Chemical industrial park distributed state estimation method and system considering multi-energy flow dynamics

Abstract

The present application belongs to the technical field of distributed state estimation, and provides a chemical industry park distributed state estimation method and system considering multi-energy substance flow dynamics, according to the time sequence coupling of electric and thermal heterogeneous energy flow and hydrogen / ammonia / alcohol multi-substance flow and the differentiated dynamic characteristics thereof, a state space model of a power distribution network, a heating network and a chemical material flow network and the like is constructed, and a hydrogen / ammonia / methanol chemical production equipment coupling element model is constructed, thereby constructing a green electricity chemical industry park real-time state estimation model considering substance-energy coupling and dynamic characteristics; taking a particle filtering algorithm as a framework, a distributed state estimation is constructed based on an asynchronous parallel architecture, thereby solving the data privacy problem between the power and chemical subsystems; a green electricity chemical industry park multi-time scale distributed state estimation architecture combined with a new information triggering mechanism is constructed. The present application solves the measurement uploading period difference problem and the calculation resource waste problem caused by the time constant difference between different networks.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of distributed state estimation, and particularly relates to a chemical industrial park distributed state estimation method and system considering multi-energy substance flow dynamics, in particular, a green electricity chemical industrial park distributed state estimation method and system considering multi-energy substance flow and heterogeneous energy flow dynamics. BACKGROUND

[0002] The statements in this section merely provide background information related to the present application and do not necessarily constitute prior art.

[0003] The green electricity chemical industrial park refers to a wind-solar-hydrogen-ammonia-alcohol integrated park that replaces fossil fuels with green electricity and green hydrogen to supply energy for heavy chemical systems. The park integrates controllable units such as distributed power generation, energy storage, and chemical production complete equipment, realizes load-following power supply for chemical production, and has important significance.

[0004] In the power and chemical subsystems of the green electricity chemical industrial park, there are spatio-temporal cross-coupling electric and thermal heterogeneous energy flows and hydrogen / ammonia / alcohol / ester multi-energy substance flows. Different types of energy flows and substance flows have great differences in dynamic characteristics and mathematical descriptions. In order to optimize the operating state of the green electricity chemical industrial park while ensuring the safety of chemical production, the real-time operating state of the power and chemical subsystems needs to be obtained first, which serves as the data basis for the optimization and scheduling of the green electricity chemical industrial park. Therefore, real-time state estimation is needed, and the current real-time state estimation method for the green electricity chemical industrial park has the following problems:

[0005] Electric and thermal heterogeneous energy flows and hydrogen / ammonia / alcohol / ester multi-energy substance flows have great differences in dynamic characteristics and mathematical descriptions. The propagation time constant of electric power flow is usually in the order of microseconds or milliseconds, while the propagation time constant of thermal and material flows is usually in the order of minutes or hours. If only the steady-state or quasi-steady-state time scale commonly used in power subsystem state estimation is used for state estimation of the chemical subsystem heating network and material network, the inherent dynamic characteristics of the system cannot be captured, resulting in poor state estimation accuracy, which affects the accuracy of subsequent scheduling and other advanced application decisions, and thus affects the safe operation of the system.

[0006] The power and chemical subsystems of a green electricity chemical industrial park have data privacy problems due to different management subjects. Due to different dynamic characteristics and estimation model properties of the power grid, heat supply network and material network, the state estimation execution cycle is different, and the estimation efficiency also has differences. In addition, the power and chemical subsystems have certain differences in sensitivity to changes in their coupled power. Therefore, it is difficult to ensure data privacy by using traditional centralized state estimation; using a single time scale synchronous parallel distributed state estimation strategy has problems such as difficulty in fully utilizing computing resources, wasting waiting time, and affecting estimation efficiency; and if state estimation is performed when the measurement value of the coupled power of any system changes, it may also lead to waste of computing resources in the case of small overall state changes. SUMMARY

[0007] In order to solve the above problems, the present application provides a chemical industrial park distributed state estimation method considering energy and mass flow dynamics and a system, which fully considers the time sequence coupling and differentiated dynamic characteristics of electric, thermal heterogeneous energy flow and hydrogen / ammonia / alcohol multi-substance flow, and realizes efficient and accurate state estimation of electric, thermal and material flow networks.

[0008] According to some embodiments, the present application adopts the following technical scheme:

[0009] A chemical industrial park distributed state estimation method considering multi-energy and mass flow dynamics, comprising the following steps:

[0010] The green electricity chemical industrial park is divided into an electric power subsystem and a chemical subsystem, wherein the electric power subsystem manages the operation of the power distribution network, and the chemical subsystem manages the operation of the heat supply network, the chemical material flow network and the chemical production equipment;

[0011] Network information and equipment information in each subsystem are obtained respectively, and the parameters of the coupling elements of each branch in the power distribution network, the heat supply network and the chemical material flow network and the hydrogen / ammonia / methanol chemical production equipment are normalized;

[0012] For the electric power subsystem, according to the power distribution network information, the state transition equation of the power distribution network under quasi-steady state is constructed by using the back-calculation method, and the power distribution network state space model is constructed in combination with the measurement model of the power distribution network;

[0013] For the chemical subsystem, according to the material flow network pipe pressure equation, material flow momentum equation and material flow pipe mass balance equation, the partial differential model of the material flow network is established and discretized, the state transition equation of the material flow network is constructed, and the material flow network state space model is constructed in combination with the measurement model of the material flow network; according to the dynamic thermodynamic model of the heat supply network considering pipe transmission loss, the partial differential model of the heat supply network is established and discretized, and the heat supply network state space model considering temperature transmission dynamic characteristics is constructed in combination with the measurement model of the hydraulic network and the heat network in the heat supply network;

[0014] For hydrogen / ammonia / methanol chemical production equipment and coupling elements, an equation constraint is constructed to describe the power consumption, raw material input, and chemical production heat release, and a coupling element state space model is constructed;

[0015] Based on the constructed power distribution network state space model, material flow network state space model, and heat supply network state space model, a real-time state estimation model of green chemical industrial park considering the dynamic characteristics of multi-element material flow and heterogeneous energy flow is constructed, and a distributed state estimation model of green chemical industrial park based on an asynchronous parallel computing architecture is constructed;

[0016] A new information triggering mechanism is introduced, and whether to perform real-time state estimation of electricity, heat, and material flow at the current time is determined based on the state estimation execution period of the electricity, heat, and material flow network and the new information triggering mechanism. If so, the distributed state estimation model of the green chemical industrial park is used to perform distributed state estimation of the corresponding network to obtain state estimation values;

[0017] Determine whether the current time is the predetermined end time. If so, the obtained state estimation values are taken as the result, otherwise, proceed to the next time and return to the previous step.

[0018] As an optional implementation, the power distribution network information includes power distribution network topology information, branch impedance and ground admittance information, node-to-ground capacitance information, power distribution network measurement information, and power distribution network state estimation execution period;

[0019] The chemical heat supply network information includes heat supply network topology information, heat supply pipeline length, pipeline cross-sectional area, pipeline water mass flow, pipeline friction coefficient, pipeline heat transfer coefficient, ambient temperature, heat supply network measurement information, and heat supply network state estimation execution period;

[0020] The material flow network takes the hydrogen supply network as an example, and the information includes hydrogen supply network topology, pipeline length, pipeline cross-sectional area, pipeline average hydrogen flow rate and hydrogen sound speed, pipeline friction coefficient, hydrogen temperature, hydrogen compression coefficient, hydrogen gas constant, hydrogen supply network measurement information, and hydrogen supply network state estimation execution period.

[0021] As an optional implementation, the state transition equation of the power distribution network under quasi-steady state is constructed using the back / forward sweep calculation method, and the process of constructing the power distribution network state space model combined with the power distribution network measurement model includes: setting the power distribution network state quantity matrix x e =[ V , θ ] T , including the voltage amplitude V i and the voltage phase angle θ i of each node, and the measurement information matrixz e =[ P ij,meas , Q ij,mes , I ij,meas , V meas , θ meas ,P i,meas ,Q i,meas ] T This includes partial branch active, reactive or current measurements uploaded by remote terminal units, partial node voltage amplitude and phase angle measurements uploaded by synchronous phasor units, and node injection active and reactive power pseudo-measurements generated based on historical load data.

[0022] The state-space model of the distribution network is constructed as follows:

[0023] ;

[0024] ;

[0025] Among them, subscript k express k time; u e,k The input variable matrix for the distribution network includes the active and reactive power injection at each node, Δ. u e,k The power changes injected into nodes between adjacent time points are obtained from load forecasting. f e (·) represents the nonlinear state transition equation of the distribution network; J e,k-1 According to k Jacobian matrix of distribution network calculated at time -1; ξ e,k The non-Gaussian process noise of the distribution network has the following error covariance matrix: ; z e,k For distribution network quantity measurement matrix; h e (·) represent the nonlinear measurement equations for the distribution network; ω e,k The non-Gaussian measurement noise matrix of the distribution network has the following error covariance matrix: R e,k .

[0026] As an alternative embodiment, the process of constructing the state transition equation of the material flow network, in combination with the measurement model of the material flow network, to construct the state space model of the material flow network comprises: setting the state quantity matrix of the material flow network x M =[ ρ , M ] T , including the hydrogen density of each gas network node ρ i and the hydrogen mass flow M i , the measurement information matrix z M =[ ρ i,meas , M i,meas , p i,meas ,0] T , including the hydrogen density, hydrogen mass flow and node gas pressure of part of the nodes, and 0 corresponds to the virtual measurement in the material flow network;

[0027] Assuming that the hydrogen temperature in the material flow network is constant and the hydrogen supply pipeline is horizontal, according to the information of the material flow network, the material flow momentum equation, the material flow pipeline mass balance equation and the pipeline pressure equation constructed based on fluid mechanics analysis, the partial differential-algebraic simplified model of the material flow network is established, that is:

[0028] ;

[0029] ;

[0030] ;

[0031] wherein, t and s represent time and space distance; v is the gas flow rate in the pipeline; p is the pipeline pressure; λ M is the friction coefficient of the hydrogen supply network pipeline, d is the pipeline diameter; c is the sound speed of hydrogen;

[0032] The above formula is discretized based on the Lax-Wendroff difference format, and the state transition equation of the material flow network is constructed, that is:

[0033] ;

[0034] wherein, ρ i,k is the node k at time ihydrogen density; M i,k with M j,k is k the length of the pipe ij at time t i with the inflow node j hydrogen mass flow rate; L ij is the length of the pipe ij , S ij is its cross-sectional area, and the corresponding pipe radius is d ij ; p i,k is the pressure of the node k at time t i ; is the average hydrogen flow rate in the pipe ij ;

[0035] The state space model of the material flow network is constructed in combination with the measurement model of the material flow network:

[0036] ;

[0037] ;

[0038] wherein, u M,k is the input variable matrix of the material flow network, including the hydrogen density of the gas source node and the hydrogen consumption load node; f M is the state transition equation of the material flow network, i.e., the above differentiation equation; ξ M,k is the non-Gaussian process noise of the material flow network, and its error covariance matrix is ; z M,k is the measurement matrix of the material flow network; h M is the measurement equation of the material flow network, respectively; ω M,k is the non-Gaussian measurement noise matrix of the material flow network, and its error covariance matrix is R M,k The gas continuity equation and the gas source pressure set value of the material flow network will be used as virtual measurements.

[0039] As an optional implementation, the state variable matrix of the heat supply network x H [ T s , T r ] T is set to include the heat supply temperature of each nodeT s,i with the heat recovery temperature T r,i , the measurement information matrix z M [ H , T s , T r ,0] T , including the node heat power, the partial node heating temperature and the heat recovery temperature, and 0 corresponds to the virtual measurement in the heating network; according to the dynamic thermodynamic model of the heating network considering the pipeline transmission loss, the partial differential model of the heating network in the heating network is established and discretized, and combined with the measurement model of the hydraulic network and the heating network in the heating network, a state space model of the heating network considering the temperature transmission dynamic characteristics is constructed;

[0040] Specifically, it is assumed that the heating network adopts the constant flow variable temperature working medium regulation mode, and according to the law of conservation of heat, the partial differential model of the heating network in the heating network is established:

[0041] ;

[0042] wherein, T is the heating network node temperature; m is the mass flow of hot water in the pipeline, T a is the ambient temperature; C p is the specific heat capacity of water, ρ water is the density of water; λ M is the heat dissipation coefficient of the hydraulic network;

[0043] The above formula is discretized based on the Lax-Wendroff difference format, and the state transition equation of the heating network is constructed, that is:

[0044] ;

[0045] Combined with the measurement model of the heating network, the state space model of the heating network is constructed:

[0046] ;

[0047] ;

[0048] wherein, u H,k is the input variable matrix of the heating network; f H (·) is the state transition equation of the heating network, that is, the above difference equation;ξ H,k The non-Gaussian process noise of the heating network has the following error covariance matrix: ; z H,k For heating network quantity measurement matrix; h H (·) represent the measurement equations for the heating network; ω H,k The non-Gaussian measurement noise matrix of the heating network has the following error covariance matrix: R H,k ;

[0049] The values ​​of the water flow continuity equation, nodal temperature mixing equation, and loop water pressure equation of the heating network will be used as virtual measurements to improve the measurement redundancy of the heating network.

[0050] As an alternative implementation, the process of constructing the state-space model of the coupling elements includes: for the electrolyzer, based on the chemical reaction equation for hydrogen production from water electrolysis, constructing an electrolyzer model that considers the coupling of power consumption-hydrogen production and hydrogen production-water consumption, i.e.:

[0051] ;

[0052] in, for k Molar flow rate of hydrogen produced in the electrolyzer at any given time; for k The electrical power consumed by the electrolytic cell at any given time; for k The amount of water used for hydrogen production in the electrolyzer at any given time; η EL Hydrogen production efficiency of the electrolyzer; α 0 represents the water consumption efficiency of the electrolyzer; q H2 It is the lower heat energy of hydrogen;

[0053] For green electricity ammonia synthesis equipment, based on the chemical reaction equation for ammonia synthesis, a model is constructed that considers the coupling of power consumption-hydrogen consumption, ammonia production-hydrogen consumption, and heat production-ammonia production, namely:

[0054] ;

[0055] in, , , and They are respectively k The molar flow rates of air, nitrogen, hydrogen, and ammonia produced by the ammonia synthesis equipment at all times; yes k The total electrical power consumed by the ammonia synthesis reactor at any given time; fork The heat release power of the ammonia synthesis equipment at time t; The k The ammonia production rate at time t; M air The molar mass of air, R H2 The hydrogen gas constant, T AS The air separation unit inlet gas temperature, p AS in The air separation unit compressor inlet gas pressure, p AS out The air separation unit compressor outlet gas pressure, η C The mechanical efficiency of the air separation unit compressor; P P2A N The fixed electrical power of the ammonia synthesis reactor, l The electrical power consumed to produce a unit molar flow of ammonia; η h,P2A The heat release rate of the ammonia synthesis equipment; q NH3 The heat release amount to produce a unit mass of ammonia;

[0056] For the methanol synthesis equipment, according to the chemical reaction equation of ammonia synthesis, a methanol synthesis equipment model is constructed considering the coupling of electrical power consumption-hydrogen consumption, alcohol production-hydrogen consumption, and heat production-hydrogen consumption:

[0057] ;

[0058] Wherein, The k The electrical power consumed by the methanol synthesis equipment at time t; , And are k The molar flow of carbon dioxide and hydrogen consumed by the methanol synthesis equipment at time t, and the synthesis of methanol; The k The heat release power of the methanol synthesis equipment at time t; η P2M The methanol synthesis efficiency; H P2M The reaction heat released by the synthesis of a unit mass of methanol;

[0059] For the hydrogen storage tank, according to its hydrogen storage tank energy balance constraint and production scheduling plan, its state transition equation is constructed, and the hydrogen storage tank model is built considering the coupling constraint of electrical power consumption-hydrogen consumption, i.e.:

[0060] ;

[0061] Wherein, Thek the charging state of the hydrogen storage tank at time t; and is k the mass flow rate of the hydrogen storage tank at time t; is k the electrical power consumed by the hydrogen storage tank compressor at time t; V HSS is the capacity of the hydrogen storage tank; η ch HSS and η dis HSS are the charging and discharging efficiencies of the hydrogen storage tank, respectively; T HSS is the temperature of the hydrogen storage tank at time t; p HSS in and p HSS out are the inlet and outlet pressures of the hydrogen storage tank compressor, respectively; η HC is the mechanical efficiency of the hydrogen storage tank; M H2 is the molar mass of hydrogen;

[0062] For the electric heating equipment, an electric heating equipment model is constructed according to the power consumption-heat generation constraint thereof, i.e.

[0063] ;

[0064] wherein, is k the electrical power consumed by the electric heating equipment at time t; is k the heat generation power of the electric heating equipment at time t.

[0065] As an optional implementation, based on the constructed state space models of the power distribution network, the material flow network and the heat supply network, the process of constructing a real-time state estimation model of a green chemical industrial park considering the dynamic characteristics of multi-element material flow and heterogeneous energy flow includes: setting independent estimators for the power subsystem power distribution network and the chemical subsystem heat supply network and material flow network, respectively; each network estimator performs its own state estimation based on the particle filtering algorithm according to the corresponding state space model constructed;

[0066] Based on the state variable estimation values obtained by performing its own state estimation, each network estimator calculates the power values of the coupling elements in the other two types of networks according to the coupling element model, as limited interaction information, and transmits them to the other two types of network estimators for asynchronous parallel iteration until the deviation between the equivalent power estimation values obtained by the coupling elements from the adjacent networks and the network power estimation values is less than a threshold value.

[0067] As an alternative embodiment, the process of constructing a distributed state estimation model of a green chemical industrial park based on a distributed state estimation strategy of an asynchronous parallel computing architecture includes: before state estimation of the current network, obtaining the interaction quantities of the other two types of networks as new pseudo-measurements, and if the information of the adjacent area is not updated, continuing to use the last measurement value for estimation;

[0068] When a network completes estimation, the state estimation value of the network is corrected using the obtained adjacent network information;

[0069] The equivalent power values of the other two types of networks are calculated using the state estimation values of the nodes where the coupling elements of the network are located, and are transmitted to the other two networks;

[0070] For a sub-network, the state estimation of the network converges only when the deviation between the power estimation value of the network at the nodes where all coupling elements are located and the equivalent power estimation value obtained from the adjacent network is less than the corresponding threshold value, otherwise the asynchronous iteration continues.

[0071] As an alternative embodiment, the process of introducing a new information trigger mechanism includes: the trigger value of the new information trigger mechanism is selected according to the state estimation error requirements of the green chemical industrial park by performing sensitivity analysis of the power change value of the power distribution network-state variable error of the heating network, the power change value of the power distribution network-state variable error of the material network, the heat power change value of the heating network-state variable error of the power distribution network, and the hydrogen power change value of the material network-state variable error of the power distribution network. If any threshold value is met, the new information trigger mechanism is triggered.

[0072] As an alternative embodiment, the process of determining whether to perform real-time state estimation of the electric, thermal, and material flow coordination at the current time according to the state estimation execution period of the electric, thermal, and material flow networks and the new information trigger mechanism includes: if the current time is the estimation time of the electric, thermal, and material flow networks and the new information trigger mechanism is met, performing distributed state estimation of the three and outputting the state estimation value; if the current time is the estimation time of the electric and material flow networks and the new information trigger mechanism is met, performing distributed state estimation of the two and outputting the state estimation value; if it is only the state estimation time of the power distribution network or the new information trigger mechanism is not met, only performing state estimation of the power distribution network and outputting the state estimation value.

[0073] A distributed state estimation system for a chemical industrial park considering the dynamics of multi-energy flow, comprising:

[0074] The green electric chemical industrial park parameter preprocessing module is used for dividing the green electric chemical industrial park into an electric power subsystem and a chemical subsystem, wherein the electric power subsystem manages the operation of the power distribution network, and the chemical subsystem manages the operation of the heat supply network, the chemical material flow network, and the chemical production equipment; the network information and the equipment information in each subsystem are respectively acquired, and the branches in the power distribution network, the heat supply network, and the chemical material flow network and the coupling elements of the hydrogen / ammonia / methanol chemical production equipment are subjected to parameter normalization processing;

[0075] The electric power subsystem state estimation model construction module is used for constructing, for the electric power subsystem, a power distribution network state transition equation under quasi-steady state according to the power distribution network information by using a back / forward sweep method, and constructing a power distribution network state space model in combination with a power distribution network measurement model;

[0076] The chemical subsystem state estimation model construction module is used for establishing a partial differential model of the material flow network by using a pipeline pressure equation, a material flow momentum equation, and a material flow pipeline mass balance equation, performing discretization processing, constructing a material flow network state transition equation, and constructing a material flow network state space model in combination with a material flow network measurement model; a partial differential model of the heat supply network is established by using a dynamic thermodynamic model of the heat supply network considering pipeline transmission loss, and discretization processing is performed, and a heat supply network state space model considering the temperature transmission dynamic characteristics is constructed in combination with measurement models of the hydraulic network and the heat network in the heat supply network;

[0077] The coupling element state space model construction module is used for constructing an equation constraint describing the power consumption, the raw material inflow, and the chemical production heat release of the hydrogen / ammonia / methanol chemical production equipment and the coupling elements, and constructing a coupling element state space model;

[0078] The distributed state estimation modeling module is used for constructing a green electric chemical industrial park real-time state estimation model considering the dynamic characteristics of the multi-element material flow and the heterogeneous energy flow based on the constructed power distribution network state space model, the material flow network state space model, and the heat supply network state space model, and constructing a green electric chemical industrial park distributed state estimation model based on a distributed state estimation strategy of an asynchronous parallel computing architecture;

[0079] The distributed state estimation execution module is used for introducing a innovation trigger mechanism, judging whether real-time state estimation of electricity, heat, and material flow is performed at the current time according to the state estimation execution period of the electricity, heat, and material flow network and the innovation trigger mechanism, and if so, performing distributed state estimation of the corresponding network by using the green electric chemical industrial park distributed state estimation model to obtain a state estimation value;

[0080] The state estimation end condition judgment module is used for judging whether the current time is a predetermined end time, and if so, taking the obtained state estimation value as a result, and if not, entering the next time and re-calling the distributed state estimation execution module.

[0081] Compared with the prior art, the present application has the following advantages:

[0082] The present application provides a green electricity chemical industrial park distributed state estimation method considering the dynamic characteristics of material flow and energy flow, which can fully consider the time sequence coupling and differentiated dynamic characteristics of electric, thermal heterogeneous energy flow and hydrogen / ammonia / alcohol multi-element material flow, realize efficient and accurate state estimation of electric, thermal and material flow network, and provide data basis for subsequent advanced application functions.

[0083] The present application combines asynchronous parallel computing architecture and innovation trigger mechanism to build a multi-time scale distributed state estimation architecture, which solves the data privacy problem between power and chemical subsystems, and solves the measurement uploading period difference problem and computing resource waste problem caused by the difference in time constants between different networks.

[0084] According to the time sequence coupling and differentiated dynamic characteristics of electric, thermal heterogeneous energy flow and hydrogen / ammonia / alcohol multi-element material flow, the present application builds state space models of power distribution network, heating network and chemical material flow network and coupling element models of hydrogen / ammonia / methanol chemical production equipment, thereby building a green electricity chemical industrial park real-time state estimation model considering material-energy coupling and dynamic characteristics; taking particle filtering algorithm as the framework, the present application builds distributed state estimation based on asynchronous parallel architecture, thereby solving the data privacy problem between power and chemical subsystems; the present application builds a green electricity chemical industrial park multi-time scale distributed state estimation architecture combined with innovation trigger mechanism, thereby solving the measurement uploading period difference problem and computing resource waste problem caused by the difference in time constants between different networks.

[0085] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the following preferred embodiments are described in detail below, and the accompanying drawings are used for explanation. BRIEF DESCRIPTION OF DRAWINGS

[0086] The drawings accompanying the specification of the present application form part of the present application and serve to provide a further understanding of the present application, the illustrative embodiments thereof and its description serve to explain the present application without imposing undue limitations on the present application.

[0087] Figure 1 The design scheme flowchart provided for an embodiment of the present application is provided.

[0088] Figure 2 The green electricity chemical industrial park multi-time scale distributed state estimation architecture combined with innovation trigger mechanism provided for an embodiment of the present application is provided. DETAILED DESCRIPTION

[0089] The present application will be further described below in combination with the drawings and embodiments.

[0090] It should be noted that the following detailed description is illustrative only, and is intended to provide further description of the application. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0091] It is also important to note that the terms "example" and "exemplary" as may be used herein mean "serving as an example, instance, or illustration," and should not necessarily be construed as preferred or advantageous over other examples. While the application has been illustrated by a description of examples, it is not the intention to restrict or in any way limit the scope of the application to such examples. Any numerical examples

[0092] The embodiments in the application and the features in the embodiments can be combined with each other under the condition of no conflict.

[0093] Embodiment one

[0094] According to Figure 1 , the embodiment provides a chemical industrial park distributed state estimation method considering multi-energy flow dynamics, and specifically includes the following steps:

[0095] Step 1: The green electricity chemical industrial park is divided into an electricity subsystem and a chemical subsystem according to the difference of management subjects, wherein the electricity subsystem manages the operation of the power distribution network, and the chemical subsystem manages the operation of the heat supply network, the chemical material flow network and the chemical production equipment; the network and equipment information in different subsystems is obtained respectively, and the parameter unitization processing is performed on the branches in the power distribution network, the heat supply network and the chemical material flow network and the coupling elements such as hydrogen / ammonia / methanol chemical production equipment.

[0096] Specifically, the power distribution network information includes power distribution network topology information, branch impedance and ground admittance information, node ground capacitance information, power distribution network measurement information, power distribution network state estimation execution period Δ t e ; the chemical heat supply network information includes heat supply network topology information, heat supply pipeline length, pipeline cross-sectional area, pipeline average flow rate, pipeline friction coefficient, pipeline heat transfer coefficient, environment temperature, heat supply network measurement information and heat supply network state estimation execution period Δ t h ; the material flow network takes a hydrogen supply network as a representative, and the information includes hydrogen supply network topology, pipeline length, pipeline cross-sectional area, pipeline average flow rate and hydrogen sound speed, pipeline friction coefficient, hydrogen temperature, hydrogen compression coefficient, hydrogen gas constant, hydrogen supply network measurement information and hydrogen supply network state estimation execution period Δ t M; hydrogen / ammonia / methanol chemical production equipment, hydrogen storage tank and electric heating equipment, etc. Coupling element electric-hydrogen conversion coefficient, hydrogen-heat conversion coefficient, raw material-hydrogen / hydrogen-product conversion coefficient, electric-heat conversion coefficient, etc.

[0097] Step 2: For the power subsystem, according to the power distribution network information, the state transition equation of the power distribution network under quasi-steady state is constructed based on the backflow calculation method, and the state space model of the power distribution network is constructed combined with the power distribution network measurement model.

[0098] Specifically, the power distribution network state quantity matrix x e =[ V , θ ] T , including the voltage amplitude V i and the voltage phase angle θ i , the measurement information matrix z e =[ P ij,meas , Q ij,mes , I ij,meas , V meas , θ meas ,P i,meas ,Q i,meas ] T , including part of the branch active, reactive or electric quantity measurement uploaded by the remote terminal unit (RTU), part of the node voltage amplitude and phase angle measurement uploaded by the synchronous phasor unit (PMU), and the node injected active and reactive power pseudo measurement generated according to the historical load data. The state space model of the power distribution network is constructed as follows

[0099] ;

[0100] ;

[0101] Wherein, the subscript k represents k time; u e,k is the input variable matrix of the power distribution network, including the active and reactive power injection of each node, Δ u e,k is the change of node injection power between adjacent time, which is obtained by load prediction. f e (·) is the state transition equation of the power distribution network; J e,k-1 is the state transition equation of the power distribution network; kThe power distribution network Jacobian matrix calculated at time instant t = 1; ξ e,k For the power distribution network non-Gaussian process noise, the error covariance matrix is ; z e,k For the power distribution network measurement matrix; h e (·) are the nonlinear measurement equations of the power distribution network, respectively; ω e,k For the power distribution network non-Gaussian measurement noise matrix, the error covariance matrix is R e,k .

[0102] Step 3: For the chemical subsystem, according to the material flow network, heat supply network and coupling element information, a state space model is constructed considering dynamic characteristics.

[0103] Step 3.1: Set the material flow network (i.e. hydrogen supply network) state quantity matrix x M =[ ρ , M ] T , including the hydrogen density ρ i and hydrogen mass flow M i of each gas network node, the measurement information matrix z M =[ ρ i,meas , M i,meas , p i,meas , 0] T , including the hydrogen density, hydrogen mass flow, node gas pressure of part of the nodes, and 0 corresponding to the virtual measurement in the material flow network. Assuming that the hydrogen temperature in the material flow network is constant and the hydrogen supply pipeline is horizontal, according to the material flow network information, the material flow momentum equation, material flow pipeline mass balance equation, pipeline pressure equation, etc. constructed based on fluid mechanics analysis, the partial differential-algebraic simplified model of the material flow network is established, i.e.

[0104] ;

[0105] ;

[0106] ;

[0107] wherein, t and s represent time and spatial distance; v is the gas flow rate in the pipeline; p is the pipeline pressure; λM The friction coefficient of the hydrogen supply network pipeline, d The diameter of the pipe; c The speed of sound in hydrogen gas.

[0108] Discretize the above equation using the Lax-Wendroff difference scheme to construct the state transition equation for the material flow network, namely:

[0109] ;

[0110] in, ρ i,k for k Time Node i The density of hydrogen gas; M i,k With M j,k for k Time Pipeline ij Outflow node i With inflow node j The hydrogen mass flow rate; L ij For pipelines ij Length, S ij Its cross-sectional area, and the corresponding pipe radius is d ij ; p i,k for k Time Node i air pressure; For pipelines ij Average flow rate of hydrogen gas inside the vessel.

[0111] The state-space model of the material flow network is constructed by combining the measurement model of the material flow network, as shown in the following equation.

[0112] ;

[0113] ;

[0114] in, u M,k Input variable matrix to the material flow network, including hydrogen density of gas source nodes and hydrogen load nodes. f M (·) is the state transition equation of the material flow network, i.e., the difference equation mentioned above; ξ M,k The noise of the material flow network is a non-Gaussian process, and its error covariance matrix is: ; z M,k For material flow network quantity measurement matrix; h M (·) represent the material flow network measurement equations;ω M,k is the measurement noise matrix of the material flow network, and its error covariance matrix is R M,k .

[0115] Further, the gas continuity equation of the material flow network and the gas source pressure set value will be used as virtual measurements, i.e.,

[0116] ;

[0117] ;

[0118] Step 3.2: Set the heating network state quantity matrix x H = T s , T r ] T , including the heating temperature of each node T s,i and the regenerative temperature T r,i , the measurement information matrix z M = H , T s , T r , 0] T , including the node heat power, part of the node heating temperature and the regenerative temperature, and 0 corresponds to the virtual measurement in the heating network. According to the dynamic thermodynamic model of the heating network considering the pipeline transmission loss, the partial differential model of the heating network in the heating network is established and discretized, combined with the measurement model of the hydraulic network and the heating network in the heating network, the state space model of the heating network considering the temperature transmission dynamic characteristics is constructed.

[0119] Specifically, it is assumed that the heating network adopts the constant flow variable temperature working medium regulation mode, and according to the law of conservation of heat, the partial differential model of the heating network in the heating network is established as follows,

[0120] ;

[0121] wherein, T is the heating network node temperature; m is the mass flow of hot water in the pipeline, T a is the ambient temperature; C p is the specific heat capacity of water, ρ water is the density of water; λ MThis represents the heat dissipation coefficient of the hydraulic network.

[0122] Discretize the above equation using the Lax-Wendroff difference scheme to construct the state transition equation for the heating network, i.e.

[0123] ;

[0124] A state-space model of the heating network is constructed by combining the measurement model of the heating network, as shown in the following equation.

[0125] ;

[0126] ;

[0127] in, u H,k Input variable matrix for heating network. f H (·) represents the state transition equation of the heating network, i.e., the difference equation mentioned above; ξ H,k The non-Gaussian process noise of the heating network has the following error covariance matrix: ; z H,k For heating network quantity measurement matrix; h H (·) represent the measurement equations for the heating network; ω H,k The non-Gaussian measurement noise matrix of the heating network has the following error covariance matrix: R H,k .

[0128] Furthermore, the values ​​of the water flow continuity equation, node temperature mixing equation, and loop water pressure equation of the heating network will be used as virtual measurements to improve the measurement redundancy of the heating network.

[0129] Step 3.3: For the coupled components such as hydrogen / ammonia / methanol chemical production equipment, hydrogen storage tanks and electric heating equipment, construct equation constraints describing the power consumption, raw material feed rate and heat release of chemical production, and the heat generation power of electric heating equipment, and construct the state space model of the coupled components.

[0130] Specifically, for the electrolyzer, according to the chemical reaction equation for producing hydrogen through water electrolysis, i.e. An electrolyzer model was constructed that considers the coupling of power consumption and hydrogen production, as well as the coupling of hydrogen production and water consumption.

[0131] ;

[0132] in, for k Molar flow rate of hydrogen produced in the electrolyzer at any given time; P (t) is the consumed electric power of the electrolyzer at time t; k P (t) is the consumed electric power of the electrolyzer at time t; W (t) is the consumed water quantity of the electrolyzer for hydrogen production at time t; k W (t) is the consumed water quantity of the electrolyzer for hydrogen production at time t; η EL η (t) is the hydrogen production efficiency of the electrolyzer at time t; α η (t) is the hydrogen production efficiency of the electrolyzer at time t; q H2 Q (t) is the low heat energy of hydrogen gas.

[0133] For green electricity synthesis ammonia equipment, according to the chemical reaction equation of synthesis ammonia, i.e. , the green electricity synthesis ammonia equipment model considering the coupling of consumed electric power-consumed hydrogen quantity, ammonia production quantity-consumed hydrogen quantity, and heat production quantity-ammonia production quantity is constructed, i.e.

[0134] ;

[0135] wherein, , , and are the molar flow rates of air, nitrogen, hydrogen, and ammonia gas generated by the synthesis ammonia equipment at time t; k P (t) is the total consumed electric power of the synthesis ammonia reactor at time t; P (t) is the total consumed electric power of the synthesis ammonia reactor at time t; k Q (t) is the heat release power of the synthesis ammonia equipment at time t; Q (t) is the heat release power of the synthesis ammonia equipment at time t; k N (t) is the ammonia gas production quantity at time t; N (t) is the ammonia gas production quantity at time t; k air M (air) is the molar mass of air, M H2 R (H2) is the hydrogen gas constant, R AS T (in) is the air separation device inlet gas temperature, T AS in is the air separation device compressor inlet gas pressure, p AS out is the air separation device compressor outlet gas pressure, p C η (mech) is the mechanical efficiency of the air separation device compressor; η P2A N is the fixed electric power of the synthesis ammonia reactor, P P (t) is the consumed electric power of the synthesis ammonia reactor at time t; l P (t) is the consumed electric power of the synthesis ammonia reactor at time t; η h,P2A Q (t) is the heat release rate of the synthesis ammonia equipment; q NH3 Q (t) is the heat release rate of the synthesis ammonia equipment;

[0136] For methanol synthesis equipment, according to the chemical reaction equation of synthesis ammonia, i.e. A methanol synthesis equipment model was constructed that takes into account the coupling of power consumption-hydrogen consumption, alcohol production-hydrogen consumption, and heat production-hydrogen consumption.

[0137] ;

[0138] in, for k The electrical power consumed by the methanol synthesis equipment at any given time; , and They are respectively k The amount of carbon dioxide and hydrogen consumed by the methanol synthesis equipment at any given time, and the molar flow rate of the synthesized methanol. for k The exothermic power of the methanol synthesis equipment at any given time; η P2M The methanol synthesis efficiency; Δ H P2M The heat of reaction released to synthesize one unit mass of methanol.

[0139] For hydrogen storage tanks, their state transition equations are constructed based on their energy balance constraints and production scheduling plan. Then, the coupled constraints of power consumption and hydrogen intake are considered to complete the model of the hydrogen storage tank.

[0140] ;

[0141] in, for k Monitor the constant gas level of the hydrogen storage tank; and for k Real-time inlet and outlet mass flow rates of hydrogen storage tank; for k The electrical power consumed by the compressor of the hydrogen storage tank at all times; V HSS This refers to the capacity of the hydrogen storage tank. η ch HSS and η dis and HSS represent the charging and discharging efficiency of the hydrogen storage tank, respectively. T HSS For the hydrogen storage tank inlet temperature, p HSS in and p HSS and out represent the inlet and outlet pressures of the hydrogen storage tank compressor, respectively. η HC The mechanical efficiency of the hydrogen storage tank; M H2 is the molar mass of hydrogen gas.

[0142] For electric heating equipment, a model is constructed based on its power consumption minus heat generation constraint, namely:

[0143] ;

[0144] in, for k The electrical power consumed by the electric heating equipment at all times; for k The heat output of the electric heating equipment at all times.

[0145] Step 4: Based on the state-space models of the power subsystem and chemical subsystem mentioned above, a real-time state estimation model for the green chemical industrial park that takes into account the differentiated dynamic characteristics of multiple material / energy flows is constructed. Then, based on the particle filter algorithm, a distributed state estimation strategy based on an asynchronous parallel computing architecture is used to construct a distributed state estimation model for the green chemical industrial park.

[0146] Step 4.1: Set up independent estimators for the power distribution network of the power subsystem and the heating network and material flow network of the chemical subsystem respectively. Each network estimator performs its own state estimation based on the particle filter algorithm according to the state space model constructed in Steps 2 and 3.

[0147] Specifically, each network collects data based on the prior distribution of its state variables. N For each particle, the weight is calculated recursively based on Bayesian theory. wjk Yes, the corresponding posterior distribution, its mean, and the weights of each particle are calculated as follows:

[0148] ;

[0149] ;

[0150] ;

[0151] in,{ X ( i ) k , i =1,2,…, N} is from the reference distribution q ( x k | x k-1 , z 1:k A set of random samples (particles) collected in ) δ (·) represents the Dirac-Delta function; for wjk Normalized importance weights. X 0:k As of the end k The system's state variable sequence at time t, and it is assumed that they are independent and identically distributed. z 1:k As of the end kThe measurement sequence of the state variable at time k. p X k z 1:k-1 is the probability density function of the state variable at time k for the prediction stage, p X k X k-1 is the prior density of the state variable, p z k X k is the measurement likelihood density, p X k z 1:k is the k posterior density of the state variable at time k. When the p X k X k-1 is taken as the reference distribution, the weight of each particle can be recursively calculated only according to its weight at time k-1 and the measurement likelihood probability w k . p z k X k , i.e.

[0152] .

[0153] Further, to solve the problem of particle degeneracy in the particle filter, the sampling-importance-resampling method is used to resample according to the normalized weight of each particle. After resampling, the normalized weight of each particle is 1 / N .

[0154] The state variable estimation value of each network state estimation based on the particle filter and the error can be given by the expectation and covariance thereof, i.e.

[0155] .

[0156] .

[0157] ​​​​​​​​​​​​Step 4.2: Based on the state variable estimates obtained in Step 4.1, each network estimator calculates the power values ​​of the coupled elements in the other two types of networks according to the coupled element model in Step 3.3. This power value is then transmitted to the other two types of network estimators as limited interaction information. Specifically, taking the distribution network in the power subsystem as an example, based on the estimated values ​​of the node voltage amplitude and phase angle of the nodes where the coupled elements are located in the distribution network, the hydrogen production / consumption power and heat production / consumption power of the hydrogen / ammonia / methanol chemical production equipment, hydrogen storage tanks, and electric heating equipment are calculated. This power value is used as limited interaction information for information exchange with the material flow network and heating network of the chemical subsystem. Asynchronous parallel iteration is performed until the deviation between the equivalent power estimates obtained by all coupled elements from adjacent networks and the power estimates of this network is less than a certain threshold.

[0158] ;

[0159] ;

[0160] ;

[0161] in, P couple,e i , k , P couple,H i , k and P couple,M i , k They respectively represent the use of k The power consumed by the coupling elements calculated based on the estimated state values ​​of the power distribution network, heating network, and material network at any given time; H couple,e i , k , H couple,H i , k and H couple,M i , k They respectively represent the use of k The heat power generated / used by the coupling elements calculated based on the estimated state values ​​of the power distribution network, heating network, and material network at all times; M couple,e i , k , M couple,H i , k and M couple,M i , k They respectively represent the use of kThe hydrogen power generated / used by the coupled components is calculated based on the estimated states of the distribution network, heating network, and material network at all times. This ensures both global consistency across multiple network states and data privacy for different management entities.

[0162] Furthermore, an asynchronous parallel computing architecture is constructed to achieve joint estimation of the electricity-heat-material flow network in the Green Power Chemical Industrial Park. This avoids the problem of wasted estimator waiting time caused by differences in the computational efficiency of different networks under synchronous iterative methods. The steps for asynchronous state estimation in the Green Power Chemical Industrial Park are as follows:

[0163] (1) Before performing state estimation in this network, obtain the interaction quantities of the other two types of networks as new pseudo-measures. If the information of the adjacent regions is not updated, continue to use the previous measurement value for estimation.

[0164] (2) After a network has been estimated, the state estimate of the current network is corrected using the neighboring network information obtained in step (1);

[0165] (3) The equivalent power values ​​of the other two types of networks are calculated using the state estimates of the nodes where the coupling elements of this network are located, and then transferred to the other two networks;

[0166] (4) For a certain sub-network, the network state estimation converges if and only if the deviation between the local network power estimate of all nodes where its coupled elements are located and the equivalent power estimate obtained from the adjacent network is less than a certain threshold. Otherwise, asynchronous iteration continues.

[0167] Step 5: Construct a multi-timescale distributed state estimation architecture for the green chemical industrial park, incorporating an innovation triggering mechanism. The execution cycle of the state estimation for the electricity, heat, and material flow networks (i.e., whether the innovation triggering mechanism is met) determines whether a real-time state estimation model for coordinated electricity, heat, and material flow should be performed at the current moment. Generally, the time constants of electricity flow, hydrogen flow, and heat flow increase in that order, meaning the distribution network state estimation execution cycle is the shortest, and the heating network state estimation execution cycle is the longest. Therefore, if the current moment is for electricity, heat, and material flow network estimation and meets the innovation triggering mechanism, then distributed state estimation for all three is performed and the state estimate is output. If the current moment is for electricity and material flow network estimation and meets the innovation triggering mechanism, then distributed state estimation for both is performed and the state estimate is output. If it is only for distribution network state estimation or does not meet the innovation triggering mechanism, then only distribution network state estimation is performed and the state estimate is output.

[0168] Further, the trigger value of the innovation trigger mechanism is selected according to the state estimation error requirement of the green chemical industrial park by performing sensitivity analysis of the power distribution network power change value-heat supply network state variable error, the power distribution network power change value-material network state variable error, the heat supply network heat power change value-power distribution network state variable error, and the material network hydrogen power change value-power distribution network state variable error. If any threshold value is met, the innovation trigger mechanism is triggered.

[0169] Step 6: Determine whether the current time is the predetermined end time. If yes, output the state estimation value of step 5 as the result. Otherwise, enter the next time and return to execute step 5.

[0170] Embodiment two

[0171] The embodiment provides a chemical industrial park distributed state estimation system considering multi-energy flow dynamics, comprising:

[0172] A green chemical industrial park parameter preprocessing module is configured to acquire information of a power distribution network, a heat supply network, a chemical material flow network, and hydrogen / ammonia / methanol chemical production equipment of the park, and perform unitization processing on the parameters.

[0173] An electric power subsystem state estimation model construction module is configured to construct a power distribution network state space model as an electric power subsystem state estimation model.

[0174] A chemical subsystem state estimation model construction module is configured to construct heat supply network and material flow network state space models considering dynamic characteristics, and construct hydrogen / ammonia / methanol chemical production equipment and electric heating equipment, hydrogen storage tank models as a chemical subsystem state estimation model.

[0175] A green chemical industrial park distributed state estimation modeling module is configured to construct a green chemical industrial park distributed state estimation model considering dynamic characteristics of multi-element material flow and heterogeneous energy flow based on a distributed state estimation strategy of an asynchronous parallel computing architecture.

[0176] A green chemical industrial park distributed state estimation execution module is configured to perform park multi-time scale distributed state estimation according to electric, heat, and material flow network state estimation execution cycle differences and park measurement innovation, in combination with an innovation trigger mechanism.

[0177] A green chemical industrial park state estimation end condition judgment module is configured to determine whether the above distributed state estimation reaches an estimation end condition.

[0178] Those skilled in the art will appreciate that embodiments of the present application can be readily used as software, hardware, or a combination of software and hardware. In a software embodiment, various software modules in accordance with embodiments of the present application are stored in a memory such as a computer program product (e.g., a disk storage) and executed by a computer processor. As such, various computer program products code, when executed, enable the computer to function as a special purpose computer programmed to carry out the steps described herein. The software modules, when executed, enable the computer to provide various example embodiments of the present application as discussed herein. CD - ROM

[0179] The present application is described in reference to the drawings using a flowchart and / or a block diagram of the method, apparatus (system) and computer program product according to embodiments of the application. It will be understood that each block of the flowchart and / or block diagram, and combinations of blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processing device or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified by one or more blocks Figure 1 of the flowchart and / or block diagram.

[0180] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified by one or more blocks Figure 1 of the flowchart and / or block diagram.

[0181] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart and / or block diagram block or blocks. Figure 1 one or more functions specified by one or more blocks Figure 1 of the flowchart and / or block diagram.

[0182] The foregoing is merely illustrative of the principles of this application and various modifications can be made by those skilled in the art without departing from the scope and spirit of the application. The above description is embodied in the form of a specific example of the present application and is not intended to represent the only or the best way to implement the present application. Those skilled in the art will readily recognize from this disclosure that alternative embodiments of the present application can be devised which fall within the spirit and scope of the application. Accordingly, the patent is not intended to be limited to the specific form set forth herein, but on the contrary, it is intended to cover such alternatives, modifications, equivalents, and equivalents, as can be included within the spirit and scope of the application.​

Claims

1. A distributed state estimation method for chemical industrial parks that considers multi-energy mass flow dynamics, characterized in that, Includes the following steps: The Green Power Chemical Industrial Park is divided into an electrical subsystem and a chemical subsystem. The electrical subsystem manages the operation of the power distribution network, while the chemical subsystem manages the heating network, the chemical material flow network, and the operation of chemical production equipment. The network information and equipment information of each subsystem are obtained separately, and the parameters of each branch in the power distribution network, heating network, chemical material flow network and coupling elements of hydrogen / ammonia / methanol chemical production equipment are standardized. For the power subsystem, based on the distribution network information, the reverse power flow calculation method is used to construct the quasi-steady-state distribution network state transition equation, and the distribution network state space model is constructed by combining the distribution network measurement model. For the chemical subsystem, a partial differential model of the material flow network is established and discretized based on the pressure equation, flow rate equation, and mass balance equation of the material flow network pipeline. The state transition equation of the material flow network is then constructed, and a state space model of the material flow network is built by combining the measurement model of the material flow network. Based on the dynamic thermodynamic model of the thermal network considering pipeline transmission losses, a partial differential model of the heating network is established and discretized. Combined with the measurement models of the hydraulic and thermal networks in the heating network, a state space model of the heating network considering the dynamic characteristics of temperature transmission is constructed. For hydrogen / ammonia / methanol chemical production equipment and coupling elements, an equation constraint describing the relationship between power consumption, raw material feed rate and heat release in chemical production is constructed, and a state-space model of the coupling elements is also constructed. Based on the constructed distribution network state space model, material flow network state space model and heating network state space model, a real-time state estimation model for the green electric chemical industrial park that takes into account the dynamic characteristics of multiple material flows and heterogeneous energy flows is constructed. A distributed state estimation model for the green electric chemical industrial park is constructed based on a distributed state estimation strategy of asynchronous parallel computing architecture. An innovation triggering mechanism is introduced. Based on the execution cycle of the state estimation of the electricity, heat and material flow network and the innovation triggering mechanism, it is determined whether to perform real-time state estimation of the electricity, heat and material flow coordination at the current moment. If so, the distributed state estimation of the corresponding network is performed using the distributed state estimation model of the green chemical industrial park to obtain the state estimation value. Determine if the current time is the predetermined end time. If so, use the obtained state estimate as the result; otherwise, proceed to the next time step and return to the previous step. Based on the constructed state space models of the power distribution network, material flow network, and heating network, the process of constructing a real-time state estimation model for a green electric chemical industrial park that takes into account the dynamic characteristics of multiple material flows and heterogeneous energy flows includes: setting independent estimators for the power distribution network of the power subsystem and the heating network and material flow network of the chemical subsystem respectively; and each network estimator performs its own state estimation based on the particle filter algorithm according to the corresponding state space model constructed. Each network estimator, based on the state variable estimates obtained from its own state estimation, calculates the power values ​​of the coupled elements in the other two types of networks according to the coupled element model. These values ​​are then passed to the other two types of network estimators as limited interaction information. The process is carried out asynchronously and in parallel until the deviation between the equivalent power estimates obtained by all coupled elements from the neighboring networks and the power estimates of the current network is less than a threshold.

2. The distributed state estimation method for chemical industrial parks considering multi-energy mass flow dynamics as described in claim 1, characterized in that, The power distribution network information includes power distribution network topology information, impedance and ground admittance information of each branch, ground capacitance information of each node, power distribution network measurement information, and power distribution network state estimation execution cycle. The information on the chemical heating network includes heating network topology information, length of each heating pipe, cross-sectional area of ​​the pipe, mass flow rate of hot water in the pipe, pipe friction coefficient, pipe heat transfer coefficient, ambient temperature, heating network measurement information, and the execution cycle of heating network status estimation. Material flow networks are represented by hydrogen supply networks. Information includes the hydrogen supply network topology, pipeline length, pipeline cross-sectional area, average hydrogen flow velocity and hydrogen sound velocity in the pipeline, pipeline friction coefficient, hydrogen temperature, hydrogen compressibility coefficient, hydrogen gas constant, hydrogen supply network measurement information, and hydrogen supply network status estimation execution cycle.

3. The distributed state estimation method for chemical industrial parks considering multi-energy mass flow dynamics as described in claim 1, characterized in that, The process of constructing the distribution network state transition equations under quasi-steady state using the reverse power flow calculation method, and then constructing the distribution network state-space model in conjunction with the distribution network measurement model, includes: setting the distribution network state quantity matrix. x e =[ V , θ ] T Including the voltage amplitude of each node V i and voltage phase angle θ i Measurement information matrix z e =[ P ij,meas , Q ij,mes , I ij,meas , V meas , θ meas ,P i,meas ,Q i,meas ] T This includes partial branch active, reactive or current measurements uploaded by remote terminal units, partial node voltage amplitude and phase angle measurements uploaded by synchronous phasor units, and node injection active and reactive power pseudo-measurements generated based on historical load data. The state-space model of the distribution network is constructed as follows: ； ； Among them, subscript k express k time; u e,k The input variable matrix for the distribution network includes the active and reactive power injection at each node, Δ. u e,k The power changes injected into nodes between adjacent time points are obtained from load forecasting. f e (·) represents the nonlinear state transition equation of the distribution network; J e,k-1 According to k Jacobian matrix of distribution network calculated at time -1; ξ e,k The non-Gaussian process noise of the distribution network has the following error covariance matrix: ; z e,k For distribution network quantity measurement matrix; h e (·) represent the nonlinear measurement equations for the distribution network; ω e,k The non-Gaussian measurement noise matrix of the distribution network has the following error covariance matrix: R e,k .

4. The distributed state estimation method for chemical industrial parks considering multi-energy mass flow dynamics as described in claim 1, characterized in that, The process of constructing the state transition equations of the material flow network and building the state space model of the material flow network by combining the measurement model of the material flow network includes: setting the state variable matrix of the material flow network. x M =[ ρ , M ] T This includes the hydrogen density at each gas network node. ρ i and hydrogen mass flow rate M i Measurement information matrix z M =[ ρ i,meas , M i,meas , p i,meas ,0] T This includes hydrogen density, hydrogen mass flow rate and node pressure at some nodes, with 0 corresponding to virtual measurements in the material flow network. Assuming the hydrogen temperature remains constant and the hydrogen supply pipeline is horizontal in the material flow network, based on the material flow network information and the material flow rate equation, material flow pipeline mass balance equation, and pipeline pressure equation constructed based on fluid dynamics analysis, a partial differential-algebraic simplified model of the material flow network is established, namely: ； ； ； in, t and s Indicates time and spatial distance; v The gas velocity inside the pipe; p For pipeline pressure; λ M The friction coefficient of the hydrogen supply network pipeline, d The diameter of the pipe; c The speed of sound in hydrogen gas; Discretize the above equation using the Lax-Wendroff difference scheme to construct the state transition equation for the material flow network, namely: ； in, ρ i,k for k Time Node i The density of hydrogen gas; M i,k With M j,k for k Time Pipeline ij Outflow node i With inflow node j The hydrogen mass flow rate; L ij For pipelines ij Length, S ij Its cross-sectional area, and the corresponding pipe radius is d ij ; p i,k for k Time Node i air pressure; For pipelines ij Average internal hydrogen flow rate; Construct a state-space model of the material flow network by combining the measurement model of the material flow network: ； ； in, u M,k Input variable matrix for material flow network, including hydrogen density of gas source nodes and hydrogen load nodes; f M (·) is the state transition equation of the material flow network, i.e., the difference equation mentioned above; ξ M,k The noise of the material flow network is a non-Gaussian process, and its error covariance matrix is: ; z M,k For material flow network quantity measurement matrix; h M (·) represent the material flow network measurement equations; ω M,k Here is the non-Gaussian measurement noise matrix of the material flow network, and its error covariance matrix is: R M,k The gas continuity equation and gas source pressure setpoint of the material flow network will be used as virtual measurements.

5. The distributed state estimation method for chemical industrial parks considering multi-energy mass flow dynamics as described in claim 1, characterized in that, Set the heating network state variable matrix x H =[ T s , T r ] T Including the heating temperature at each node T s,i With regeneration temperature T r,i Measurement information matrix z M =[ H , T s , T r ,0] T This includes node thermal power, partial node heating temperature and regeneration temperature, with 0 corresponding to virtual measurements in the heating network. Based on the dynamic thermodynamic model of the heating network considering pipeline transmission losses, a partial differential model of the thermal network in the heating network is established and discretized. Combined with the measurement models of the hydraulic network and thermal network in the heating network, a state-space model of the heating network considering the dynamic characteristics of temperature transmission is constructed. Specifically, assuming the heating network adopts a constant flow rate and variable temperature working fluid regulation method, a partial differential model of the thermal network in the heating network is established based on the law of conservation of heat: ； in, T Temperature at heating network nodes; m This refers to the mass flow rate of hot water in the pipe. T a The ambient temperature; C p The specific heat capacity of water, ρ water The density of water; λ H The heat dissipation coefficient of the hydraulic network; t and s S represents time and spatial distance. ij Let be the cross-sectional area of ​​pipe ij; Discretize the above equation using the Lax-Wendroff difference scheme to construct the state transition equation for the heating network, namely: ； Construct a state-space model of the heating network by combining the measurement model of the heating network: ； ； in, u H,k Input variable matrix for heating network; f H (·) represents the state transition equation of the heating network, i.e., the difference equation mentioned above; ξ H,k The non-Gaussian process noise of the heating network has the following error covariance matrix: ; z H,k For heating network quantity measurement matrix; h H (·) represent the measurement equations for the heating network; ω H,k The non-Gaussian measurement noise matrix of the heating network has the following error covariance matrix: R H,k ; The values ​​of the water flow continuity equation, nodal temperature mixing equation, and loop water pressure equation of the heating network will be used as virtual measurements to improve the measurement redundancy of the heating network.

6. The distributed state estimation method for chemical industrial parks considering multi-energy mass flow dynamics as described in claim 1, characterized in that, The process of constructing the state-space model of the coupled elements includes: For the electrolyzer, based on the chemical reaction equation for hydrogen production from water electrolysis, constructing an electrolyzer model that considers the coupling of power consumption and hydrogen production, and hydrogen production and water consumption, i.e.: ； in, for k Molar flow rate of hydrogen produced in the electrolyzer at any given time; for k The electrical power consumed by the electrolytic cell at any given time; for k The amount of water used for hydrogen production in the electrolyzer at any given time; η EL Hydrogen production efficiency of the electrolyzer; For the water consumption efficiency of the electrolyzer; It is the lower heat energy of hydrogen; For green electricity ammonia synthesis equipment, based on the chemical reaction equation for ammonia synthesis, a model is constructed that considers the coupling of power consumption-hydrogen consumption, ammonia production-hydrogen consumption, and heat production-ammonia production, namely: ； in, , , and They are respectively k The molar flow rates of air, nitrogen, hydrogen, and ammonia produced by the ammonia synthesis equipment at all times; yes k The total electrical power consumed by the ammonia synthesis reactor at any given time; for k The heat release power of the ammonia synthesis equipment at all times; for k Ammonia production at any given time; M air The molar mass of air, Where is the gas constant of hydrogen. This refers to the inlet air temperature of the air separator. This refers to the intake air pressure of the air separator compressor. This refers to the outlet air pressure of the air separator compressor. The mechanical efficiency of the compressor in the air separation device; For the fixed electrical power of the ammonia synthesis reactor, l The electrical power consumed per unit molar flow rate of ammonia produced; η h,P2A The heat release rate of the ammonia synthesis equipment; q NH3 The heat released per unit mass of ammonia produced; For methanol synthesis equipment, based on the chemical reaction equation for ammonia synthesis, a methanol synthesis equipment model considering the coupling of power consumption-hydrogen consumption, alcohol production-hydrogen consumption, and heat production-hydrogen consumption is constructed: ； in, for k The electrical power consumed by the methanol synthesis equipment at any given time; , and They are respectively k The amount of carbon dioxide and hydrogen consumed by the methanol synthesis equipment at any given time, and the molar flow rate of the synthesized methanol. for k The exothermic power of the methanol synthesis equipment at any given time; η P2M For methanol synthesis efficiency; The heat of reaction released for the synthesis of one unit mass of methanol; For hydrogen storage tanks, their state transition equations are constructed based on their energy balance constraints and production scheduling plan. Taking into account the coupling constraints of power consumption and hydrogen intake, the hydrogen storage tank model is completed, i.e.: ； in, for k Monitor the constant gas level of the hydrogen storage tank; and for k Real-time inlet and outlet mass flow rates of hydrogen storage tank; for k The electrical power consumed by the compressor of the hydrogen storage tank at all times; This refers to the capacity of the hydrogen storage tank. and These are the charging and discharging efficiencies of the hydrogen storage tank, respectively. T HSS For the hydrogen storage tank inlet temperature, and These are the inlet and outlet pressures of the hydrogen storage tank compressor, respectively. η HC For the mechanical efficiency of the hydrogen storage tank; This refers to the molar mass of hydrogen gas. For electric heating equipment, a model is constructed based on its power consumption minus heat generation constraint, namely: ； in, for k The electrical power consumed by the electric heating equipment at all times; for k The heat output of the electric heating equipment at all times.

7. The distributed state estimation method for chemical industrial parks considering multi-energy mass flow dynamics as described in claim 1, characterized in that, The process of constructing a distributed state estimation model for the Green Power Chemical Industrial Park based on an asynchronous parallel computing architecture includes: before performing state estimation on the current network, obtaining the interaction quantities of the other two types of networks as new pseudo-measurements; if the information of adjacent areas has not been updated, then continuing to use the previous measurement value for estimation. Once a network has completed its estimation, the estimated state of the current network is corrected using the information obtained from neighboring networks. The equivalent power values ​​of the other two types of networks will be calculated using the state estimates of the nodes where the coupling elements of this network are located, and then transferred to the other two networks. For a given subnetwork, the network state estimation converges if and only if the deviation between the local network power estimate of all nodes containing its coupled elements and the equivalent power estimate obtained from the neighboring network is less than the corresponding threshold; otherwise, asynchronous iteration continues.

8. A distributed state estimation method for chemical industrial parks considering multi-energy mass flow dynamics as described in claim 1, characterized in that it introduces... The process of the innovation triggering mechanism includes: the trigger value of the innovation triggering mechanism is selected based on the sensitivity analysis of the power change value of the distribution network - the state variable error of the heating network, the power change value of the distribution network - the state variable error of the material network, the heat power change value of the heating network - the state variable error of the distribution network, and the hydrogen power change value of the material network - the state variable error of the distribution network. If any threshold is met, the innovation triggering mechanism is triggered. The process of determining whether to perform real-time state estimation of electricity, heat, and material flow in coordination based on the execution cycle of the electricity, heat, and material flow network state estimation and the innovation triggering mechanism includes: if the current time is the time for electricity, heat, and material flow network estimation and the innovation triggering mechanism is met, then distributed state estimation of the three is performed and the state estimation value is output; if the current time is the time for electricity and material flow network estimation and the innovation triggering mechanism is met, then distributed state estimation of the two is performed and the state estimation value is output; if it is only the time for power distribution network state estimation or the innovation triggering mechanism is not met, then only power distribution network state estimation is performed and the state estimation value is output.

9. A distributed state estimation system for chemical industrial parks that considers multi-energy mass flow dynamics, using the method described in claim 1, characterized in that: include: The Green Power Chemical Industrial Park parameter preprocessing module is used to divide the Green Power Chemical Industrial Park into an electrical subsystem and a chemical subsystem. The electrical subsystem manages the operation of the power distribution network, while the chemical subsystem manages the operation of the heating network, the chemical material flow network, and the chemical production equipment. The module acquires network and equipment information from each subsystem and performs per-unit parameter standardization on each branch of the power distribution network, heating network, chemical material flow network, and coupling components of the hydrogen / ammonia / methanol chemical production equipment. The power subsystem state estimation model construction module is used to construct the quasi-steady-state distribution network state transition equations based on distribution network information and the reverse power flow calculation method for the power subsystem, and to construct the distribution network state space model in combination with the distribution network measurement model. The chemical subsystem state estimation model construction module is used to establish a partial differential model of the material flow network for chemical subsystems based on the pressure equation, flow rate equation, and mass balance equation of the material flow network pipeline, and then discretize it to construct the state transition equation of the material flow network. Combined with the measurement model of the material flow network, a state space model of the material flow network is constructed. Based on the dynamic thermodynamic model of the thermal network considering pipeline transmission losses, a partial differential model of the heating network is established and discretized. Combined with the measurement models of the hydraulic and thermal networks in the heating network, a state space model of the heating network considering the dynamic characteristics of temperature transmission is constructed. The coupling element state space model construction module is used to construct equation constraints describing the relationship between power consumption, raw material feed rate and heat release in chemical production for hydrogen / ammonia / methanol chemical production equipment and coupling elements, and to construct the state space model of coupling elements. The distributed state estimation modeling module is used to construct a real-time state estimation model for the green chemical industrial park that takes into account the dynamic characteristics of multiple material flows and heterogeneous energy flows, based on the constructed distribution network state space model, material flow network state space model and heating network state space model. It also constructs a distributed state estimation model for the green chemical industrial park based on the distributed state estimation strategy of asynchronous parallel computing architecture. The distributed state estimation execution module is used to introduce an innovation triggering mechanism. Based on the execution cycle of the state estimation of the electricity, heat and material flow network and the innovation triggering mechanism, it determines whether to perform real-time state estimation of the electricity, heat and material flow coordination at the current moment. If so, the distributed state estimation of the corresponding network is performed using the distributed state estimation model of the Green Electric Chemical Industrial Park to obtain the state estimation value. The state estimation termination condition judgment module is used to determine whether the current time is the predetermined termination time. If so, the obtained state estimation value is used as the result; otherwise, the process proceeds to the next time step and the distributed state estimation execution module is called again.

Images