain possible state that emerges only when multiple variables in a variable combination are jointly observed, and cannot be obtained from any subset of the variable combination; the causal leakage components represent the part of information about the target variable in a certain possible state that is jointly participated in by the entire observation information expansion matrix but cannot be explained, the part of information originating from latent variables not included in the current observation or purely random fluctuations;

[0192] The full-state-space causal contribution integration module 105 is used to integrate the causal contributions of the full-state space based on the decomposition results.

[0193] The global causal contribution processing module 106 is used to normalize the global causal contribution.

[0194] The main control factor and coordination mechanism analysis module 107 is used to generate optimization suggestions based on the main control factors and coordination mechanisms of the global causal contribution identification system after normalization processing.

[0195] The present invention also provides a memory that stores multiple instructions for implementing the method as described in Embodiment 1.

[0196] like Figure 10 As shown, the present invention also provides an electronic device, including a processor 301 and a memory 302 connected to the processor 301. The memory 302 stores a plurality of instructions, which can be loaded and executed by the processor to enable the processor to perform methods as described in Embodiments 2 and 3.

[0197] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing the controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system, based on causal component decomposition to achieve causal inference and quantification, characterized in that... include: S1, acquire multi-field coupling historical data of reaction units in thermochemical energy storage system, and determine a structured multi-field data matrix based on the multi-field coupling historical data; wherein, the multi-field coupling historical data comes from numerical simulation output or experimental monitoring system, and includes time series data under one or more operating conditions; S2, construct a structured variable set, which includes a target variable set and an observation variable set, and form an observation information extension matrix based on the observation variable set, wherein the observation information extension matrix includes the observation variables at the current time and the historical state of the observation variables at the current time; S3, calculate mutual information and specific mutual information based on the target variable and the set of observed variables; wherein the specific mutual information is the input of fine causal decomposition; S4, perform refined causal decomposition, including: for each target state, perform cooperative-unique-redundant decomposition on the specific mutual information based on incremental sorting and recursive allocation to obtain the decomposition result, decomposing the specific mutual information into redundant causal contribution, unique causal contribution, cooperative causal contribution, and causal leakage component; wherein the redundant causal contribution represents the causal contribution of all possible non-empty subsets of the observation information expansion matrix to the target variable in a certain possible state; the unique causal contribution represents information about a certain possible state that is unique to a single variable and cannot be obtained from any other single variable in the set of observed variables; the cooperative causal contribution represents additional information about a certain possible state that emerges only when multiple variables in the variable combination are jointly observed, and cannot be obtained from any subset of the variable combination; the causal leakage component represents the part of information about the target variable in a certain possible state that is jointly participated in by the entire observation information expansion matrix but cannot be explained, the part of information originating from latent variables not included in the current observation or purely random fluctuations; S5, Integrate the causal contributions of the entire state space based on the decomposition results; S6, normalize the global causal contribution; S7. Based on the normalized global causal contribution identification system's main control factors and collaborative mechanisms, optimization suggestions are generated.

2. The method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system according to claim 1, characterized in that, S1 includes: S11, acquire multi-field coupling historical data of the reaction unit in the thermochemical energy storage system; the multi-field coupling historical data includes: temperature field information, water vapor concentration field information, flow velocity field information, pore structure parameters, reaction progress and external heat and mass input conditions; wherein the external heat and mass input conditions include boundary heat flux density, boundary vapor pressure, energy storage efficiency and energy release power; S12, preprocessing the multi-field coupling historical data, including: for the data output from the numerical simulation, deriving the evolution data of each physical field at discrete spatiotemporal nodes from the established fluid-thermal-chemical multi-field coupling cross-scale model, which can analyze the key processes from the pore scale to the reaction unit scale; for the data from the experimental monitoring system, synchronously acquiring time-series signals through a sensor network deployed within the reaction unit; performing high-dimensional data feature extraction operations on the evolution data of each physical field at discrete spatiotemporal nodes and the time-series signals, and using statistical methods to compress the spatially distributed high-dimensional data of the temperature field, concentration field, and velocity field to extract feature values ​​that can represent the overall state of the physical field; S13, determining a structured multi-field data matrix based on the preprocessed multi-field coupled historical data, including: performing time interpolation and spatial alignment preprocessing on the sensor data of each sensor in the sensor network to form the spatiotemporally matched structured multi-field data matrix; and performing independent normalization processing on the variable data of each dimension in the structured multi-field data matrix.

3. The method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system according to claim 2, characterized in that, S2 includes: S21, Select the core performance indicators as the target variables; S22, Select relevant physical field variables to construct the set of observation variables, and form an observation information matrix containing historical information based on the set of observation variables.

4. The method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system according to claim 3, characterized in that, S3 includes: S31, Based on information theory, calculate the mutual information between the target variable and the set of observed variables; S32, based on the mutual information, further calculate the specific mutual information for each possible state of the target variable.

5. The method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system according to claim 4, characterized in that, S3 further includes: S33, calculate the expectation based on the specific mutual information of all possible states of the target variable to restore the total mutual information.

6. The method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system according to claim 5, characterized in that, S4 includes: S41, Initialize the remaining variable set, which contains all observation variable indices, including: creating a list containing all observation variable indices, called the remaining variable set, which will be dynamically updated during the allocation process; S42, calculate the specific mutual information of each non-empty subset in the observation information extended matrix with respect to the target state; S43, each non-empty subset is sorted in ascending order according to the value of specific mutual information, and the information increment between adjacent sorted non-empty subsets is calculated by subtracting the information of the previous subset from the information of the next subset. The information increment is used to characterize the information gain brought about by introducing new variables or combinations of variables on the basis of existing information. S44, recursively allocate the information increment, including: processing each non-empty subset and its corresponding information increment in sequence according to the sorting order in S43; S45, defining unique contributions, including: after completing the recursive allocation according to S44, in all redundant contribution dictionaries, the entries with a single variable as the key are redefined as the unique contributions corresponding to that variable. S46, the causal leakage component is obtained by subtracting the sum of all allocated redundant contributions and cooperative contributions from the total specific mutual information, wherein all allocated redundant contributions include items that have been converted into unique contributions.

7. The method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system according to claim 6, characterized in that, S5 includes: S51 integrates all possible target states into a full state space; S52, for each possible target state in the full state space, perform a probability-weighted average of the corresponding decomposition results to obtain the global causal contribution and global causal leakage that reflect the average causal strength of the multi-field coupled thermochemical energy storage system, wherein the global causal contribution includes global redundancy, uniqueness and cooperative contribution.

8. The method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system according to claim 7, characterized in that, S6 includes: S61, normalize the global causal contribution to a proportion relative to the total mutual information; S62 normalizes causal leakage to a ratio relative to the entropy of the target variable.

9. The method for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system according to claim 8, characterized in that, S7 includes: S71, by comparing the magnitude of the normalized global causal contribution, unique controlling factors, synergistic reinforcement mechanisms, and redundant variable combinations are identified, and the modeling completeness of the thermochemical energy storage system is evaluated. The criteria for identifying unique controlling factors and synergistic mechanisms include: if the normalized unique contribution of a variable is significantly higher than that of other variables, then the variable is inferred to be a unique controlling factor; if the normalized synergistic contribution of a variable combination is significantly greater than zero, then the combination is identified to have a synergistic reinforcement mechanism; if the normalized redundant contribution of a variable combination is significant, then it indicates that the information of these variables overlaps and can be used to guide model simplification; if the normalized causal leakage is high, then it suggests the existence of important unobserved variables and the need for further model improvement or experimental observation. S72, based on the quantitative identification conclusions of the normalized global causal components, main control factors and synergistic mechanisms, combined with the physical mechanism of the reaction unit, provides optimization suggestions for the optimization design and operation strategy formulation of thermochemical energy storage reaction units, including: implementing precise control of unique main control factors, designing engineering schemes to promote the joint effect of synergistic variable groups, simplifying monitoring systems or simulation models based on redundancy analysis, and supplementing key variable observations or deepening mechanism research for high causal leakage indication directions.

10. A system for analyzing the main controlling factors and synergistic mechanisms of a multi-field coupled thermochemical energy storage system, used to implement the method described in any one of claims 1-9, characterized in that, include: The structured multi-field data matrix construction module (101) is used to acquire the multi-field coupling historical data of the reaction unit in the thermochemical energy storage system, and to determine the structured multi-field data matrix based on the multi-field coupling historical data; wherein, the multi-field coupling historical data comes from the numerical simulation output or the experimental monitoring system, and includes time series data under one or more operating conditions; The structured variable set construction module (102) is used to construct a structured variable set, which includes a target variable set and an observation variable set, and forms an observation information extension matrix based on the observation variable set, wherein the observation information extension matrix includes the observation variable at the current time and the historical state of the observation variable at the current time; The mutual information and specific mutual information calculation module (103) is used to calculate mutual information and specific mutual information based on the target variable and the set of observed variables; wherein the specific mutual information is the input of fine causal decomposition; The refined causal decomposition module (104) is used to perform refined causal decomposition, including: for each target state, performing a cooperative-unique-redundant decomposition on the specific mutual information based on incremental sorting and recursive allocation to obtain the decomposition result, decomposing the specific mutual information into redundant causal contribution, unique causal contribution, cooperative causal contribution, and causal leakage component; wherein the redundant causal contribution represents the causal contribution of all possible non-empty subsets of the observation information expansion matrix to the target variable in a certain possible state; the unique causal contribution represents information about a certain possible state that is unique to a single variable and cannot be obtained from any other single variable in the set of observation variables; the cooperative causal contribution represents additional information about a certain possible state that emerges only when multiple variables in the variable combination are jointly observed, and cannot be obtained from any subset of the variable combination; the causal leakage component represents the part of information about the target variable in a certain possible state that is jointly participated in by the entire observation information expansion matrix but cannot be explained, the part of information originating from latent variables not included in the current observation or purely random fluctuations; The full-state-space causal contribution integration module (105) is used to integrate the causal contributions of the full-state space based on the decomposition results. The global causal contribution processing module (106) is used to normalize the global causal contribution. The main control factor and coordination mechanism analysis module (107) is used to generate optimization suggestions based on the main control factors and coordination mechanisms of the global causal contribution identification system after normalization.

Images