CN114428606A

CN114428606A - Unit model analytic system based on EO variable

Info

Publication number: CN114428606A
Application number: CN202111574845.8A
Authority: CN
Inventors: 章佳明; 王吉; 苏井井; 陈佳伊; 陈卓
Original assignee: Zhejiang Supcon Technology Co Ltd
Current assignee: Zhejiang Supcon Technology Co Ltd
Priority date: 2021-12-21
Filing date: 2021-12-21
Publication date: 2022-05-03

Abstract

The invention relates to a unit model analysis system based on EO variables, which is a distributed architecture of a client and a server, wherein the server comprises: the intermediate data processing layer is used for converting the flow chart information of the client into structured data, selecting a corresponding processing strategy according to a processing instruction of the client to interact with the computing engine, acquiring a processing result and converting the processing result into information which can be identified by the client so as to display the client; the calculation engine is used for creating an information structure table according to the structural data converted by the intermediate data layer, and interactively calling the information structure table, the intermediate data layer and each dynamic link library so as to perform distributed iteration and solve EO variables or results in the information structure table, residual errors of equations and Jacobian matrixes, and the dynamic link libraries are connected with the calculation engine through respective abstract interfaces to realize distributed iteration. The analysis system simplifies the design and coding of the existing unit model, reduces the coupling with the unit model code and improves the calculation speed.

Description

Unit model analytic system based on EO variable

Technical Field

The invention relates to the technical field of data processing, in particular to a unit model analysis system based on EO variables.

Background

In chemical engineering process design, engineers usually use a tool (process simulation software such as Aspen) similar to a flow chart to abstract a production process in a plant, and abstract the process flow into a complex flow chart, and constituent elements in the flow chart usually include equipment, measuring instruments, pipelines and the like in the plant, and these elements are called unit models by those skilled in the art. When the parameters of the elements in the flow chart are configured correctly, the calculation of the flow chart can be performed to simulate the production process, and there are two general calculation processes: sequential and EO; the sequence is calculated according to the arrow sequence (topology) of the flow chart, and the former model outputs the calculation result to the next connection model; EO describes the calculated data in the flow model (i.e., the flow chart) as variables and the calculations in the model as equations describing the correlations between the variables, and the input-output relationships between the models are given by the connection equations between the variables.

Most of domestic process simulation software is based on sequential calculation, but the sequential calculation has many defects, such as the computational inflexibility, difficult convergence when many loops exist in the process, unsuitability for optimization control on operation variables, and the like.

The traditional EO-oriented unit model is constructed by using MATLAB coding to realize simple flow calculation and optimization due to the lack of development framework, and the main defects are as follows: 1) the unit model development expansibility is poor; 2) code reusability is poor; 3) the coding repetition workload is large; 4) generally, the method can only be used for the test calculation of small-scale flow; 5) the lack of organization for EO variables and equations makes localization and management difficult.

Disclosure of Invention

Technical problem to be solved

In view of the above-mentioned shortcomings and drawbacks of the prior art, the present invention provides a cell model analysis system based on EO variables.

(II) technical scheme

In order to achieve the purpose, the invention adopts the main technical scheme that:

in a first aspect, embodiments of the present invention provide a unit model parsing system based on EO variables,

the unit model analysis system is a distributed architecture of a client and a server, and the server comprises:

the intermediate data processing layer is used for receiving the flow chart information of the client and the processing instruction corresponding to the flow chart, which are transmitted by the server, converting the flow chart information into structured data used for computing service, selecting a processing strategy corresponding to the processing instruction according to the processing instruction to interact with the computing engine, and acquiring a processing result corresponding to the processing instruction;

converting the processing result into information which can be identified by the client, and transmitting the client display through the server;

the computing engine is used for creating an information structure table corresponding to the structured data according to the structured data converted by the intermediate data layer; and/or based on the processing instruction, the structured data and the information structure table, the intermediate data layer and each dynamic link library, interactively calling to perform distributed iteration and solve EO variables or results in the information structure table, and residual errors or Jacobian matrixes of equations;

a dynamically linked library, the dynamically linked library comprising at least: the dynamic link library is connected with the calculation engine through respective abstract interfaces, so that the calculation engine calls the dynamic link library to realize distributed iteration according to the requirements of processing instructions and structured data, and the result of transmitting the calculation engine is obtained.

Optionally, the processing instruction is an instruction to be executed by the server, triggered by the user at the client based on the flow chart information selected by the display interface, and the processing instruction includes: compiling, initial value estimation, initial value import or operation;

the information structure table created by the calculation engine comprises one or more of the following: the method comprises the following steps of a unit model object list, a connection information table, unit model library information, solver library information, an EO variable and equation general table, global configuration information, component thermodynamic information, calculation process information, EO variable and equation positioning information.

Optionally, when the processing instruction is a compiling instruction, the intermediate data processing layer is specifically configured to:

receiving the complete information of the flow chart transmitted by the server, and converting the received complete information of the flow chart into structured data; calling an initialization process of the calculation engine, creating a unit model object corresponding to each node in the flow chart according to the structured data and the compiling instruction, generating EO variables and equations corresponding to the unit model objects based on the created unit model objects, and enabling the calculation engine to store the EO variables and equations to which the unit model objects belong in an information structure table; the calculation engine updates/creates the positioning information of EO variables and equations in the information structure table for subsequent calling or accessing;

the intermediate data layer converts the information in the information structure table into data which can be identified and viewed by a client and sends the data to the client for display through a server;

and/or, the flowchart integrity information includes one or more of the following: configuration information, group grouping information, total group grouping information, connection information, group configuration information, group grouping configuration information and global configuration information corresponding to each node (namely a unit model) in the flow chart.

Optionally, when the processing instruction is an initial value estimation instruction, the intermediate data processing layer is specifically configured to:

converting the first associated information of the flow chart transmitted by the server into structured data, performing sequential operation in the intermediate data processing layer, calling EO variables and equations from an information structure table of a calculation engine based on the identification of all unit model objects of the flow chart, performing initial processing, determining the connection of the EO variables and input/output stream information according to a predefined input/output rule, realizing the transmission of the EO variables of all the unit model objects and obtaining the estimated initial value of each EO variable;

converting the transmission information of the estimated initial value and the EO variable into data which can be identified and checked by a client, and sending the data to the client for display through a server;

the first associated information of the flow chart comprises: flow chart complete information, EO variables and equations.

Optionally, the input and output rules of the intermediate data processing layer include:

all EO variables belonging to a given stream of unit model objects need to be continuous,

the arrangement among EO variable groups needs to be specified when a single unit model object is created;

the connected set of EO variables specific to the unit model object needs to be continuous.

Optionally, when the processing instruction is an initial value import instruction, the intermediate data processing layer is specifically configured to: and receiving initial value information of the EO variables transmitted by the server, configuring the EO variables of each unit model object to which the flow chart belongs, converting the configured information into data which can be identified and checked by the client, and sending the data to the client for display through the server.

Optionally, when the processing instruction is an operation instruction, the intermediate data processing layer is specifically configured to:

converting the second associated information of the flow chart transmitted by the server into structured data, and transmitting the converted structured data and the operation instruction to a computing engine;

the method comprises the steps that a calculation engine constructs a mathematical model corresponding to structured data and capable of being identified and processed by a solver according to the solver and solver parameter information in the structured data, the mathematical model is distributed to the corresponding solver by means of an abstract interface connected with each solver, asynchronous solution of cooperation of each solver and a dynamic link library is achieved, and the solver calls information of an information structure table of the calculation engine through the corresponding abstract interface in distributed solution to process;

the calculation engine receives process information and result information of each solver in each processing process, stores and combines the process information and the result information, and acquires final result information through set iteration times;

the intermediate data layer converts the stored process information, the result information of each iteration and the final result information into data which can be identified and checked by a client and sends the data to the client for display through a server;

the second associated information of the flowchart includes: complete information of the flow chart, EO variables and equations, solver and parameter information of the solver.

Optionally, the unit model library stores the positioning information/location information of the EO variables to which the unit model objects belong in the information structure table of the computation engine;

the processing functions of the compute engine include: the method comprises the steps of generating an EO variable and an equation, estimating an initial value of the EO variable, calculating equation residual and calculating a Jacobian matrix.

Optionally, the intermediate data layer converts the received flow chart complete information into structured data; calling an initialization process of the calculation engine to generate EO variables and equations, and processing the EO variables and the equations according to the generation rules of the EO variables and the equations;

the rule of generating EO variables and equations includes:

all EO variables and equations need to be stored continuously;

EO variables and equations need to be stored in groups, and the variables and equations in the groups occupy continuous storage space;

for the same configuration information, generating an object/routine requires generating the same EO variables and equation information.

(III) advantageous effects

The method of the invention can ensure that the full-flow simulation software in the prior art has better expansibility and practicability, overcomes the defect that the traditional flow chart directly carries out numerical calculation based on MATLAB, extracts the same characteristics (such as thermodynamic configuration and EO variable generation) in various unit models and puts the same characteristics into a common base class, and uses a connecting flow polymerization class to carry out grouping management on EO variables, thereby greatly reducing the difficulty and the code quantity of unit model development and improving the code reuse rate.

The system of the invention utilizes the construction scheme to realize a flow chart calculation engine in a short time, and the unit model construction only needs to pay attention to the flow type, the self variable and the equation calculation, thereby greatly shortening the development period.

Drawings

FIGS. 1A and 1B are partial schematic diagrams of a server architecture in an EO variable-based unit model parsing system according to an embodiment of the invention;

FIG. 2 is a diagram illustrating an architecture of an information structure table of the computing engine of FIG. 1;

FIG. 3 is a schematic diagram of the interaction between the metamodel and the compute engine of FIG. 1;

FIG. 4 is a schematic diagram of a process for constructing flash tanks based on FIG. 1;

FIG. 5 is a schematic diagram of the functions of a unit model;

FIG. 6 is a partial schematic diagram of a server process flow in a unit model analytic system based on EO variables.

Detailed Description

For the purpose of better explaining the present invention and to facilitate understanding, the present invention will be described in detail by way of specific embodiments with reference to the accompanying drawings.

In order to overcome the sequential defect, embodiments of the present invention provide a full-flow simulation software, that is, an EO variable-based unit model analysis system, which uses an EO variable calculation strategy to perform group management on EO variables, improve code reuse rate, and implement equation-oriented distributed iteration.

Some words in the embodiments of the present invention are defined or explained for better understanding.

The components are as follows: the common chemical substances form the components in the calculation process, such as methane, oxygen, water, nitrogen, air, benzene and the like. The total components are as follows: the current flow calculates the set of all components used. Grouping the components: a non-empty subset of the total composition. Physical properties: a property parameter of the component, such as the average molecular weight. Unit model: various devices in an actual plant are abstracted from the flow diagrams. Extracting an image of equipment in a factory in a flow chart as model information, inputting a stream, outputting the stream and configuring information; the meta-model in the calculation will be further abstracted into EO variables and EO equations (describing the relationship between its own EO variables) according to the information in the flow chart. Unit model types are mainly active stream, flash tank, mixing, separator, heater, etc.

Model information: including information about the model type (source stream, flash tank, etc.), model ID, etc. of the unit. And (2) flow: the stream is generally like a pipe for transporting a mixture of a certain concentration, and the stream in the flow chart is generally required to store the temperature, pressure, flow rate, total enthalpy of the mixture, and molar concentration of each component in the mixture as its attributes. The stream data is typically embedded in the cell model input and described in the output.

Connecting the flow: the connecting flow can describe the connection relation between the two unit models, belongs to data of flow chart configuration, and can be described by < a source unit model, a target unit model, a connecting flow ID, a source unit model flow ID and a target unit model ID >, and the flow of the source unit model is connected to the appointed flow of the target unit model through the connecting flow. The connecting stream directly relates (by equation) the flow-related properties of the source cell model (such as temperature pressure, abstracted as EO variables) in the calculation to the corresponding properties in the target cell model stream. In addition, the connecting flow can also be information in the unit model, the connecting flow in the unit model is used for representing the corresponding relation between pins in the unit model and the outer flow, and the connecting flow belongs to the configuration of the unit model, namely the connecting flow in the unit model can be directly called as the flow in practical application by calculation of an object per se.

Model configuration information: the method comprises interface configuration information related to unit models, such as model names, component groups, thermodynamics, input flow quantity/ID, output flow quantity/ID and the like, and different unit model configuration information, such as reaction configuration of a reactor, characteristic curve configuration of a compressor and the like, wherein the configuration information determines the abstraction (EO variables and equations) of the models in a calculation layer.

Global configuration information: contains some configuration information that all unit models share, such as the current time, unit model calculation options, etc.

Solving configuration information: the method comprises some configuration information of the solver, such as the selection of the solver and the configuration information related to the solver.

Calculating a mode: the method mainly comprises simulation calculation and optimization calculation, wherein the simulation calculation mainly solves an equation with the degree of freedom of 0, and the optimization problem mainly solves the optimization problem with the degree of freedom of more than 0, and mainly minimizes cost or maximizes economic benefit.

EO variables: the bottom layer of the numerical attributes related to the flow and the unit model is abstract, such as flow of a flow mixture, molar concentrations of various components, gas phase fraction of a flash tank, separation coefficient of a separator, isentropic efficiency of a compressor, heat exchange coefficient of a heat exchanger and the like, and different EO variables can be abstracted by different unit models under different model configurations. EO variables in the flow chart are mainly divided into 2, fixed variables and calculated variables. The fixed variables are constants in the calculation process, and the calculated variables are usually mathematically solved by the fixed variables and the EO equations. The fixed variables in the optimized calculation mode can be selected as optimized variables.

EO equation: to describe the mathematical relationship between EO variables.

Univariate connection: for describing the connection relationship (typically an equality relationship) between different EO variables, one connection relationship will add one equation in the calculation. May be configured by the client or fixed by the system (generated by the system itself or the unit model requirements).

Variable connection group: in order to facilitate the introduction of variable connection management, the variable connection management system is a group formed by a plurality of single variable connections. May be configured by the client or fixed by the system. If the connected stream is actually a variable connected set (system fixed) at the bottom, the relationship (equality) of the specified stream EO variables (temperature, pressure, flow, mixture concentration) in the source/target cell model is established. Specific connection sets can be designated for connection at the time of designing each unit model according to actual requirements, such as the EO variable representing heat can be exposed to the upper layer for connection when designing the flash tank.

Degree of freedom: and the variable quantity-equation quantity is expressed by calculation variable quantity. And 0 represents that the equation number is equal to the variable number, the equation is a square equation at the moment, the equation can be directly solved by a numerical method at the moment, and if the degree of freedom is more than 0, an optimized variable exists, and the corresponding optimization problem is solved. The simulation mode generally requires a degree of freedom of 0, and the optimization mode requires a degree of freedom of 0 or more.

And (3) sub-process: a special type of unit model. The flow diagram itself can also be abstracted as a unit model, since it can be connected to the outside flow diagram unit model by means of connecting streams, also EO variables and equations.

Solver: and (3) solving a mathematical model, such as solving an equation F (x) 0 or an optimization problem min f (x) s.t.g (x) 0, l < (x) > u.

A unit model library: the library of the computation end is a set formed by dynamic link libraries for realizing the universal interfaces of the unit models.

Solver library: and (4) solving the set formed by the dynamic link library of the solver.

Thermodynamic library: various thermodynamic calculations dynamically link collections of libraries.

Model solution sequence: describing a model sequence participating in the flowchart solution, wherein all unit models in the flowchart do not necessarily participate in calculation when the solution is performed, and part of sub-processes may be selected to perform virtual connection solution through configuration variable connection groups in the middle, so that the solution sequence is required to describe the unit models participating in the solution.

And (3) calculation service: the method is mainly used for computing resource management and remote interaction of clients, such as distribution of computing tasks and parallel/distributed computing thread management.

An intermediate data layer: the method is mainly used for processing the data of the client into a high-efficiency structure required by calculation and organizing and processing result information returned by the process calculation engine into a structure required by the client

A flow calculation engine: and performing related modeling calculation according to the data provided by the intermediate data layer, and returning the calculation result to the intermediate data layer for processing.

Example one

As shown in fig. 1A and fig. 1B, the present embodiment provides an EO variable-based unit model analysis system, where the unit model analysis system is a distributed architecture of client-server, and a server of the distributed architecture includes: the system comprises an intermediate data processing layer, a calculation engine and a dynamic link library;

converting the processing result into information which can be identified by the client, and transmitting client display through the server;

the computing engine is used for creating an information structure table corresponding to the structured data according to the structured data converted by the intermediate data layer; and/or based on the processing instruction, the structured data and the information structure table, the intermediate data layer and each dynamic link library are called interactively to perform distributed iteration and solve the initial values or the results of the EO variables and the equations in the information structure table,

a dynamically linked library, the dynamically linked library comprising at least: the dynamic link library is connected with the computing engine through respective abstract interfaces, so that the computing engine calls the dynamic link library to realize distributed iteration according to the requirements of processing instructions and structured data, and the initial value or result is obtained, as shown in fig. 1B.

Understandably, the unit model library stores the positioning information/position information of the EO variable to which the unit model object belongs in an information structure table of a calculation engine;

the processing functions of the compute engine include: the method comprises the steps of generating EO variables and equations, estimating initial values of the EO variables, calculating residual equations and calculating Jacobian.

In practical application, a server can receive related data transmitted by a client through a receiving module, further create a thread through a computing service, and transmit the thread to an intermediate data processing layer through the computing service. Through the calling and the interaction, the distributed iterative computation is realized, the code reuse rate is improved, and the development period is greatly shortened.

In addition, in this embodiment, the processing instruction is an instruction to be executed by the server, triggered by the user at the client based on the flow chart information selected by the display interface, and the processing instruction includes: compiling, initial value estimation, initial value import or operation; as shown in fig. 6.

The information structure table created by the calculation engine comprises one or more of the following: a unit model object list, a connection information table, unit model library information, solver library information, EO variable and equation summary table, global configuration information, component thermodynamic information, calculation process information, EO variable and equation positioning information, as shown in fig. 2.

In a specific implementation process, when the processing instruction is a compiling instruction, the intermediate data processing layer is specifically configured to:

the complete information of the flow chart comprises one or more of the following: configuration information, group grouping information, total grouping information, connection information, group configuration information, group grouping configuration information and global configuration information corresponding to each node. The complete information of the flow diagram can be configured for the user at the client, corresponding to the configuration process in fig. 6, which can select the total components, set a plurality of groups, configure thermodynamics, group flow diagram (e.g. a source stream OS1, a flash tank B1, connected by a connecting stream S1), configure cell model objects (e.g. configure source stream temperature, pressure, composition parameters (mainly used with initial value estimation), configure flash tank effective phase, entrainment (affecting EO variables and equations), flash type, temperature, pressure, etc. (used for initial value estimation)), and generate group and thermodynamics data, cell model data and connecting stream data corresponding to the flow diagram data, etc. in the configuration process.

It is understood that operations that the compilation process may trigger include: detecting the configuration, if the configuration is correct, sending the configuration data to an intermediate data layer by the client through the computing service of the server, compiling the configuration data, analyzing the configuration data by the intermediate data layer, performing unit conversion on configuration parameters, finally forming a structured solving structure and a solving sequence, and transmitting the processed data to a computing engine for compiling; the calculation engine establishes a unit model and a connection object in the flow chart according to the data of the intermediate data layer, calls an EO variable and equation generating interface of each unit model according to the solving sequence to generate an EO variable and equation, and returns the EO variable/equation to the intermediate data layer; and the intermediate data layer processes the returned EO variables and equations and finally transmits the processed EO variables and equations to the client for display.

In addition, the intermediate data layer converts the received complete information of the flow chart into structured data; calling an initialization process of the calculation engine to generate EO variables and equations, and processing the EO variables and the equations according to the generation rules of the EO variables and the equations; i.e., the arrangement of EO variables is not arbitrary, the following rules are defined in this embodiment in order to improve the efficiency and scalability of the system.

The rule of generating EO variables and equations includes: rule 1): all EO variables and equations need to be stored continuously; since the variable in the stream is usually a variable connected set, the subsequent connection operation can be facilitated if it is continuous. Rule 2) EO variables and equations need to be stored in groups, and the variables and equations in the group occupy continuous storage space, so that subsequent connection operation as a whole is facilitated; rule 3) for the same configuration information, generating an object/routine requires generating the same EO variables and equation information.

The rules realize the arrangement among the specified variable groups in the single unit model, for example, the flash tank generally discharges the inlet flow, the self variable of the flash tank, the gas phase outlet and the liquid phase outlet in sequence, so that the sparsity of the equation Jacobian in the decomposition process is ensured, and the solving efficiency is improved.

In a second implementable manner, when the processing instruction is an initial value estimation instruction, the intermediate data processing layer is specifically configured to: converting the first associated information of the flow chart transmitted by the server into structured data, performing sequential operation in the intermediate data processing layer, calling EO variables and equations from an information structure table of a calculation engine based on the identification of all unit model objects of the flow chart, performing initial processing, determining the connection of the EO variables and input/output stream information according to a predefined input/output rule, realizing the transmission of the EO variables of all the unit model objects and obtaining the estimated initial value of each EO variable; converting the transmission information of the estimated initial value and the EO variable into data which can be identified and checked by a client, and sending the data to the client for display through a server;

it will be appreciated that after compiling to generate EO variables, the user can directly import previously calculated EO variable values from elsewhere, and if no values are available then an initial value estimate is needed which will trigger the following operations: and the intermediate data layer performs data conversion, calculates a topological sequence of the flow chart, calls a unit model initial value estimation interface of the calculation engine according to the topological sequence, after the current unit model is calculated, connects the flow related variable (variable connection group) to the flow variable appointed by the next unit model according to the connection flow information, continues to calculate until all the unit models needing to be estimated are calculated, and finally returns the value of the EO variable to the client for display.

The first related information of the flowchart includes: flow chart complete information, EO variables and equations. Specifically, the input/output rule of the intermediate data processing layer includes: a) all EO variables belonging to the stream specified by the unit model object need to be continuous, b) the arrangement in the middle of the EO variable group needs to be specified when a single unit model object is created; c) the connected set of EO variables specific to the unit model object needs to be continuous.

Of course, when the processing instruction is an initial value import instruction, the intermediate data processing layer is specifically configured to: and receiving initial value information of the EO variables transmitted by the server, configuring the EO variables of each unit model object to which the flow chart belongs, converting the configured information into data which can be identified and checked by the client, and sending the data to the client for display through the server.

In a third possible implementation manner, when the processing instruction is an execution instruction, the intermediate data processing layer is specifically configured to: converting the second associated information of the flow chart transmitted by the server into structured data, and transmitting the converted structured data and the operation instruction to a computing engine;

the calculation engine receives the process information and the result information of each solver in each processing process, stores and combines the process information and the result information, and acquires final result information through set iteration times;

After the EO variable initial value is obtained, EO-related configuration may be performed, for example, operations such as changing variable attributes (exchanging fixed variables and calculation variables), changing upper and lower limits of variables, changing values of fixed variables to perform special calculation, selecting a calculation mode, and configuring a solver.

The EO-related configuration is calculated and the process calculation (run button in video, for example in analog mode) will trigger the following operations:

the method comprises the steps that a server transmits data such as freedom degree configuration information, configuration information and global configuration information of a client to a middle data layer for simulation scene calculation, the middle data layer conducts structuralization processing on the configuration information data and the global configuration information of the client to generate a structure (unit model structure data (including ID, group/thermodynamic data, in-out stream description, configuration information …) needed by a solving engine, a solving sequence and variable connection group information are generated, the solving configuration information (solver ID, solver attribute) is solved, and the data are transmitted to the computing engine for simulation mode solving; the calculation engine constructs a mathematical model of the flow chart according to the data of the middle layer, and transmits the mathematical model to a bottom solver selected in the solving configuration for calculation;

and continuously updating the value of the current EO variable by the solver and calling the residual error of the unit model, obtaining the residual error value and the Jacobian matrix under the value of the current EO variable by the Jacobian calculation interface to carry out iterative calculation, asynchronously informing the intermediate data layer of the solving result of each round, and finally completing the solving of the mathematical model.

The calculation engine processes the result returned by the solver to obtain a solving result of the flow chart, and returns the result to the intermediate data layer; the intermediate data layer organizes and processes the result and finally returns the result to the client for display. After obtaining the final solution result, the user usually reconfigures or performs EO allocation according to the result, and then recalculates until obtaining a satisfactory result.

The client in the unit model parsing system of this embodiment may be any client for creating a process flow diagram, and the client is mainly used for displaying an operation result and receiving an operation instruction of a user. For example, the system may include: the display interface is used for displaying and receiving user trigger operation at a client, and a group area, a total group area, a unit model configuration area, a thermodynamic configuration area, an editing area and the like are displayed in the display interface;

the component group area mainly displays a plurality of chemical substances to form components in a calculation flow chart; the total group partition area mainly displays information of common components used in a plurality of calculation flow charts; the unit model configuration area mainly shows a plurality of EO variables and/or EO equations abstracted by the unit models in the flow chart; the thermodynamic configuration area is mainly used for configuring thermodynamic attribute information in a thermodynamic diagram; the editing area is mainly used for displaying elements in the group grouping area, the total group division area, the unit model configuration area and the thermodynamic configuration area triggered by the user, and combining and/or displaying a result fed back by the server according to an operation instruction selected by the user.

For example, the total component may be selected in the client, such as selection C₆H₆，C₇H₈，H₂O, etc., and also to configure the thermodynamics, calculation subroutines specifying the properties of each thermodynamics, group flow diagrams, such as source stream OS1, a flash drum B1, connected by connecting stream S1; unit models can also be configured, such as configuration of source stream temperature, pressure, composition parameters (mainly used for initial value estimation), configuration of flash tank effective phase state, entrainment (influencing EO variables and equations), flash type, temperature, pressure and other attributes (used for initial value estimation) and the like.

The calculation engine in this embodiment records the positioning information of the EO variables of each unit model object in the summary table, and distributes the calculation to each unit model according to the positioning information, so that the calculation of each unit model only processes the relationship of the EO variables of the unit model, and the relationship of different unit models is realized by the association of connection equations.

In the prior art, when EO variables and equations are compiled, the EO variables and equations are designed to be stored in a unit model object memory, which causes the EO variable equations to be dispersed in the memory, and all EO variables can be obtained only by traversing the unit models in the process. In the embodiment of the invention, a general table is distributed by a calculation engine during compiling, each unit model object is filled with EO variables and equations, and the calculation engine records the positioning information of the EO variables of each unit model object in the general table.

The grouping management of the EO variables simplifies the work of the intermediate data layer, and the variables in the group can be directly connected without considering specific EO variables when the variable connection group is processed.

The unit model design can independently abstract the connecting streams into classes and aggregate the connecting streams into the unit model, and the unit model design has the following advantages: the coupling with the unit model code is reduced; a model stream library can be constructed as required, and the code reuse rate is increased; the design of the unit model is simplified; aggregation with the unit model can enable an equation calculation routine to rapidly access EO variables of the streams of the specified inlet and outlet, and pointer conversion can be directly carried out without table lookup; variable positioning information can be conveniently generated by a public base class, and only the stream EO variables are sequentially generated and the variable offset information is recorded; virtual grouping of cell model EO variables can be achieved using special join stream classes, so that complex functions in some processes can be simplified (e.g. join equations between variable sets).

Meanwhile, the common base class of the unit model realizes interaction between all model common functions (such as thermodynamic configuration and attribute configuration) and the engine, and model development only needs to pay attention to variables and equations, so that the design and coding of the unit model are greatly simplified, as shown in fig. 3.

In addition, the calculation engine disperses the calculation to each single unit model, and the calculation is further adapted through the organization of some information and the interface of a solver, so that the calculation process of the unit model only needs to pay attention to the relation between EO variables of the unit model, and the relation before different unit models is only described through a connection equation.

Example two

In order to better explain the contents of the above-mentioned first embodiment, the following describes the technical solution of the present application from another point of view.

Description of the Overall computing architecture

The server of the unit model analysis system based on the EO variables can comprise a calculation engine, a unit model library and a solver library; as shown in fig. 1A.

The calculation engine integrates various unit models (such as flash tanks, separators and the like) and solvers through abstract interfaces. Various unit models are connected and interacted with the engine through abstract interfaces; the solver library also interacts with the compute engine through a dedicated solution abstraction interface.

The specific structure of the calculation engine shown in fig. 2 is mainly responsible for management of a unit model library and a solver library, establishment of a flow chart, management of EO variables and equations, and calculation modes of various scenes.

The flow chart information is represented in a calculation engine as a unit model information list and a connection stream list, and the connection is mainly represented as a source unit model, a target unit model, a source connection stream and a tuple of a target connection stream.

In order to facilitate management of EO variables in the calculation engine, all EO variables of the unit model need to be continuously stored in the calculation engine, and the unit model can be quickly accessed by only storing the positions of the EO variables in the unit model in the whole summary table.

In the aspect of calculation, a great amount of iterative calculation is often needed at a solver end, each round of iterative calculation engine organizes and dispersedly updates current iterative information of the solver to each individual unit model for calculation, and after all unit models are calculated, calculation results are organized together and transmitted to the solver for continuous iteration; the framework can disperse the solving task into each single unit model, and each unit model only needs to provide a self-calculating method, so that the parallelism of the system can be improved, and the expansion of a subsequent analysis system is facilitated.

In this embodiment, from the division of the calculation function, the unit model may be mainly divided into the generation of EO variables/equations, the initial value estimation of EO variables, the calculation of equation residuals, and the calculation of jacobian matrices. The EO variables/equations of the cell model are generated according to the current configuration information of the cell model, and the generation of the EO variables/equations meets the following rules:

1) all EO variables/equations need to be stored continuously

2) EO variables/equations require storage in groups, with the variables/equations in the group occupying contiguous storage space

3) For the same configuration information, the generation routine needs to generate the same EO variable/equation information,

rule 1) is mainly used for access performance improvement (access through array indexes) and convenience for subsequent unified management, rule 2) is mainly used for configuration calculation function extension and convenience for variable group management (group variables/equations can be expressed by index intervals), for example, when correlation among some special EO variables/equations among unit models needs to be established during calculation, the special variables/equations can be placed in an independent group, the index intervals are directly returned to engine calculation, and rule 3) is a constraint condition for configuration calculation.

In the compiling stage, the EO variable/equation generation routine generates a variable/equation table according to the current configuration in the flow chart transmitted by the client. In the initial value estimation stage, the EO variable initial value estimation needs to calculate the value and the related attribute of the generated EO variable according to the current configuration in the flow chart transmitted by the client, in the operation stage, the equation residual error calculation needs to calculate the residual value of each equation according to the initial value of the EO variable of the current model and the current unit model configuration, and the jacobian calculation needs to calculate the first derivative of each equation to the EO variable according to the initial value of the EO variable of the current unit model and the current unit model configuration. The initial value estimation of EO variables, the equation residual calculation and the calculation of the Jacobian matrix need to access the configuration related information of the unit model, the EO variables of the unit model and the equation of the unit model.

The configuration information of the cell model mainly comprises connection flow information among the cell models, connection flow information inside the cell models, component information, thermodynamic information and configuration attributes related to the model, and the configuration information can determine EO variables/equations of the model. The information of the connection flow between the unit models mainly describes the information related to the connection of the unit models, and mainly comprises the pin number of the unit models, the connection number, the connection flow number, the flow component group and the thermodynamics. The connecting streams can be divided into outlet streams, inlet streams, outlet gas phase streams, outlet liquid phase streams and the like according to purposes, different streams contain different types of EO variables, and the EO variables belonging to the same stream need to form a group. The group related information is mainly used for EO variable/equation generation, and thermodynamics is mainly used for initial value estimation, residual error and Jacobian calculation. Configuration attributes related to the unit model mainly affect the EO variable/equation generation, initial value estimation, equation residual and jacobian matrix calculation, and the attributes are generally expressed by < key, value > pairs. That is, the connection flow information between the cell models is used for the generation of EO variable connection groups, or the overall calculation is implemented. The sequential operation of the middle data layer uses a connecting stream between the unit models, and the computing engine can add an EO variable connection group of the connecting stream between the unit model objects when the computing results of all the unit model objects are combined, namely the connecting stream between the unit models is mainly used for computing EO variables and equations, equation residual errors and Jacobian matrixes when the unit model objects are compiled to generate the EO variables and the equations;

in addition, the connection flow is information in the unit model, the connection flow in the unit model is used for representing the corresponding relation between pins in the unit model and the external flow, and the connection flow belongs to the configuration of the unit model, namely the connection flow in the unit model can be directly called as the flow in practical application through calculation of an object per se.

In order to meet the above requirements of the unit model, the server of the embodiment provides a reasonable construction scheme of the unit model. Because the only difference in function between the unit models is EO variable/equation generation, initial value estimation and equation residual error/Jacobian matrix calculation, all other functions (such as attribute acquisition, component acquisition and connected stream acquisition) can be put into the basic class, and the corresponding interface is led out to be accessed by the subsequent derived class. As shown in FIG. 3, all configuration related information such as group grouping, thermodynamics, model attributes, etc. can be placed in the base class (directly created by engine interaction), without the need for subsequent derivative class creation and modification, so that the derivative class only needs to be concerned with EO variables/equations and related computation routines. The unit model EO variables can be divided into a flow-related EO variable group, a model EO variable group and a special EO variable group, a special flow connection class can be defined, connection-related information structures are stored and used for generating the flow EO variable group and the special EO variable group (virtual flow), a generation rule of the group of EO variables is provided, and a model variable generation function is responsible for generating the EO variables of the model. Thus, only the unit model is required to derive and provide an EO variable generation rule table of each group, the basic class can generate an integral EO variable table according to the table in sequence according to the designated sequence, and meanwhile, the position information of the EO variables of the group can be maintained. In order to improve the access performance, the rule list can be integrated into the unit model base class in an aggregation mode, and the time loss caused by subsequent table lookup is avoided. In order to hide the interaction between the cell model and the engine, all the interactive interfaces can be completely implemented in the basic class, and the interactive interfaces can be implemented by polymorphic access to functional interfaces in the derived class (the use of CRTP in C + + can be considered). The basic class is also responsible for giving default realization of equation generation, initial value estimation and residual jacobian calculation, and mainly utilizes a polymorphic mode to call the realization of the derived class. The interface functionality provided by the base class is shown in fig. 5.

In view of the unit model shown in fig. 3, a simple flash tank model can be easily assembled with the help of the presentation interface of the client. The flash tank model mainly comprises an inlet (variable number, at least one), a gas phase outlet and a liquid phase outlet, 3 variable generation rules can be determined to generate an EO variable group according to the configuration information of the current stream class, the flash tank can be derived from the basic class and the 3 rules are transmitted to the basic class, and then only EO variable/equation generation, initial value estimation and residual error/Jacobian interface need to be realized, and each method can be easily realized by using the method provided by the basic class. The flash tank EO variables and equations constructed in the way not only meet the calculation related constraints, but also can be directly controlled and operated by an engine, but only the variable generation and equation calculation related implementation in the derived classes, and the corresponding construction relationship is shown in FIG. 4.

It should be noted that in the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the terms first, second, third and the like are for convenience only and do not denote any order. These words are to be understood as part of the name of the component.

Furthermore, it should be noted that in the description of the present specification, the description of the term "one embodiment", "some embodiments", "examples", "specific examples" or "some examples", etc., means that a specific feature, structure, material or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.

While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, the claims should be construed to include preferred embodiments and all changes and modifications that fall within the scope of the invention.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention should also include such modifications and variations.

Claims

1. An EO variable-based unit model analytic system, wherein the unit model analytic system is a client-server distributed architecture, and a server comprises:

2. The unit model parsing system of claim 1, wherein the processing instruction is an instruction to be executed by the server, triggered by the user at the client based on the flowchart information selected by the presentation interface, and the processing instruction comprises: compiling, initial value estimation, initial value import or operation;

the information structure table created by the calculation engine comprises one or more of the following: the method comprises the following steps of a unit model object list, a connection information table, unit model library information, solver library information, an EO variable and equation summary table, global configuration information, component thermodynamic information, calculation process information, EO variable and equation positioning information.

3. The unit model parsing system of claim 2, wherein when the processing instruction is a compiling instruction, the intermediate data processing layer is specifically configured to:

receiving the complete information of the flow chart transmitted by the server, and converting the received complete information of the flow chart into structured data; calling an initialization process of the calculation engine, creating a unit model object corresponding to each node in the flow chart according to the structured data and the compiling instruction, generating EO variables and equations corresponding to the unit model objects based on the created unit model objects, and enabling the calculation engine to store the EO variables and equations to which the unit model objects belong in an information structure table; the calculation engine updates/creates EO variables and equations in the information structure table for the positioning information of subsequent calling or accessing;

and/or, the flowchart integrity information includes one or more of the following: configuration information, group grouping information, total grouping information, connection information, group configuration information, group grouping configuration information and global configuration information corresponding to each node in the flow chart.

4. The unit model parsing system of claim 3, wherein, when the processing instruction is an initial value estimation instruction, the intermediate data processing layer is specifically configured to:

5. The unit model parsing system of claim 4, wherein the input and output rules of the intermediate data processing layer comprise:

when a single unit model object is created, the arrangement among EO variable groups needs to be specified;

6. The unit model parsing system of claim 4,

when the processing instruction is an initial value import instruction, the intermediate data processing layer is specifically configured to: and receiving initial value information of the EO variables transmitted by the server, configuring the EO variables of each unit model object to which the flow chart belongs, converting the configured information into data which can be identified and checked by the client, and sending the data to the client for display through the server.

7. The unit model parsing system of claim 4,

when the processing instruction is an operation instruction, the intermediate data processing layer is specifically configured to:

8. The unit model parsing system of claim 4,

the unit model library stores the positioning information/position information of the EO variable which the unit model object belongs to in an information structure table of a calculation engine;

9. The unit model parsing system of claim 3,

the intermediate data layer converts the received complete information of the flow chart into structured data; calling an initialization process of the calculation engine to generate EO variables and equations, and processing the EO variables and the equations according to the generation rules of the EO variables and the equations;

the rule of generating EO variables and equations includes:

all EO variables and equations need to be stored continuously;