Simulation method and platform for at-211 online separation process based on digital twinning

Abstract

The application provides an At-211 online separation process simulation method and platform based on digital twinning, relates to the technical field of data processing, and identifies a separation behavior evolution mode based on a data change correlation relationship in a continuous time interval by collecting process monitoring data reflecting material distribution states and separation result changes in a separation process, and divides the separation process into multiple operation segments accordingly. Different digital twinning simulation sub-models are constructed for different separation behavior evolution modes, with different state variable composition modes and state change constraint conditions. In the online simulation process, the correspondence between the current separation behavior evolution mode and the digital twinning simulation sub-model is continuously determined, and when the correspondence is consistent, continuous simulation is maintained, and when the correspondence is inconsistent, the existing state is continued and switched to the matched digital twinning simulation sub-model, and the subsequent simulation evolution process is re-established, so that the simulation of the multi-stage operation state of the At-211 online separation process is realized.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a simulation method and platform for the At-211 online separation process based on digital twins. Background Technology

[0002] In existing technologies, digital twin technology has been widely applied to online simulation and optimization of industrial processes, especially in the modeling and real-time monitoring of continuous or discrete processes in chemical engineering and manufacturing. By constructing a simulation model in digital space corresponding to the physical device or process flow, and combining it with historical data and real-time sensor data, a synchronous mapping of the actual operating state can be achieved, thereby enabling process monitoring, operational prediction, and decision support. In the field of separation processes, for processes such as solvent extraction and adsorption separation, there are already digital twin-based systems used for visual monitoring and parameter simulation of separation processes to support operators in analyzing and managing the status of separation processes.

[0003] However, when applying the aforementioned digital twin online simulation technology to continuous separation processes, existing simulation models often struggle to effectively depict the actual separation behavior throughout the entire process, as the material state and separation behavior change with each stage of operation. Particularly when feed composition changes, separation medium gradually becomes saturated, or elution conditions switch, the evolutionary characteristics of the separation process undergo phased transformations. Existing digital twin systems typically continue to simulate the state evolution path formed in the previous separation stage, leading to the continuation of historical states no longer applicable to the current separation behavior characteristics into subsequent simulations. This results in systematic simulation biases at the boundaries between separation stages, affecting the accuracy of the simulation results in reflecting the actual separation process. Summary of the Invention

[0004] The purpose of this invention is to provide a simulation method and platform for the online separation process of At-211 based on digital twins, in order to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] Firstly, a simulation method for the online separation process of At-211 based on digital twins, the method comprising:

[0007] Process monitoring data was collected during the online separation process of At-211 to characterize the material distribution state and the changes in separation results. Based on the correlation of changes in process monitoring data over a continuous time interval, the state evolution characteristics of the separation process were analyzed, and the separation behavior evolution pattern that characterizes the evolution law of the separation process operation state was identified.

[0008] Based on different separation behavior evolution modes, the At-211 online separation process is divided into multiple separation process operation segments, and each separation process operation segment corresponds to a separation behavior evolution mode that remains relatively stable within a predetermined time range;

[0009] Based on each separation behavior evolution mode, a digital twin simulation sub-model is constructed to describe the state change relationship of the At-211 separation process under that separation behavior evolution mode. The different digital twin simulation sub-models differ from each other in the composition of process state variables and / or state change constraints.

[0010] During the online simulation of the digital twin, based on the real-time collected process monitoring data, it is continuously determined whether the evolution mode of the separation behavior corresponding to the current separation process corresponds to the running digital twin simulation sub-model.

[0011] When the two are consistent, the continuous simulation of the separation process state by the digital twin simulation sub-model is maintained; when the two are inconsistent, the continuation of the existing process state in the current digital twin simulation sub-model is terminated, and the model is switched to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode. Based on the process state variables and their state change constraints in the digital twin simulation sub-model, the process state of the subsequent separation process is re-deduced.

[0012] Preferably, based on the correlation of changes in process monitoring data over a continuous time interval, the state evolution characteristics of the separation process are analyzed, and the separation behavior evolution patterns characterizing the evolution law of the separation process's operating state are identified, including:

[0013] Multiple process monitoring parameters used to characterize the material distribution and changes in separation results are collected during the separation process according to a preset sampling period. The process monitoring parameters are sorted according to the collection time to form multi-parameter time series data reflecting the continuous operation status of the separation process.

[0014] Based on multi-parameter time series data, the direction, magnitude, and synchronicity of change of each process monitoring parameter are processed interval by interval within a continuous time interval to obtain joint change characteristics that characterize the coordinated change relationship of each process monitoring parameter within the corresponding time interval.

[0015] Based on the joint change characteristics, the state evolution trajectory formed by the change of process state over time during the separation process is analyzed, and a set of state evolution characteristics is generated to characterize the stability and consistency of the state evolution trajectory.

[0016] Based on the set of state evolution features, the state evolution trajectories in different time intervals are distinguished, and the state evolution trajectories with the same state evolution features are grouped into the same separation behavior evolution pattern, forming a separation behavior evolution pattern identification result used to characterize the evolution law of the separation process operation state.

[0017] Preferably, for each separation behavior evolution mode, a digital twin simulation sub-model is constructed to describe the state change relationship of the At-211 separation process under that separation behavior evolution mode. The different digital twin simulation sub-models differ from each other in the composition of process state variables and / or state change constraints, including:

[0018] Based on the identified separation behavior evolution pattern, analyze the participation of each process monitoring parameter in the state evolution trajectory corresponding to the separation behavior evolution pattern on the state change of the separation process, select process monitoring parameters that match the separation behavior evolution pattern, and map the selected process monitoring parameters to the corresponding process state variables to form a set of process state variables;

[0019] Based on the set of process state variables, the process state variables are organized according to their roles and mutual influences in the separation process, and a state variable organization structure is constructed to describe the state transfer relationship between process state variables.

[0020] Based on the organizational structure of state variables, determine the state change constraints, including the order of changes, boundaries, and mutual relationships of the state variables in the limiting process, and form state change rules applicable to the evolutionary pattern of this separation behavior.

[0021] A digital twin simulation sub-model corresponding to the evolution mode of the separated behavior is constructed by the set of process state variables, the organizational structure of state variables, and the constraints of state changes.

[0022] Preferably, when the two are inconsistent, the continuation of the existing process state in the current digital twin simulation sub-model is terminated, and the model is switched to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode. Based on the process state variables and their state change constraints in the digital twin simulation sub-model, the process state of the subsequent separation process is re-deduced, including:

[0023] After determining that the evolution mode of the separation behavior corresponding to the current separation process is inconsistent with the running digital twin simulation sub-model, stop the continued deduction of the separation process state of subsequent time steps based on the digital twin simulation sub-model, and block the transmission of the process state already formed in the digital twin simulation sub-model to the subsequent simulation process.

[0024] Based on the newly determined separation behavior evolution mode, a digital twin simulation sub-model corresponding to the separation behavior evolution mode is selected as the digital twin simulation sub-model used for subsequent online simulation of the separation process.

[0025] Based on the real-time collected process monitoring data, the process state variables in the selected digital twin simulation sub-model are initialized, and the process state of the subsequent separation process is gradually deduced under the corresponding state change constraints, thereby re-deducing the subsequent separation process.

[0026] Preferably, based on the joint change characteristics, the state evolution trajectory formed by the change of process state over time during the separation process is analyzed to generate a set of state evolution characteristics to characterize the stability and consistency of the state evolution trajectory, including:

[0027] Based on the joint change characteristics, the process states in the continuous time interval are connected in chronological order to form a state evolution trajectory consisting of multiple continuous process state points, where each process state point corresponds to the value result of the joint change characteristics in the corresponding time interval.

[0028] Based on the state evolution trajectory, the relationship between adjacent process state points is analyzed segment by segment to determine the degree of continuity of process state changes within adjacent time intervals, and to generate continuity analysis results to characterize the local continuity of the state evolution trajectory.

[0029] Based on the continuity analysis results, the overall direction and consistency of change of the state evolution trajectory in multiple continuous time intervals are analyzed, and trend consistency analysis results are generated to characterize the overall change law of the state evolution trajectory.

[0030] Based on the results of continuity analysis and trend consistency analysis, stability features for characterizing the stability of state evolution trajectories and consistency features for characterizing the consistency of state evolution trajectories are generated respectively. The stability features and consistency features are then combined to form a set of state evolution features.

[0031] Preferably, based on the set of state evolution features, the state evolution trajectories within different time intervals are distinguished, and state evolution trajectories with the same state evolution features are grouped into the same separation behavior evolution pattern, forming a separation behavior evolution pattern identification result used to characterize the evolution law of the separation process's operating state, including:

[0032] Based on the set of state evolution features corresponding to different time intervals, the stability and consistency features of each state evolution trajectory are aligned to form a feature description result that characterizes the features of each state evolution trajectory.

[0033] Based on the feature description results, the stability features and consistency features between different state evolution trajectories are compared one by one to determine the consistency relationship between the feature composition of each state evolution trajectory and generate the feature consistency determination result between the state evolution trajectories.

[0034] Based on the characteristic consistency determination results, state evolution trajectories with the same characteristics are grouped into the same trajectory set, and each trajectory set corresponds to a state evolution law.

[0035] Each trajectory set is identified as a separation behavior evolution pattern, and a correspondence is established between the separation behavior evolution pattern and the corresponding state evolution feature set to form the separation behavior evolution pattern recognition result.

[0036] Preferably, based on the identified separation behavior evolution pattern, the participation of each process monitoring parameter in the state evolution trajectory corresponding to the separation behavior evolution pattern on the state changes of the separation process is analyzed. Process monitoring parameters matching the separation behavior evolution pattern are selected, and the selected process monitoring parameters are mapped to corresponding process state variables to form a set of process state variables, including:

[0037] Based on the state evolution trajectory corresponding to the separation behavior evolution mode, extract the parameter change trajectory of each process monitoring parameter during the formation of the state evolution trajectory, and form the parameter change description result corresponding to each process monitoring parameter;

[0038] Based on the parameter change description results, the correspondence between each process monitoring parameter and the state change of the separation process during the formation of the state evolution trajectory is analyzed, and participation analysis results are generated to characterize the degree of participation of each process monitoring parameter in the state change of the separation process.

[0039] Based on the participation analysis results, process monitoring parameters that play a major role in characterizing the state changes of the separation process under the separation behavior evolution model were selected to form a candidate process monitoring parameter set;

[0040] Each process monitoring parameter in the candidate process monitoring parameter set is mapped to a corresponding process state variable based on its physical meaning and functional role in the separation process, forming a set of process state variables corresponding to the evolution mode of the separation behavior.

[0041] Secondly, a digital twin-based online separation process simulation platform for At-211, the platform comprising:

[0042] The separation behavior evolution pattern recognition module is used to collect process monitoring data that characterizes the material distribution state and separation result changes during the online separation process of At-211, and analyze the state evolution characteristics of the separation process based on the correlation of changes in process monitoring data within a continuous time interval, and identify the separation behavior evolution pattern that characterizes the evolution law of the separation process operation state.

[0043] The segment division module is used to divide the At-211 online separation process into multiple separation process segments according to different separation behavior evolution modes. Each separation process segment corresponds to a separation behavior evolution mode that remains relatively stable within a predetermined time range.

[0044] The digital twin simulation sub-model construction module is used to construct a digital twin simulation sub-model to describe the state change relationship of the At-211 separation process under each separation behavior evolution mode. Different digital twin simulation sub-models differ from each other in the composition of process state variables and / or state change constraints.

[0045] The separation behavior evolution mode determination module is used to continuously determine whether the separation behavior evolution mode corresponding to the current separation process corresponds to the running digital twin simulation sub-model based on real-time collected process monitoring data during the online simulation of the digital twin.

[0046] The simulation control module is used to maintain the continuous simulation of the separation process state by the digital twin simulation sub-model when the two are consistent; when the two are inconsistent, it terminates the continuation of the existing process state in the current digital twin simulation sub-model, switches to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode, and re-deduces the process state of the subsequent separation process based on the process state variables and their state change constraints in the digital twin simulation sub-model.

[0047] The above-described solution of the present invention has at least the following beneficial effects:

[0048] First, by introducing a separation behavior evolution pattern recognition method based on the correlation of changes in process monitoring data, digital twin simulation no longer relies solely on a single continuous state model to describe the separation process. Instead, it can start from the separation process operation data itself to characterize the evolution characteristics of material distribution and separation results over time, thus providing a basis for distinguishing the states of different operating stages in continuous separation processes.

[0049] Secondly, by dividing the online separation process into multiple separation process segments according to different separation behavior evolution modes, each segment corresponds to a separation behavior feature that remains relatively stable within a predetermined time range. This avoids mixing and processing inapplicable state evolution features from different operation stages, which helps improve the simulation process's ability to adapt to the phased changes in the continuous separation process.

[0050] Furthermore, by constructing digital twin simulation sub-models with different process state variable composition methods and state change constraints for different separation behavior evolution modes, the digital twin simulation sub-models can match the corresponding separation behavior characteristics at the state description level, avoiding the deviation caused by using a single simulation model to uniformly characterize multiple separation operation states, and improving the rationality of expressing the state change relationship of the separation process.

[0051] Furthermore, during the online simulation of the digital twin, by continuously determining the correspondence between the current separation behavior evolution mode and the running digital twin simulation sub-model, when the separation behavior evolution mode changes, the continuation of the existing process state is terminated and switched to the matching digital twin simulation sub-model, and the subsequent simulation evolution process is re-established. This avoids directly inheriting the state evolution path formed in the previous separation stage to the new separation stage, which helps to reduce the systematic simulation deviation at the boundary of the separation stages.

[0052] Through the aforementioned progressive processing method, the online digital twin simulation process can dynamically adjust the simulation evolution process as the separation process operation status changes, thereby improving the simulation results' ability to reflect the actual continuous separation process operation status and facilitating the analysis and understanding of the separation process. Attached Figure Description

[0053] Figure 1 This is a flowchart of the At-211 online separation process simulation method based on digital twin provided in the embodiments of the present invention. Detailed Implementation

[0054] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0055] like Figure 1 As shown, embodiments of the present invention propose a simulation method for the online separation process of At-211 based on digital twins, the method comprising:

[0056] Process monitoring data was collected during the online separation process of At-211 to characterize the material distribution state and the changes in separation results. Based on the correlation of changes in process monitoring data over a continuous time interval, the state evolution characteristics of the separation process were analyzed, and the separation behavior evolution pattern that characterizes the evolution law of the separation process operation state was identified.

[0057] Based on different separation behavior evolution modes, the At-211 online separation process is divided into multiple separation process operation segments, and each separation process operation segment corresponds to a separation behavior evolution mode that remains relatively stable within a predetermined time range;

[0058] Based on each separation behavior evolution mode, a digital twin simulation sub-model is constructed to describe the state change relationship of the At-211 separation process under that separation behavior evolution mode. The different digital twin simulation sub-models differ from each other in the composition of process state variables and / or state change constraints.

[0059] During the online simulation of the digital twin, based on the real-time collected process monitoring data, it is continuously determined whether the evolution mode of the separation behavior corresponding to the current separation process corresponds to the running digital twin simulation sub-model.

[0060] When the two are consistent, the continuous simulation of the separation process state by the digital twin simulation sub-model is maintained; when the two are inconsistent, the continuation of the existing process state in the current digital twin simulation sub-model is terminated, and the model is switched to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode. Based on the process state variables and their state change constraints in the digital twin simulation sub-model, the process state of the subsequent separation process is re-deduced.

[0061] In this embodiment of the invention, by introducing continuous perception and analysis of the operating status of the At-211 online separation process in the digital space, the simulation process no longer relies on a single continuous state evolution path. Instead, it can characterize the state evolution characteristics of the separation process based on the changes in material distribution and separation results over time, thereby identifying the separation behavior evolution pattern that reflects the evolution law of the separation process's operating status. This enables the simulation process to identify the stage-by-stage changes in the separation process, providing a foundation for subsequent segmented processing in the simulation process.

[0062] Based on the identified different separation behavior evolution patterns, the At-211 online separation process is divided into multiple separation process segments, with each segment corresponding to a relatively stable separation behavior evolution pattern within a predetermined time range. In this way, the originally continuous separation process, which involves internal state transitions, is broken down into multiple operating intervals with consistent evolutionary characteristics. This avoids carrying over historical evolutionary characteristics that are not applicable to the current operating state into subsequent simulations, thereby improving the adaptability of the simulation process to the actual separation process operating state.

[0063] For different separation behavior evolution modes, corresponding digital twin simulation sub-models are constructed to match the corresponding separation behavior evolution mode in terms of the composition of process state variables and state change constraints. By introducing multiple digital twin simulation sub-models with different structures, the simulation process can adopt different state description methods and evolution rules for different separation behavior evolution characteristics, thereby avoiding simulation bias caused by using a single model to uniformly describe all operating states.

[0064] During the online digital twin simulation, the evolution mode of the separation behavior corresponding to the current separation process is continuously determined based on the real-time collected process monitoring data. When the determination result indicates that the current separation process is still in a separation behavior evolution mode that matches the running digital twin simulation sub-model, the existing digital twin simulation sub-model is maintained to continuously simulate the state of the separation process, so that the simulation process can smoothly reflect the state evolution of the separation process within this running segment.

[0065] When the determination result indicates that the evolutionary pattern of the separation behavior corresponding to the current separation process has changed and is inconsistent with the currently running digital twin simulation sub-model, the continuation of the existing process state in the current digital twin simulation sub-model is terminated, and the system switches to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolutionary pattern. Simultaneously, the simulation evolution process for subsequent separation processes is re-established. By reconstructing the simulation state when the separation behavior evolutionary pattern changes, the system avoids directly inheriting the state evolution path formed in the previous running segment to the new running segment, thereby reducing simulation deviations at the boundary of separation stages.

[0066] For example, in the continuous separation process of At-211, when the feed composition changes or the separation medium gradually reaches saturation, the material distribution and separation results exhibit different characteristics over time compared to the initial operating stage. In this embodiment of the invention, through continuous analysis of process monitoring data, the change in the separation behavior evolution mode corresponding to this change can be identified, and the separation process can be divided into different operating segments. In the initial operating segment, a digital twin simulation sub-model matching the initial separation behavior evolution mode is used for state evolution simulation; when the separation behavior evolution mode changes, the corresponding new digital twin simulation sub-model is switched in a timely manner and the simulation evolution process is re-established, so that the entire online simulation process can maintain consistency with the actual operating state of the At-211 separation process at different operating stages, facilitating the analysis and understanding of the separation process.

[0067] In a preferred embodiment of the present invention, the At-211 online separation process is divided into multiple separation process segments according to different separation behavior evolution modes. Each separation process segment corresponds to a separation behavior evolution mode that remains relatively stable within a predetermined time range, including:

[0068] After identifying the separation behavior evolution pattern, the time axis of continuous operation during the online separation process is scanned, and the separation behavior evolution pattern corresponding to each moment is used as the running status identifier at that moment.

[0069] According to the time sequence, time intervals in which the evolutionary patterns of separate behaviors remain consistent across multiple consecutive time points are merged to form a continuous time period;

[0070] A continuous time period is defined as a segment of the separation process, in which the evolution pattern of separation behavior remains unchanged within the segment of the separation process.

[0071] When a change in the separation behavior evolution pattern is detected, the current separation process segment ends, and the time point corresponding to the new separation behavior evolution pattern is used as the starting point of the next separation process segment.

[0072] By using the above method, the entire At-211 online separation process is divided into multiple separation process segments arranged in chronological order, so that each segment corresponds to a stable separation behavior evolution mode.

[0073] In a preferred embodiment of the present invention, during the online simulation of a digital twin, based on real-time collected process monitoring data, it is continuously determined whether the evolution mode of the separation behavior corresponding to the current separation process corresponds to the running digital twin simulation sub-model, including:

[0074] During the online simulation of the digital twin, process monitoring data at the current moment is acquired in real time according to a preset update cycle;

[0075] Based on the process monitoring data at the current moment, the separation behavior evolution mode corresponding to the current separation process is determined according to the identification method of separation behavior evolution mode;

[0076] Obtain the separation behavior evolution pattern corresponding to the currently running digital twin simulation sub-model;

[0077] Compare the separation behavior evolution pattern corresponding to the current separation process with the separation behavior evolution pattern corresponding to the current digital twin simulation sub-model;

[0078] When the two are consistent, it is determined that the current digital twin simulation sub-model is still applicable to the current separation process operation state; when the two are inconsistent, it is determined that the current digital twin simulation sub-model is no longer applicable to the current separation process operation state, and subsequent simulation switching processing is triggered.

[0079] In a preferred embodiment of the present invention, the predetermined time range is a time interval used to determine whether the evolutionary pattern of separation behavior remains stable, including:

[0080] In the process of identifying the evolutionary pattern of separation behavior, in order to avoid the impact of instantaneous data fluctuations on the determination of separation behavior, a minimum duration for determining the stability of the evolutionary pattern of separation behavior is preset.

[0081] When a certain separation behavior evolution pattern is identified as the same pattern in multiple consecutive time points and the duration reaches the minimum duration, it is determined that the separation behavior evolution pattern remains stable in the corresponding time interval.

[0082] The time interval during which the duration meets the stability requirements is used as a predetermined time range to form a segment of the separation process operation;

[0083] When the evolution pattern of separation behavior changes and the duration does not reach the minimum duration, a new separation process segment is not immediately divided. Instead, subsequent time points are judged until the stability requirements are met before the segment is divided.

[0084] By using the above methods, the division of the separation process segments is based on the stable maintenance of the separation behavior evolution pattern within a certain time range, thereby avoiding frequent switching of operation segments due to short-term fluctuations.

[0085] In a preferred embodiment of the present invention, based on the correlation of changes in process monitoring data over a continuous time interval, the state evolution characteristics of the separation process are analyzed, and the separation behavior evolution pattern characterizing the evolution law of the separation process's operating state is identified, including:

[0086] Multiple process monitoring parameters used to characterize the material distribution and changes in separation results are collected during the separation process according to a preset sampling period. The process monitoring parameters are sorted according to the collection time to form multi-parameter time series data reflecting the continuous operation status of the separation process.

[0087] Based on multi-parameter time series data, the direction, magnitude, and synchronicity of change of each process monitoring parameter are processed interval by interval within a continuous time interval to obtain joint change characteristics that characterize the coordinated change relationship of each process monitoring parameter within the corresponding time interval.

[0088] Based on the joint change characteristics, the state evolution trajectory formed by the change of process state over time during the separation process is analyzed, and a set of state evolution characteristics is generated to characterize the stability and consistency of the state evolution trajectory.

[0089] Based on the set of state evolution features, the state evolution trajectories in different time intervals are distinguished, and the state evolution trajectories with the same state evolution features are grouped into the same separation behavior evolution pattern, forming a separation behavior evolution pattern identification result used to characterize the evolution law of the separation process operation state.

[0090] In this embodiment of the invention, the state evolution characteristics of the separation process are analyzed layer by layer based on the correlation of changes in process monitoring data over continuous time intervals. This allows the identification of the separation process's operating state to no longer rely on preset process stages or manual experience divisions, but rather to be naturally formed by the synergistic changes of process monitoring parameters over time. By constructing a set of multi-parameter time series, joint change characteristics, and state evolution characteristics, the state evolution trajectories of the separation process in different time intervals can be distinguished and classified, thereby obtaining a separation behavior evolution pattern that reflects the evolution law of the separation process's operating state. This approach allows the identification results of separation behavior to directly originate from the operating data of the separation process itself, improving the ability to characterize the state change characteristics of the separation process.

[0091] In a preferred embodiment of the present invention, multiple process monitoring parameters for characterizing the material distribution state and changes in separation results are collected during the separation process according to a preset sampling period, and the process monitoring parameters are sorted according to the collection time to form multi-parameter time series data reflecting the continuous operation state of the separation process, including:

[0092] During the operation of the At-211 online separation process, several process monitoring parameters that need to be used to characterize the operating status of the separation process are determined. The process monitoring parameters include at least the operating parameters that can reflect changes in material distribution and separation results.

[0093] According to the set sampling period, the numerical information of each process monitoring parameter at each sampling moment is synchronously acquired at that moment.

[0094] Multiple process monitoring parameters acquired at the same sampling time are treated as a set of process monitoring data, and the corresponding acquisition time is labeled for this set of process monitoring data.

[0095] According to the chronological order of the collection time, multiple sets of process monitoring data obtained at consecutive sampling times are arranged so that each process monitoring parameter forms a continuous data record in the time dimension.

[0096] The above methods generate multi-parameter time series data that reflects the changes of monitoring parameters of each process during continuous operation of the separation process, providing a data foundation for subsequent correlation analysis.

[0097] In a preferred embodiment of the present invention, based on multi-parameter time series data, the direction, magnitude, and synchronicity of change of each process monitoring parameter are processed interval by interval within a continuous time interval to obtain joint change characteristics characterizing the coordinated change relationship of each process monitoring parameter within the corresponding time interval, including:

[0098] Multi-parameter time series data is divided into several consecutive time intervals according to time order, and each time interval contains multiple adjacent sampling results;

[0099] Within each time interval, the changes in the values ​​of each process monitoring parameter at the beginning and end of the interval are compared to determine the direction of change of the process monitoring parameter within that time interval.

[0100] Based on determining the direction of change, the degree of change of each process monitoring parameter within the time interval is analyzed to characterize the magnitude of change of the process monitoring parameter within the time interval.

[0101] By comparing and analyzing the direction and magnitude of change of multiple process monitoring parameters within the same time interval, it is possible to determine whether the process monitoring parameters show a related trend of change within the same time interval, thereby obtaining the synchronicity of change that reflects the coordinated change of multiple process monitoring parameters;

[0102] By combining the direction, magnitude, and synchronicity of change of each process monitoring parameter within each time interval, a joint change characteristic is formed to characterize the coordinated change relationship of each process monitoring parameter within that time interval.

[0103] In a preferred embodiment of the present invention, the preset sampling period is a time interval used to determine the acquisition frequency of process monitoring parameters, including:

[0104] Before the At-211 online separation process begins, the sampling time interval for collecting process monitoring parameters is determined based on the separation process running speed and the time scale of changes in process monitoring parameters.

[0105] The sampling time interval should cover the time range during which the material distribution and separation results change during the separation process, thereby ensuring that continuous sampling data can reflect the changing trend of the separation process.

[0106] During the separation process, process monitoring parameters are periodically collected according to sampling time intervals to ensure that the process monitoring data between adjacent sampling times remain continuous in time.

[0107] By using a fixed sampling period, the resulting multi-parameter time series data has a consistent interval structure in the time dimension, which facilitates subsequent analysis and processing of the relationship between process monitoring parameters over continuous time intervals.

[0108] In a preferred embodiment of the present invention, a digital twin simulation sub-model is constructed to describe the state change relationship of the At-211 separation process under each separation behavior evolution mode. The different digital twin simulation sub-models differ from each other in the composition of process state variables and / or the constraints on state changes, including:

[0109] Based on the identified separation behavior evolution pattern, analyze the participation of each process monitoring parameter in the state evolution trajectory corresponding to the separation behavior evolution pattern on the state change of the separation process, select process monitoring parameters that match the separation behavior evolution pattern, and map the selected process monitoring parameters to the corresponding process state variables to form a set of process state variables;

[0110] Based on the set of process state variables, the process state variables are organized according to their roles and mutual influences in the separation process, and a state variable organization structure is constructed to describe the state transfer relationship between process state variables.

[0111] Based on the organizational structure of state variables, determine the state change constraints, including the order of changes, boundaries, and mutual relationships of the state variables in the limiting process, and form state change rules applicable to the evolutionary pattern of this separation behavior.

[0112] A digital twin simulation sub-model corresponding to the evolution mode of the separated behavior is constructed by the set of process state variables, the organizational structure of state variables, and the constraints of state changes.

[0113] In this embodiment of the invention, digital twin simulation sub-models are constructed according to different separation behavior evolution modes, ensuring that the composition of state variables and the constraints of state changes in the digital twin simulation sub-models match the corresponding separation behavior evolution characteristics. By analyzing the participation of process monitoring parameters in the state changes of the separation process under each separation behavior evolution mode, and determining the process state variables and their organizational structure accordingly, different digital twin simulation sub-models exhibit differences in state description, thereby avoiding the use of a single simulation model to uniformly describe all separation operation states. This approach enables the simulation model to reflect the state change relationships of the separation process separately for different separation behavior evolution modes, improving the adaptability of the simulation process to the multi-stage operational characteristics of the separation process.

[0114] In a preferred embodiment of the present invention, based on the set of process state variables, the process state variables are organized according to their roles and mutual influences in the separation process to construct a state variable organization structure for describing the state transfer relationships between process state variables, including:

[0115] After obtaining the set of process state variables that match a certain separation behavior evolution pattern, analyze the functional role of each process state variable in the separation process, and clarify the role of each process state variable in material transfer, separation action and result formation.

[0116] Based on the order of action of each process state variable in the separation process, the direction of state transmission between process state variables is determined so that state changes can be transmitted according to the actual operating logic of the separation process.

[0117] Further analysis was conducted to determine whether there were any mutual constraints or influences among the process state variables, and process state variables with correlations were identified as related variables.

[0118] Based on the above analysis results, the process state variables are hierarchically or grouped to form a state variable organization structure that can reflect the state transmission path and interrelationship between process state variables, which is used to describe the overall composition of the separated process states.

[0119] In a preferred embodiment of the present invention, based on the organizational structure of state variables, state change constraints that define the order of changes, boundaries, and mutual relationships of process state variables are determined, forming state change rules applicable to the evolutionary pattern of the separated behavior, including:

[0120] Based on the constructed state variable organization structure, the order in which each process state variable is allowed to change during the separation process is determined, so that the state changes conform to the actual operation flow of the separation process;

[0121] By combining the operating conditions and technological constraints of the separation process, the range of values ​​that each process state variable is allowed to change under the corresponding separation behavior evolution mode is determined, thereby limiting the boundary of the process state variable change.

[0122] Based on the mutual influence relationships between process state variables, determine the constraint relationships on other process state variables when a certain process state variable changes, so as to avoid conflicts between state changes that do not conform to the actual separation process.

[0123] The order of changes, boundaries of changes, and mutual constraints are uniformly organized to form state change rules used to limit the way process state variables change.

[0124] In a preferred embodiment of the present invention, a digital twin simulation sub-model corresponding to the separation behavior evolution mode is constructed from the set of process state variables, the organizational structure of the state variables, and the constraints of state changes, including:

[0125] Based on the set of process state variables, a state description framework is established to characterize the operating state of the separated process, so that each process state variable can be used as a state description element in the digital twin simulation sub-model.

[0126] Introducing the state variable organization structure into the state description framework is used to determine the state transfer paths and interrelationships between process state variables;

[0127] The state change rules are loaded into the digital twin simulation sub-model to constrain the way and order of change of state variables in each process during the simulation.

[0128] By means of the above methods, a digital twin simulation sub-model is constructed that matches the specific separation behavior evolution mode, so that the digital twin simulation sub-model can reflect the changing relationship of the separation process state under the separation behavior evolution mode.

[0129] In a preferred embodiment of the present invention, when the two are inconsistent, the continuation of the existing process state in the current digital twin simulation sub-model is terminated, and the system switches to a digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode. Based on the process state variables and their state change constraints in the digital twin simulation sub-model, the process state of the subsequent separation process is re-deduced, including:

[0130] After determining that the evolution mode of the separation behavior corresponding to the current separation process is inconsistent with the running digital twin simulation sub-model, stop the continued deduction of the separation process state of subsequent time steps based on the digital twin simulation sub-model, and block the transmission of the process state already formed in the digital twin simulation sub-model to the subsequent simulation process.

[0131] Based on the newly determined separation behavior evolution mode, a digital twin simulation sub-model corresponding to the separation behavior evolution mode is selected as the digital twin simulation sub-model used for subsequent online simulation of the separation process.

[0132] Based on the real-time collected process monitoring data, the process state variables in the selected digital twin simulation sub-model are initialized, and the process state of the subsequent separation process is gradually deduced under the corresponding state change constraints, thereby re-deducing the subsequent separation process.

[0133] In this embodiment of the invention, by continuously determining the correspondence between the current separation behavior evolution mode and the running digital twin simulation sub-model during the online digital twin simulation, the simulation process can promptly perceive changes in the separation process's operating state. When the separation behavior evolution mode changes, the continuation of the existing process state is terminated and switched to the corresponding digital twin simulation sub-model, while simultaneously re-establishing the simulation evolution process. This avoids directly carrying over the state evolution relationship formed in the previous operating state to the new operating state. This method enables the simulation process to re-match the new state evolution rules when the separation behavior evolution mode switches, thereby reducing simulation deviations caused by state mismatches at the boundary of separation stages.

[0134] In a preferred embodiment of the present invention, after determining that the evolution mode of the separation behavior corresponding to the current separation process is inconsistent with the running digital twin simulation sub-model, the continued deduction of the separation process state of subsequent time steps based on the digital twin simulation sub-model is stopped, and the transmission of the process state already formed in the digital twin simulation sub-model to the subsequent simulation process is blocked, including:

[0135] During the online simulation of the digital twin, when the current separation behavior evolution mode identified based on real-time process monitoring data is inconsistent with the separation behavior evolution mode corresponding to the currently running digital twin simulation sub-model, a simulation switching trigger signal is generated.

[0136] Upon receiving the simulation switching trigger signal, the current digital twin simulation sub-model immediately terminates its state inference operation for subsequent time points, so that the current digital twin simulation sub-model no longer generates new simulation states based on existing states.

[0137] At the same time, the process states that have been formed in the current digital twin simulation sub-model are frozen so that the process states are no longer used as the initial state or reference state for subsequent simulation calculations.

[0138] By using the above method, the state continuation path of the current digital twin simulation sub-model is interrupted, preventing the state evolution results formed under the previous separation behavior evolution mode from being carried into the subsequent mismatched simulation process.

[0139] In a preferred embodiment of the present invention, based on the newly determined separation behavior evolution mode, a digital twin simulation sub-model corresponding to the separation behavior evolution mode is selected as the digital twin simulation sub-model used for subsequent online simulation of the separation process, including:

[0140] After re-evaluating the current separation behavior evolution pattern, the identified separation behavior evolution pattern will be used as the basis for selecting the simulation model.

[0141] In a pre-established set of digital twin simulation sub-models, a digital twin simulation sub-model that corresponds to the evolutionary pattern of the separation behavior is retrieved.

[0142] When a digital twin simulation sub-model that matches the evolution pattern of separation behavior is retrieved, that digital twin simulation sub-model is set as the digital twin simulation sub-model used in subsequent online simulations.

[0143] By using the above methods, the digital twin simulation sub-model used in the subsequent simulation process can be kept consistent with the current separation behavior evolution mode in terms of the composition of state variables and the constraints of state changes.

[0144] In a preferred embodiment of the present invention, based on real-time collected process monitoring data, the process state variables in the selected digital twin simulation sub-model are initialized, and under the constraints of corresponding state changes, the process state of the subsequent separation process is progressively deduced, thereby re-deducing the subsequent separation process, including:

[0145] After selecting a new digital twin simulation sub-model, obtain real-time process monitoring data corresponding to the current time point;

[0146] Based on real-time process monitoring data, the monitoring data corresponding to the process state variables in the digital twin simulation sub-model are mapped to initial state values, which are used to initialize each process state variable in the digital twin simulation sub-model.

[0147] After the state initialization is completed, the separation process state at subsequent time points is gradually deduced according to the pre-set state variable organization structure and state change constraints in the digital twin simulation sub-model.

[0148] By using the above method, the state evolution path of the separation process is re-established under the new separation behavior evolution mode, so that the simulation process can reflect the state change relationship under the current separation process operation state.

[0149] In a preferred embodiment of the present invention, based on the joint change characteristics, the state evolution trajectory formed by the change of process state over time during the separation process is analyzed to generate a set of state evolution characteristics characterizing the stability and consistency of the state evolution trajectory, including:

[0150] Based on the joint change characteristics, the process states in the continuous time interval are connected in chronological order to form a state evolution trajectory consisting of multiple continuous process state points, where each process state point corresponds to the value result of the joint change characteristics in the corresponding time interval.

[0151] Based on the state evolution trajectory, the relationship between adjacent process state points is analyzed segment by segment to determine the degree of continuity of process state changes within adjacent time intervals, and to generate continuity analysis results to characterize the local continuity of the state evolution trajectory.

[0152] Based on the continuity analysis results, the overall direction and consistency of change of the state evolution trajectory in multiple continuous time intervals are analyzed, and trend consistency analysis results are generated to characterize the overall change law of the state evolution trajectory.

[0153] Based on the results of continuity analysis and trend consistency analysis, stability features for characterizing the stability of state evolution trajectories and consistency features for characterizing the consistency of state evolution trajectories are generated respectively. The stability features and consistency features are then combined to form a set of state evolution features.

[0154] In this embodiment of the invention, the state evolution trajectory formed by the change of process state over time during the separation process is analyzed based on joint change characteristics. State evolution characteristics are extracted from both local continuity and overall change patterns, so that the changes in the operating state of the separation process are no longer merely manifested as discrete parameter fluctuations, but are transformed into a set of trajectory features that reflect the characteristics of state evolution. By characterizing the stability and consistency of the state evolution trajectory, the inherent laws of the state evolution of the separation process within different time intervals are clearly depicted, thus providing a structured state feature basis for subsequent differentiation and determination of the separation behavior evolution pattern.

[0155] In a preferred embodiment of the present invention, based on the state evolution trajectory, the changing relationship between adjacent process state points is analyzed segment by segment to determine the degree of continuity of process state changes within adjacent time intervals, generating a continuity analysis result for characterizing the local continuity of the state evolution trajectory, including:

[0156] After obtaining the state evolution trajectory consisting of multiple continuous process state points, two adjacent process state points are selected as an analysis unit in chronological order.

[0157] Compare the changes in the values ​​of the corresponding process state variables at two adjacent process state points, and analyze whether the process state variables remain stable within adjacent time intervals;

[0158] When most process state variables change in the same direction and degree in adjacent time intervals, the adjacent time intervals are determined to be time intervals of continuous state change.

[0159] The above analysis is performed sequentially on all adjacent time intervals in the state evolution trajectory, and the results are summarized to obtain the continuity analysis results that reflect the continuous changes of the state evolution trajectory within a local time range, which are used to describe the local continuity characteristics of the state evolution trajectory.

[0160] In a preferred embodiment of the present invention, based on the continuity analysis results, the overall direction and consistency of change of the state evolution trajectory over multiple continuous time intervals are analyzed to generate trend consistency analysis results characterizing the overall change pattern of the state evolution trajectory, including:

[0161] After obtaining the continuity analysis results corresponding to the state evolution trajectory, the continuity determination results in multiple consecutive time intervals are combined and analyzed.

[0162] Based on combinatorial analysis, it is determined whether the state evolution trajectory exhibits a stable trend over a longer period of time, that is, whether the process state continues to evolve along the same direction of change.

[0163] When the state evolution trajectory maintains a consistent direction of change in multiple consecutive time intervals, the state evolution trajectory within that time interval is judged to have a consistent overall trend of change.

[0164] Through the above methods, trend consistency analysis results are generated to characterize the degree of consistency in the direction of change of state evolution trajectories over the overall time scale, so as to reflect the overall change law of state evolution trajectories.

[0165] In a preferred embodiment of the present invention, based on the continuity analysis results and the trend consistency analysis results, stability features for characterizing the stability of the state evolution trajectory and consistency features for characterizing the consistency of the state evolution trajectory are generated respectively. The stability features and consistency features are then combined to form a set of state evolution features, including:

[0166] Based on the results of continuity analysis, the cases in which the state evolution trajectory remains continuously changing within adjacent time intervals are summarized to form stability characteristics for describing the smoothness of changes in the state evolution trajectory at local time scales.

[0167] Based on the trend consistency analysis results, the cases in which the state evolution trajectory maintains a consistent direction of change in multiple consecutive time intervals are summarized to form consistency characteristics for describing the degree of consistency of the direction of change of the state evolution trajectory on the overall time scale.

[0168] By combining stability features and consistency features as two dimensions to describe the characteristics of state evolution trajectory, a set of state evolution features that can comprehensively reflect the characteristics of state evolution trajectory changes is formed.

[0169] The above methods enable a unified representation of the local and overall change characteristics of the state evolution trajectory, providing a basis for distinguishing and determining subsequent separation behavior evolution patterns.

[0170] In a preferred embodiment of the present invention, based on a set of state evolution features, state evolution trajectories within different time intervals are distinguished, and state evolution trajectories with the same state evolution features are grouped into the same separation behavior evolution pattern, forming a separation behavior evolution pattern identification result used to characterize the evolution law of the separation process's operating state, including:

[0171] Based on the set of state evolution features corresponding to different time intervals, the stability and consistency features of each state evolution trajectory are aligned to form a feature description result that characterizes the features of each state evolution trajectory.

[0172] Based on the feature description results, the stability features and consistency features between different state evolution trajectories are compared one by one to determine the consistency relationship between the feature composition of each state evolution trajectory and generate the feature consistency determination result between the state evolution trajectories.

[0173] Based on the characteristic consistency determination results, state evolution trajectories with the same characteristics are grouped into the same trajectory set, and each trajectory set corresponds to a state evolution law.

[0174] Each trajectory set is identified as a separation behavior evolution pattern, and a correspondence is established between the separation behavior evolution pattern and the corresponding state evolution feature set to form the separation behavior evolution pattern recognition result.

[0175] In this embodiment of the invention, by aligning, comparing, and merging state evolution trajectories within different time intervals based on a set of state evolution features, the state evolution patterns exhibited by the separation process in different time intervals can be systematically distinguished. By merging state evolution trajectories with the same state evolution features into the same separation behavior evolution pattern, the formation process of the separation behavior evolution pattern is directly based on the consistency of state evolution features, thereby avoiding the need to artificially segment the separation process by relying on fixed time divisions or operating condition markers. This method enables the identification results of the separation behavior evolution pattern to truly reflect the evolutionary patterns of the separation process's operating state.

[0176] In a preferred embodiment of the present invention, based on the feature description results, the stability features and consistency features between different state evolution trajectories are compared one by one to determine the consistency relationship in feature composition between each state evolution trajectory, and a feature consistency determination result between state evolution trajectories is generated, including:

[0177] After obtaining the feature description results corresponding to different state evolution trajectories, any two state evolution trajectories are selected as comparison objects;

[0178] For the two state evolution trajectories, their corresponding stability features and consistency features are extracted respectively, and the corresponding features are compared item by item.

[0179] Determine whether two state evolution trajectories exhibit the same degree of stability in terms of stability characteristics, and whether they exhibit the same overall direction of change in terms of consistency characteristics;

[0180] When two state evolution trajectories are consistent in both stability and consistency characteristics, they are determined to have a consistent relationship in terms of feature composition.

[0181] By sequentially performing the above comparison process on the evolution trajectories of different states, a feature consistency determination result reflecting the feature consistency relationship between the evolution trajectories of each state is formed, which is used for subsequent trajectory merging processing.

[0182] In a preferred embodiment of the present invention, based on an identified separation behavior evolution pattern, the participation of each process monitoring parameter in the state evolution trajectory corresponding to the separation behavior evolution pattern on the state changes of the separation process is analyzed. Process monitoring parameters matching the separation behavior evolution pattern are selected, and the selected process monitoring parameters are mapped to corresponding process state variables to form a set of process state variables, including:

[0183] Based on the state evolution trajectory corresponding to the separation behavior evolution mode, extract the parameter change trajectory of each process monitoring parameter during the formation of the state evolution trajectory, and form the parameter change description result corresponding to each process monitoring parameter;

[0184] Based on the parameter change description results, the correspondence between each process monitoring parameter and the state change of the separation process during the formation of the state evolution trajectory is analyzed, and participation analysis results are generated to characterize the degree of participation of each process monitoring parameter in the state change of the separation process.

[0185] Based on the participation analysis results, process monitoring parameters that play a major role in characterizing the state changes of the separation process under the separation behavior evolution model were selected to form a candidate process monitoring parameter set;

[0186] Each process monitoring parameter in the candidate process monitoring parameter set is mapped to a corresponding process state variable based on its physical meaning and functional role in the separation process, forming a set of process state variables corresponding to the evolution mode of the separation behavior.

[0187] In this embodiment of the invention, for a specific separation behavior evolution mode, the change trajectories of each process monitoring parameter in its corresponding state evolution trajectory are analyzed. Based on the correspondence between each process monitoring parameter and the state changes of the separation process, process monitoring parameters are selected, thus establishing a correlation between the determination process of process state variables and the specific separation behavior evolution mode. By mapping the selected process monitoring parameters to process state variables, the set of process state variables used under different separation behavior evolution modes becomes targeted, thereby providing a foundation for constructing a digital twin simulation sub-model that matches the separation behavior evolution mode, ensuring that the simulation model maintains consistency with the corresponding separation process operating characteristics at the state description level.

[0188] In a preferred embodiment of the present invention, based on the state evolution trajectory corresponding to the separation behavior evolution mode, the parameter change trajectory of each process monitoring parameter during the formation of the state evolution trajectory is extracted to form a parameter change description result corresponding to each process monitoring parameter, including:

[0189] After determining a certain separation behavior evolution pattern, obtain the state evolution trajectory corresponding to that separation behavior evolution pattern;

[0190] For the time interval covered by the state evolution trajectory, the monitoring data of each process monitoring parameter at each time point are extracted in chronological order.

[0191] By connecting the monitoring data of the same process monitoring parameter that changes over time within a time interval, a parameter change trajectory reflecting the change process of the monitoring parameter is formed;

[0192] The above processing was performed on different process monitoring parameters to generate parameter change descriptions for each process monitoring parameter, which were then used for subsequent participation analysis.

[0193] In a preferred embodiment of the present invention, based on the parameter change description results, the correspondence between each process monitoring parameter and the state change of the separation process during the formation of the state evolution trajectory is analyzed, and participation analysis results are generated to characterize the degree to which each process monitoring parameter participates in the state change of the separation process, including:

[0194] After obtaining the parameter change description results corresponding to each process monitoring parameter, the parameter change trajectory is time-aligned with the state evolution trajectory corresponding to the separation process;

[0195] Analyze whether the monitoring parameters of each process change synchronously within the time interval of the change in the state evolution trajectory, and whether the changing trends remain consistent;

[0196] When the change trajectory of a process monitoring parameter maintains a synchronous relationship with the state evolution trajectory within multiple state change intervals, it is determined that the process monitoring parameter has a high degree of participation in the state changes of the separation process.

[0197] When there is a lack of synchronization between the change trajectory of a process monitoring parameter and the state evolution trajectory, it is determined that the process monitoring parameter has a low degree of participation in the state change of the separation process.

[0198] By performing the above analysis on each process monitoring parameter in sequence, a participation analysis result is formed to characterize the degree to which each process monitoring parameter participates in the state change of the separation process, providing a basis for subsequent parameter selection.

[0199] In a preferred embodiment of the present invention, based on the participation analysis results, process monitoring parameters that play a major characterizing role in the state changes of the separation process under the separation behavior evolution mode are screened to form a candidate process monitoring parameter set, including:

[0200] After obtaining the participation analysis results for each process monitoring parameter, all process monitoring parameters are organized according to their degree of participation in the state changes of the separation process.

[0201] Based on the participation analysis results, process monitoring parameters that can continuously reflect the state change characteristics of the separation process during the state evolution trajectory are identified;

[0202] Process monitoring parameters that can reflect the characteristics of changes in the state of the separation process are identified as candidate process monitoring parameters, and they are distinguished from process monitoring parameters with low participation and weak correlation with changes in the state of the separation process.

[0203] Through the above screening process, a set of candidate process monitoring parameters is formed to characterize the state change features of the separation process under the current separation behavior evolution mode, providing a basis for the determination of subsequent process state variables.

[0204] In a preferred embodiment of the present invention, each process monitoring parameter in the candidate process monitoring parameter set is mapped to a corresponding process state variable according to its physical meaning and functional role in the separation process, forming a set of process state variables corresponding to the evolution mode of the separation behavior, including:

[0205] After obtaining the set of candidate process monitoring parameters, analyze the physical meaning reflected by each candidate process monitoring parameter and its functional role in the separation process, and clarify the meaning of the separation process state corresponding to each process monitoring parameter.

[0206] Based on their physical meaning and functional role, each candidate process monitoring parameter is mapped to a specific state description element in the separation process, so that the process monitoring parameters can be used to characterize the changes in the state of the separation process.

[0207] The process monitoring parameters that have completed the establishment of the correspondence are transformed into process state variables and uniformly incorporated into the separate process state description system;

[0208] In this way, a set of process state variables that matches the current separation behavior evolution mode is formed, so that the digital twin simulation sub-model constructed subsequently can reflect the operation characteristics of the separation process under this separation behavior evolution mode at the state description level.

[0209] Embodiments of the present invention also provide an online At-211 separation process simulation platform based on digital twins, the platform comprising:

[0210] The separation behavior evolution pattern recognition module is used to collect process monitoring data that characterizes the material distribution state and separation result changes during the online separation process of At-211, and analyze the state evolution characteristics of the separation process based on the correlation of changes in process monitoring data within a continuous time interval, and identify the separation behavior evolution pattern that characterizes the evolution law of the separation process operation state.

[0211] The segment division module is used to divide the At-211 online separation process into multiple separation process segments according to different separation behavior evolution modes. Each separation process segment corresponds to a separation behavior evolution mode that remains relatively stable within a predetermined time range.

[0212] The digital twin simulation sub-model construction module is used to construct a digital twin simulation sub-model to describe the state change relationship of the At-211 separation process under each separation behavior evolution mode. Different digital twin simulation sub-models differ from each other in the composition of process state variables and / or state change constraints.

[0213] The separation behavior evolution mode determination module is used to continuously determine whether the separation behavior evolution mode corresponding to the current separation process corresponds to the running digital twin simulation sub-model based on real-time collected process monitoring data during the online simulation of the digital twin.

[0214] The simulation control module is used to maintain the continuous simulation of the separation process state by the digital twin simulation sub-model when the two are consistent; when the two are inconsistent, it terminates the continuation of the existing process state in the current digital twin simulation sub-model, switches to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode, and re-deduces the process state of the subsequent separation process based on the process state variables and their state change constraints in the digital twin simulation sub-model.

[0215] It should be noted that this system is a system corresponding to the above method. All implementation methods in the above method embodiments are applicable to this embodiment and can achieve the same technical effect.

[0216] Embodiments of the present invention also provide a computing device, including: a processor and a memory storing a computer program, wherein the computer program, when executed by the processor, performs the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0217] Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method described above. All implementations in the above method embodiments are applicable to this embodiment and can achieve the same technical effects.

[0218] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A simulation method for the online separation process of At-211 based on digital twins, characterized in that, The method includes: Process monitoring data was collected during the online separation process of At-211 to characterize the material distribution state and the changes in separation results. Based on the correlation of changes in process monitoring data over a continuous time interval, the state evolution characteristics of the separation process were analyzed, and the separation behavior evolution pattern that characterizes the evolution law of the separation process operation state was identified. Based on different separation behavior evolution modes, the At-211 online separation process is divided into multiple separation process operation segments, and each separation process operation segment corresponds to a separation behavior evolution mode that remains relatively stable within a predetermined time range; Based on each separation behavior evolution mode, a digital twin simulation sub-model is constructed to describe the state change relationship of the At-211 separation process under that separation behavior evolution mode. The different digital twin simulation sub-models differ from each other in the composition of process state variables and / or state change constraints. During the online simulation of the digital twin, based on the real-time collected process monitoring data, it is continuously determined whether the evolution mode of the separation behavior corresponding to the current separation process corresponds to the running digital twin simulation sub-model. When the two are consistent, the continuous simulation of the separation process state by the digital twin simulation sub-model is maintained; when the two are inconsistent, the continuation of the existing process state in the current digital twin simulation sub-model is terminated, and the model is switched to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode. Based on the process state variables and their state change constraints in the digital twin simulation sub-model, the process state of the subsequent separation process is re-deduced. Based on the correlation of changes in process monitoring data over a continuous time interval, the state evolution characteristics of the separation process are analyzed, and the separation behavior evolution patterns that characterize the evolution of the separation process's operating state are identified, including: Multiple process monitoring parameters used to characterize the material distribution and changes in separation results are collected during the separation process according to a preset sampling period. The process monitoring parameters are sorted according to the collection time to form multi-parameter time series data reflecting the continuous operation status of the separation process. Based on multi-parameter time series data, the direction, magnitude, and synchronicity of change of each process monitoring parameter are processed interval by interval within a continuous time interval to obtain joint change characteristics that characterize the coordinated change relationship of each process monitoring parameter within the corresponding time interval. Based on the joint change characteristics, the state evolution trajectory formed by the change of process state over time during the separation process is analyzed, and a set of state evolution characteristics is generated to characterize the stability and consistency of the state evolution trajectory. Based on the set of state evolution features, the state evolution trajectories in different time intervals are distinguished, and the state evolution trajectories with the same state evolution features are grouped into the same separation behavior evolution pattern, forming a separation behavior evolution pattern identification result used to characterize the evolution law of the separation process operation state; For each separation behavior evolution mode, a digital twin simulation sub-model is constructed to describe the state change relationship of the At-211 separation process under that separation behavior evolution mode. The different digital twin simulation sub-models differ from each other in the composition of process state variables and / or the constraints on state changes, including: Based on the identified separation behavior evolution pattern, analyze the participation of each process monitoring parameter in the state evolution trajectory corresponding to the separation behavior evolution pattern on the state change of the separation process, select process monitoring parameters that match the separation behavior evolution pattern, and map the selected process monitoring parameters to the corresponding process state variables to form a set of process state variables; Based on the set of process state variables, the process state variables are organized according to their roles and mutual influences in the separation process, and a state variable organization structure is constructed to describe the state transfer relationship between process state variables. Based on the organizational structure of state variables, determine the state change constraints, including the order of changes, boundaries, and mutual relationships of the state variables in the limiting process, and form state change rules applicable to the evolutionary pattern of this separation behavior. A digital twin simulation sub-model corresponding to the evolution mode of the separated behavior is constructed by the set of process state variables, the organizational structure of state variables, and the constraints of state changes.

2. The simulation method for the At-211 online separation process based on digital twins according to claim 1, characterized in that, When the two are inconsistent, the continuation of the existing process state in the current digital twin simulation sub-model is terminated, and the system switches to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode. Based on the process state variables and their state change constraints in the digital twin simulation sub-model, the process state of the subsequent separation process is re-deduced, including: After determining that the evolution mode of the separation behavior corresponding to the current separation process is inconsistent with the running digital twin simulation sub-model, stop the continued deduction of the separation process state of subsequent time steps based on the digital twin simulation sub-model, and block the transmission of the process state already formed in the digital twin simulation sub-model to the subsequent simulation process. Based on the newly determined separation behavior evolution mode, a digital twin simulation sub-model corresponding to the separation behavior evolution mode is selected as the digital twin simulation sub-model used for subsequent online simulation of the separation process. Based on the real-time collected process monitoring data, the process state variables in the selected digital twin simulation sub-model are initialized, and the process state of the subsequent separation process is gradually deduced under the constraints of the corresponding state change conditions.

3. The simulation method for the At-211 online separation process based on digital twins according to claim 1, characterized in that, Based on the joint change characteristics, the state evolution trajectory formed by the change of process state over time during the separation process is analyzed, generating a set of state evolution features to characterize the stability and consistency of the state evolution trajectory, including: Based on the joint change characteristics, the process states in the continuous time interval are connected in chronological order to form a state evolution trajectory consisting of multiple continuous process state points, where each process state point corresponds to the value result of the joint change characteristics in the corresponding time interval. Based on the state evolution trajectory, the relationship between adjacent process state points is analyzed segment by segment to determine the degree of continuity of process state changes within adjacent time intervals, and to generate continuity analysis results to characterize the local continuity of the state evolution trajectory. Based on the continuity analysis results, the overall direction and consistency of change of the state evolution trajectory in multiple continuous time intervals are analyzed, and trend consistency analysis results are generated to characterize the overall change law of the state evolution trajectory. Based on the results of continuity analysis and trend consistency analysis, stability features for characterizing the stability of state evolution trajectories and consistency features for characterizing the consistency of state evolution trajectories are generated respectively. The stability features and consistency features are then combined to form a set of state evolution features.

4. The simulation method for the online separation process of At-211 based on digital twins according to claim 1, characterized in that, Based on the set of state evolution features, the state evolution trajectories within different time intervals are distinguished. State evolution trajectories with the same state evolution features are grouped into the same separation behavior evolution pattern, forming a separation behavior evolution pattern identification result used to characterize the evolution law of the separation process's operating state, including: Based on the set of state evolution features corresponding to different time intervals, the stability and consistency features of each state evolution trajectory are aligned to form a feature description result that characterizes the features of each state evolution trajectory. Based on the feature description results, the stability features and consistency features between different state evolution trajectories are compared one by one to determine the consistency relationship between the feature composition of each state evolution trajectory and generate the feature consistency determination result between the state evolution trajectories. Based on the characteristic consistency determination results, state evolution trajectories with the same characteristics are grouped into the same trajectory set, and each trajectory set corresponds to a state evolution law. Each trajectory set is identified as a separation behavior evolution pattern, and a correspondence is established between the separation behavior evolution pattern and the corresponding state evolution feature set to form the separation behavior evolution pattern recognition result.

5. The simulation method for the At-211 online separation process based on digital twins according to claim 1, characterized in that, Based on the identified separation behavior evolution pattern, the participation of each process monitoring parameter in the state evolution trajectory corresponding to the separation behavior evolution pattern on the state changes of the separation process is analyzed. Process monitoring parameters that match the separation behavior evolution pattern are selected, and the selected process monitoring parameters are mapped to corresponding process state variables to form a set of process state variables, including: Based on the state evolution trajectory corresponding to the separation behavior evolution mode, extract the parameter change trajectory of each process monitoring parameter during the formation of the state evolution trajectory, and form the parameter change description result corresponding to each process monitoring parameter; Based on the parameter change description results, the correspondence between each process monitoring parameter and the state change of the separation process during the formation of the state evolution trajectory is analyzed, and participation analysis results are generated to characterize the degree of participation of each process monitoring parameter in the state change of the separation process. Based on the participation analysis results, process monitoring parameters that play a major role in characterizing the state changes of the separation process under the separation behavior evolution model were selected to form a candidate process monitoring parameter set; Each process monitoring parameter in the candidate process monitoring parameter set is mapped to a corresponding process state variable based on its physical meaning and functional role in the separation process, forming a set of process state variables corresponding to the evolution mode of the separation behavior.

6. A digital twin-based online separation process simulation platform for At-211, characterized in that, The platform, used in the method as described in any one of claims 1 to 5, comprises: The separation behavior evolution pattern recognition module is used to collect process monitoring data that characterizes the material distribution state and separation result changes during the online separation process of At-211, and analyze the state evolution characteristics of the separation process based on the correlation of changes in process monitoring data within a continuous time interval, and identify the separation behavior evolution pattern that characterizes the evolution law of the separation process operation state. The segment division module is used to divide the At-211 online separation process into multiple separation process segments according to different separation behavior evolution modes. Each separation process segment corresponds to a separation behavior evolution mode that remains relatively stable within a predetermined time range. The digital twin simulation sub-model construction module is used to construct a digital twin simulation sub-model to describe the state change relationship of the At-211 separation process under each separation behavior evolution mode. Different digital twin simulation sub-models differ from each other in the composition of process state variables and / or state change constraints. The separation behavior evolution mode determination module is used to continuously determine whether the separation behavior evolution mode corresponding to the current separation process corresponds to the running digital twin simulation sub-model based on real-time collected process monitoring data during the online simulation of the digital twin. The simulation control module is used to maintain the continuous simulation of the separation process state by the digital twin simulation sub-model when the two are consistent; when the two are inconsistent, it terminates the continuation of the existing process state in the current digital twin simulation sub-model, switches to the digital twin simulation sub-model corresponding to the newly determined separation behavior evolution mode, and re-deduces the process state of the subsequent separation process based on the process state variables and their state change constraints in the digital twin simulation sub-model.

7. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 5.

Images