Multi-time-scale state estimation method based on hydrogen-mixed electricity-gas integrated energy system

Abstract

The invention discloses a multi-time-scale state estimation method based on a hydrogen-mixed electricity-gas comprehensive energy system. According to the method, the influence of hydrogen injection on operation parameters of all nodes of a natural gas network is considered, a mixed hydrogen net dynamic state estimation model is provided, and a mixed hydrogen net pressure-flow and hydrogen advection-diffusion dynamic equation is constructed. Different dynamic response time scales of a power grid and a gas grid are considered, a segmented state estimation algorithm based on a time sequence unification theory is adopted, different state estimation time scales of the power grid and the gas grid are selected, and a state estimation model of the hydrogen-mixed electricity-gas integrated energy system is constructed on the basis of a self-adaptive volume Kalman filtering algorithm.

Description

Technical Field

[0001] The present invention proposes a multi-time scale state estimation method based on a hydrogen-blended electric-gas integrated energy system, belonging to the field of integrated energy systems. Background Art

[0002] Under the background of "dual carbon", new energy sources mainly based on wind power and photovoltaic power are highly proportionally connected to the power grid. While improving the energy structure, problems such as new energy consumption and unstable operation are brought. Technologies such as hydrogen production by electrolysis and natural gas hydrogen blending provide new solutions to such problems. With the continuous development of the hydrogen-blended electric-gas integrated energy system, the scale and proportion of natural gas hydrogen blending are continuously increasing. Different hydrogen blending ratios will result in different combustion indices and performance indices, bringing difficulties to pipeline transportation and energy regulation. Accurately monitoring the hydrogen blending ratio at each node of the hydrogen-blended gas network can effectively improve the operation safety of the integrated energy system, achieve accurate billing for the gas network, and optimize the operation strategy.

[0003] There are currently no relatively mature management and monitoring means for natural gas hydrogen-blended pipelines, and there are differences such as different time scales and different management entities among the subsystems of the hydrogen-blended electric-gas integrated energy system. The data collected by existing measurement devices cannot accurately and comprehensively cover the requirements in actual applications. To ensure the safe, efficient, and stable operation of the hydrogen-blended integrated energy system, it is particularly important to perceive the state variables of each node in real time and accurately predict them. State estimation can filter the original data of the hydrogen-blended electric-gas integrated energy system, eliminate the bad data in the measurement information, and turn the initial measurement data into reliable processed data. On this basis, unified management and optimal scheduling of each energy flow subsystem can be realized.

[0004] In recent years, the research scope of experts and scholars has mostly been in the fields of energy flow calculation of hydrogen-blended gas networks and state estimation of integrated energy systems. The energy flow calculation of hydrogen-blended gas networks mainly constructs a dynamic energy flow model of the hydrogen-blended gas network based on the significant changes in the gas quality, node pressure, flow rate, etc. of each gas network pipeline after hydrogen blending, and obtains the node state variables of each time and space of the hydrogen-blended electric-gas integrated energy system through calculation; the state estimation of the integrated energy system uses methods such as the weighted least squares method, the Kalman filter algorithm, and the sequential cooperative estimation method with a unified time scale to perform efficient and accurate state estimation on the state variables of integrated energy systems such as electric-gas and electric-gas-heat. However, the current research on the state estimation of hydrogen-blended electric-gas integrated energy systems is still relatively blank. Combining the above two aspects, namely constructing a dynamic model of the hydrogen-blended electric-gas network and a dynamic state estimation algorithm, an effective state estimation of the hydrogen-blended electric-gas integrated energy system can be obtained. Therefore, the present invention proposes a segmented cubature Kalman filter state estimation method based on a dynamic model of the hydrogen-blended gas network, constructs a state estimation model of the hydrogen-blended electric-gas integrated energy system, and uses a segmented adaptive cubature Kalman filter algorithm based on the unified theory of time series for state estimation, which can effectively improve the estimation accuracy and algorithm performance. Summary of the Invention

[0005] Object of the Invention. In view of the above problems, the present invention provides a multi-time scale state estimation method based on a hydrogen-blended electric-gas integrated energy system, which can provide relatively accurate state estimation information.

[0006] Technical Solution. In order to achieve the above object of the invention, the present invention proposes a multi-time scale state estimation method based on a hydrogen-blended electric-gas integrated energy system, including the following steps:

[0007] Step 1: Obtain the topological structure, coupling information and boundary conditions of the power grid and the hydrogen-blended gas network of the hydrogen-blended electric-gas integrated energy system, and construct the overall architecture of the hydrogen-blended electric-gas integrated energy system;

[0008] Step 2: Establish a dynamic state estimation model for the hydrogen-blended electric-gas integrated energy system based on the system architecture, including a dynamic state estimation model for the power system based on the two-exponential smoothing method, a dynamic state estimation model for the hydrogen-blended gas network based on the implicit finite difference method, a hydrogen advection-diffusion model, a coupling device model established considering the connection between the power grid and the hydrogen-blended gas network, and an energy balance equation;

[0009] Step 3: Let the state estimation time of the power system be Δt, and the state estimation time of the hydrogen-blended gas network be nΔt. When the running time is a multiple of Δt, only the power grid state estimation is performed alone, that is, the dynamic state estimation model of the power system constructed in Step 2 is used, and the dynamic state estimation result of the power system is solved by the cubature Kalman filter algorithm; when the running time is a multiple of nΔt, both the power grid and the hydrogen-blended gas network perform state estimation, that is, the dynamic state estimation models of the power system and the hydrogen-blended gas network in Step 2 and the cubature Kalman wave algorithm are used to solve the dynamic state estimation results of the power grid and the hydrogen-blended gas network, and the boundary information between the power grid and the hydrogen-blended gas network is exchanged considering the boundary coupling relationship and the energy balance equation to realize the iterative update of the state information;

[0010] Step 4: By using different state estimation time scales for different subsystems, the dynamic state estimation result of the segmented hydrogen-blended electric-gas integrated energy system is obtained.

[0011] Further, the obtaining of the topological structure, coupling information and boundary conditions of the power grid and the hydrogen-blended gas network of the hydrogen-blended electric-gas integrated energy system in Step 1, and the construction of the overall architecture of the hydrogen-blended electric-gas integrated energy system are specifically as follows:

[0012] Using the 24-node Belgian power grid and the 20-node hybrid hydrogen gas network as the basic systems, P2H devices are installed at nodes 2, 3, 5, 6, 7, 8, and 9 in the power grid. Hydrogen is introduced into the natural gas network at nodes 2, 4, 7, and 10. The gas turbines at nodes 2, 6, 8, 15, 17, and 22 in the power grid are connected to the natural gas network at nodes 4, 6, 11, and 13. Wind turbines are connected to nodes 1 to 4 of the power system, and photovoltaic units are connected to nodes 5 to 10. The measured values of state estimation are obtained by adding Gaussian noise with a mean of zero and a standard deviation to the power flow calculation; the initial standard deviations of voltage magnitude, voltage phase angle, active power, and reactive power are set to 0.005, 0.002, 0.02, and 0.02 respectively; the initial standard deviation of the measurement error of the hybrid hydrogen gas network is set to 0.01.

[0013] Furthermore, the dynamic state estimation model of the hybrid hydrogen power-gas integrated energy system established based on the system architecture in step 2 includes a dynamic state estimation model of the power system based on the two-exponential smoothing method, a dynamic state estimation model of the hybrid hydrogen gas network based on the implicit finite difference method, a hydrogen advection-diffusion model, a coupling device model established considering the connection between the power grid and the hybrid hydrogen gas network, and the energy balance equation. The process is as follows:

[0014] (2.1) Dynamic state estimation model of the power system

[0015] The power system usually uses Holt's two-parameter exponential smoothing method to construct the state space equation, reassign the weights of the estimated value and the predicted value at time k-1, and obtain the state and predicted value at time k. The specific formula is as follows:

[0016]

[0017] In the formula, is the predicted value of the state quantity at time k; α k-1 is the horizontal component at time k-1; b k-1 is the vertical component at time k-1; α H and β H are smoothing parameters, and their value ranges are in [0,1];

[0018] According to the power flow calculation formula of the power system, the measurement equation of the power grid state estimation can be obtained as follows:

[0019]

[0020] In the formula, P i , Q i are the active power and reactive power of node i respectively; V i , V j are the voltage magnitude of node i and the voltage magnitude of node j respectively; θ i , θ jThey are the voltage phases of node i and node j respectively; G ij , B ij They are the conductance between lines ij and the susceptance between lines ij respectively; j ∈ i represents all the lines connected to node i as line ij;

[0021] (2.2) Dynamic state estimation model of hydrogen - mixed gas network

[0022] The technology of hydrogen - mixed natural gas relies on the operation of existing natural gas pipelines. To construct the state estimation model of the hydrogen - mixed gas network, it is necessary to first construct the natural gas dynamic state estimation model. Considering the flow of natural gas as a one - dimensional fluid, the following assumptions are made according to the operating characteristics of natural gas: ① The radial change of natural gas is ignored, and it is only considered to flow along the axial direction; ② The operating speed and temperature of the natural gas flow remain unchanged, and the flow velocity is less than the speed of sound; ③ The pipeline is placed horizontally, ignoring the influence brought by the pipeline angle;

[0023] According to the law of conservation of mass and Newton's second law of motion, a natural gas dynamic state estimation model is constructed. Considering the influence on pipeline pressure drop and ignoring the square term of natural gas flow velocity, the following equations are obtained:

[0024]

[0025] In the formula, ρ is the density of natural gas; t is the time; u is the flow velocity of natural gas; x is the pipeline distance; p is the pressure of each node in the natural gas network; λ is the hydraulic friction coefficient; d is the pipeline diameter;

[0026] The gas in the pipeline is a mixed gas of hydrogen and natural gas. Considering the relationship:

[0027] M = ρuA

[0028] and

[0029] p = ω 2 ρ = zRTρ

[0030] In the formula, M is the mass flow rate of the mixed gas; A is the cross - sectional area of the pipeline; ω is the speed of sound propagation in the gas, z is the compression factor of the mixed gas; R is the gas constant; T is the temperature;

[0031] Substitute ρu and ρ in the natural gas dynamic state estimation model with the above two equations to get the following formula:

[0032]

[0033] Hydrogen is injected into the natural gas pipeline, changing the gas operation state and physical property parameters in the pipeline, that is, changing the density, compression factor and hydraulic friction coefficient of the mixed gas. According to the characteristics of hydrogen - mixed natural gas, calculate the values of each parameter of the mixed gas. First, calculate the compression factor z mix :

[0034]

[0035] In the formula, T mix and T 0 are the critical temperature of the mixed gas and the temperature under standard conditions, respectively; p mix are the average pressures at both ends of the pipeline and the critical pressure of the mixed gas, respectively; p i and p j are the pressures at node i and node j, respectively; T g and T H2 are the critical temperature of natural gas and the critical temperature of hydrogen, respectively; p g and p H2 are the critical pressure of natural gas and the critical pressure of hydrogen, respectively; θ H2 is the volume ratio of hydrogen in the mixed gas;

[0036] Calculate the hydraulic friction coefficient λ mix :

[0037]

[0038] Among them, the value of the Reynolds number Re is:

[0039]

[0040] In the formula, K is the absolute roughness of the inner wall of natural gas, taking 0.0508 mm; Q ij is the volume flow rate of the mixed gas between nodes i and j; ρ mix is the standard density of the mixed gas; μ mix is the viscosity of the mixed gas;

[0041] Among them, the density ρ mix of the mixed gas is calculated by the formula:

[0042]

[0043] Use the Wilke semi-empirical relationship to calculate the viscosity μ mix of the mixed gas:

[0044]

[0045] In the formula, ρ H2 and ρ g are the densities of hydrogen and natural gas, respectively; μ H2 and μ g are the viscosities of hydrogen and natural gas, respectively; M H2 and M g are the molar masses of hydrogen and natural gas, respectively;

[0046] Substitute the compressibility factor and hydraulic friction coefficient calculated after considering hydrogen injection into the natural gas dynamic state estimation model, and use the implicit finite difference method to process this model. The finally discretized dynamic state estimation model of the hydrogen-mixed gas network is as follows:

[0047]

[0048] In the formula, Δt is the unit time; Δx is the unit distance; u ij is the average flow velocity of the mixed gas between pipelines i and j; p j,t+1 is the pressure at node j at time t + 1; p i,t+1 is the pressure at node i at time t + 1; p j,t is the pressure at node j at time t; p i,t is the pressure at node i at time t; M j,t+1 is the mass flow rate of the mixed gas at node j at time t + 1; M i,t+1 is the mass flow rate of the mixed gas at node i at time t + 1; M j,t is the mass flow rate of the mixed gas at node j at time t; M i,t is the mass flow rate of the mixed gas at node i at time t;

[0049] (2.3) Hydrogen Advection-Diffusion Model

[0050] To characterize the relationship between the hydrogen volume fraction and the flow velocity after hydrogen injection into the natural gas pipeline, as well as the gas quality differences of each pipeline and node, considering the advection-diffusion equation, a partial differential equation for the transmission process of hydrogen in the mixed gas due to advection and diffusion is constructed;

[0051] The advection-diffusion equation considering only one-dimensional gas flow is:

[0052]

[0053] In the formula, C is the substance concentration; t is the time; u is the gas flow velocity; D is the diffusion coefficient;

[0054] When it is assumed that the gas flow velocity between pipelines ij is the average flow velocity u ij at this time, the above expression is simplified to:

[0055]

[0056] In the formula, x is the pipeline distance, and u ij The calculation formula of is as follows:

[0057]

[0058] In the formula, M ij is the mass flow rate of the mixed gas between nodes i and j; ρ ijis the density of the mixed gas between nodes i and j; A ij is the cross-sectional area of the pipeline between nodes i and j;

[0059] Introduce the auxiliary variable m H2 , and use the hydrogen mass flow rate of the mixed gas to replace the concentration C. The equation is transformed into:

[0060]

[0061] The calculation formula for the hydrogen mass flow rate is as follows:

[0062]

[0063] Using the implicit finite difference method to process the hydrogen advection-diffusion model, the final model is as follows:

[0064]

[0065] In the formula, is the hydrogen mass flow rate at node j at time t + 1; is the hydrogen mass flow rate at node i at time t + 1; is the hydrogen mass flow rate at node j at time t; is the hydrogen mass flow rate at node i at time t;

[0066] (2.4) Coupling equipment model

[0067] The relationship between the electric energy consumed by the P2H device and the converted gas energy is as follows:

[0068] E i,t =η P2G P k,t

[0069]

[0070] In the formula, η k,i is the energy conversion efficiency of the P2G device; P k,t is the electric power consumed by the P2G device at node k in the power system at time t; E k,t is the hydrogen energy converted by the P2G device to node i of the hydrogen-blended natural gas network at time t; F k,t is the hydrogen flow rate converted by the P2G device to node i of the hydrogen-blended gas network at time t; H H2 is the calorific value of hydrogen;

[0071] (2.5) Energy balance equation

[0072] Considering that hydrogen injection will affect the operating characteristics of the gas network, an energy balance equation is constructed with energy flow as the variable instead of the volume flow used in the pure gas network system. The equation is constructed based on the fact that the inflow energy at each node is equal to the outflow energy, as follows:

[0073]

[0074] In the formula, G w (m) represents the set of natural gas source points connected to node m; G h (m) represents the set of power-to-hydrogen devices connected to node m; G e (m) represents the set of gas loads connected to node m; G g (m) represents the set of gas turbine units connected to node m; G(m) represents the set of natural gas pipelines connected to node m; G k (m) represents the set of compressors connected to node m; is the percentage of the gas flow consumed by compressor k in the transported flow; is the energy provided by natural gas source w at time t; is the energy provided by power-to-hydrogen device h at time t; is the actual energy consumed by natural gas load e at time t; is the gas energy consumed by gas turbine unit v at time t; E mn,t is the gas energy in pipeline mn at time t; is the gas energy passing through compressor k at time t;

[0075] There is an equation relationship between energy flow E and volume flow F, that is, E = FH, where H is the calorific value of the gas. Assuming that the gas flowing out of the node is completely mixed, the energy conservation equation is obeyed before and after hydrogen mixing at each node:

[0076]

[0077] In the formula, F mn,t is the gas volume flow in pipeline mn at time t; is the calorific value of the mixed gas in pipeline mn at time t; is the volume flow provided by power-to-hydrogen device h at time t; is the calorific value of hydrogen; is the volume flow provided by natural gas source w at time t; H gas is the calorific value of natural gas; H m,t is the mixed calorific value of node m at time t.

[0078] Further, the power system state estimation time in step 3 is set as Δt, and the hybrid hydrogen network state estimation time is set as nΔt. When the running time is a multiple of Δt, only the power grid state estimation is performed separately, that is, the power system dynamic state estimation model constructed in step 2 is adopted, and the power system dynamic state estimation result is solved by the cubature Kalman filter algorithm. When the running time is a multiple of nΔt, both the power grid and the hybrid hydrogen network perform state estimation, that is, the power system dynamic state estimation model, the hybrid hydrogen network dynamic state estimation model, and the cubature Kalman wave algorithm in step 2 are used to solve the dynamic state estimation results of the power grid and the hybrid hydrogen network, and the boundary information exchange between the power grid and the hybrid hydrogen network is carried out considering the boundary coupling relationship and the energy balance equation to realize the iterative update of the state information. The specific method is as follows:

[0079] (3.1) Multi-time scale state estimation

[0080] The dynamic state estimations of the two sub-networks are carried out by setting different time scales respectively. The time scales for the dynamic state updates of the power grid and the hybrid hydrogen network are Δt and nΔt. If only the power grid state estimation time scale is satisfied at this moment, the boundary state of the power grid is directly determined according to the gas network state and the boundary coupling constraint conditions, and the power grid dynamic state estimation is carried out separately. If the power grid and gas network state estimation time scales are satisfied simultaneously at this moment, the dynamic state estimations of the power grid and the hybrid hydrogen network are carried out simultaneously, and information is exchanged through the coupling equipment and boundary conditions;

[0081] (3.2) Adaptive cubature Kalman filter

[0082] The adaptive cubature Kalman filter algorithm is used for the state estimations of the power grid and the hybrid hydrogen network at each time scale. The steps of this algorithm are as follows:

[0083] 1) Prediction step: Estimate the state quantity and covariance matrix at the current moment through the previously obtained power system and hybrid hydrogen network dynamic state estimation models. Among them, the state quantity of the power grid dynamic state estimation is x e = [V, θ], where V is the power grid node voltage and θ is the power grid node phase angle; the state quantity of the hybrid hydrogen network is p is the gas network node pressure, is the hydrogen flow rate of the gas network node;

[0084] 2) Update step: Update the covariance matrix, gain, and state estimation result according to the measurement and the dynamic state estimation model. Among them, the power grid measurement is y e = [P, Q, V, θ], where P is the active power of the power grid node and Q is the reactive power of the power grid node; the hybrid hydrogen network measurement is M is the mixed gas flow rate of the gas network;

[0085] 3) Adaptive: By introducing the fading memory exponentially weighted method, the noise covariance matrix is updated, and finally the state estimation results of the segmented hydrogen - blended electric - gas integrated energy system are obtained.

[0086] Furthermore, the dynamic state estimation results of the segmented hydrogen - blended electric - gas integrated energy system obtained by using different state estimation time scales for different subsystems in step 4 are as follows:

[0087] Calculate the root - mean - square error and the measurement estimation error respectively. If the sum of the root - mean - square error and the measurement estimation error is less than the preset value, the state estimation meets the requirements.

[0088] 1) Introduce the root - mean - square error formula:

[0089]

[0090] In the formula, is the estimated value; is the true value; N is the total number of state estimation samples;

[0091] 2) The formula for the ratio of measurement estimation errors:

[0092]

[0093] Among them, S H is the measurement error statistic, and S M is the estimation error statistic.

[0094] Beneficial effects: Compared with the prior art, the technical solution of the present invention has the following beneficial technical effects:

[0095] Aiming at the problems that the model constructed for the hydrogen - blended gas network is not accurate enough, there are information barriers between subsystems, and there is a lack of measurement equipment and relevant standard specifications, etc., the present invention first proposes a multi - time - scale state estimation method based on the hydrogen - blended electric - gas integrated energy system. This method first refines the hydrogen - blended electric - gas integrated energy system model. Compared with the commonly used steady - state model at present, this model can more accurately describe the operation of each state variable. Considering the different time scales of the power grid and the gas network, a segmented state estimation algorithm based on the unified theory of time series is adopted, which can effectively solve the problem of inconsistent dynamic response time scales of the power grid and the gas network and improve the efficiency of information interaction between subsystems. Finally, based on the adaptive cubature Kalman filter algorithm, the state estimation of the hydrogen - blended electric - gas network is carried out. The adaptive cubature Kalman filter algorithm has high robustness and can provide a higher - precision state prediction result for the currently immature hydrogen - blended electric - gas integrated energy system. The algorithm proposed by the present invention can obtain more accurate state estimation results of the hydrogen - blended electric - gas integrated energy system, providing technical support for the subsequent consideration of the safe and economic operation of the hydrogen - blended integrated energy system. Description of the Drawings

[0096] Figure 1 It is a segmented state estimation structure diagram;

[0097] Figure 2 It is a diagram of 20 gas network nodes - 24 power grid nodes in Belgium;

[0098] Figure 3 It is a comparison diagram of voltage magnitude state estimation results;

[0099] Figure 4 It is a comparison diagram of voltage phase angle state estimation results;

[0100] Figure 5 It is a comparison diagram of gas network node pressure state estimation results;

[0101] Figure 6 It is a comparison diagram of hydrogen mass flow state estimation results. Specific implementation manners

[0102] The technical solution of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0103] The present invention proposes a multi-time scale state estimation method based on a hydrogen-blended electric-gas integrated energy system, including the following steps:

[0104] Step 1: Obtain the power grid and hydrogen-blended gas network topological structures, coupling information, and boundary conditions of the hydrogen-blended electric-gas integrated energy system, and construct the overall architecture of the hydrogen-blended electric-gas integrated energy system;

[0105] Step 2: Establish a dynamic state estimation model for the hydrogen-blended electric-gas integrated energy system based on the system architecture, including a dynamic state estimation model for the power system based on the two-exponential smoothing method, a dynamic state estimation model for the hydrogen-blended gas network based on the implicit finite difference method, a hydrogen advection-diffusion model, a coupling device model established considering the connection between the power grid and the hydrogen-blended gas network, and an energy balance equation;

[0106] Step 3: Set the power system state estimation time as Δt and the hydrogen-blended gas network state estimation time as nΔt. When the running time is a multiple of Δt, only the power grid state estimation is performed alone, that is, the dynamic state estimation model for the power system constructed in Step 2 is used, and the dynamic state estimation results of the power system are solved by the cubature Kalman filter algorithm; when the running time is a multiple of nΔt, both the power grid and the hydrogen-blended gas network perform state estimation, that is, the dynamic state estimation model for the power system, the dynamic state estimation model for the hydrogen-blended gas network, and the cubature Kalman wave algorithm in Step 2 are used to solve the dynamic state estimation results of the power grid and the hydrogen-blended gas network, and the boundary information between the power grid and the hydrogen-blended gas network is exchanged considering the boundary coupling relationship and the energy balance equation to realize the iterative update of the state information;

[0107] Step 4: Obtain the dynamic state estimation results of the segmented hydrogen - blended electric - gas integrated energy system by adopting different state - estimation time scales for different subsystems.

[0108] Furthermore, the obtaining of the topological structures, coupling information, and boundary conditions of the power grid and the hydrogen - blended gas network of the hydrogen - blended electric - gas integrated energy system, and the construction of the overall architecture of the hydrogen - blended electric - gas integrated energy system in Step 1 are specifically as follows:

[0109] Adopt the Belgian 24 - node power network and 20 - node hydrogen - blended gas network as the basic systems. The P2H devices are installed at nodes 2, 3, 5, 6, 7, 8, and 9 in the power grid. Hydrogen is introduced into the natural gas network at nodes 2, 4, 7, and 10. The gas turbines at nodes 2, 6, 8, 15, 17, and 22 in the power grid are connected to the natural gas network at nodes 4, 6, 11, and 13. The wind turbines are connected to nodes 1 to 4 of the power system, and the photovoltaic units are connected to nodes 5 to 10. The measurement values of the state estimation are obtained by adding Gaussian noise with a mean of zero and a variance of the standard deviation to the power flow calculation. The initial standard deviations of the voltage amplitude, voltage phase angle, active power, and reactive power are set to 0.005, 0.002, 0.02, and 0.02 respectively; the initial standard deviation of the measurement error of the hydrogen - blended gas network is set to 0.01.

[0110] Furthermore, the establishment of the dynamic state estimation model of the hydrogen - blended electric - gas integrated energy system in Step 2 includes the dynamic state estimation model of the power system based on the two - exponential smoothing method, the dynamic state estimation model of the hydrogen - blended gas network based on the implicit finite - difference method, the hydrogen advection - diffusion model, the coupling device model established considering the connection between the power grid and the hydrogen - blended gas network, and the energy balance equation. The process is as follows:

[0111] (2.1) Power system dynamic state estimation model

[0112] The power system usually adopts Holt's two - parameter exponential smoothing method to construct the state - space equation, re - distribute the weights of the estimated value and the predicted value at time k - 1 to obtain the state and predicted value at time k. The specific formula is as follows:

[0113]

[0114] In the formula, is the predicted value of the state quantity at time k; α k-1 is the horizontal component at time k - 1; b k-1 is the vertical component at time k - 1; α H and β H are smoothing parameters, and their value ranges are in [0, 1];

[0115] According to the power flow calculation formula of the power system, the measurement equation of the power grid state estimation can be obtained as follows:

[0116]

[0117] In the formula, P i , Q i are the active power and reactive power of node i respectively; V i , V j are the voltage amplitude of node i and the voltage amplitude of node j respectively; θ i , θ j are the voltage phase of node i and the voltage phase of node j respectively; G ij , B ij are the conductance between lines ij and the susceptance between lines ij respectively; j ∈ i represents all the lines connected to node i and is expressed as line ij;

[0118] (2.2) Dynamic State Estimation Model of Hydrogen - Mixed Gas Network

[0119] The natural gas hydrogen - mixing technology relies on the operation of existing natural gas pipelines. To construct the state estimation model of the hydrogen - mixed gas network, it is necessary to first construct the natural gas dynamic state estimation model. Considering the flow of natural gas as one - dimensional fluid, the following assumptions are made according to the operating characteristics of natural gas: ① The radial change of natural gas is ignored, and only the axial flow is considered; ② The operating speed and temperature of the natural gas flow remain unchanged, and the flow velocity is less than the speed of sound; ③ The pipeline is placed horizontally, and the influence brought by the pipeline angle is ignored;

[0120] According to the law of conservation of mass and Newton's second law of motion, a natural gas dynamic state estimation model is constructed. Considering the influence on pipeline pressure drop and ignoring the square term of natural gas flow velocity, the following equations are obtained:

[0121]

[0122] In the formula, ρ is the density of natural gas; t is time; u is the flow velocity of natural gas; x is the pipeline distance; p is the pressure of each node in the natural gas network; λ is the hydraulic friction coefficient; d is the pipeline diameter;

[0123] The gas in the pipeline is a mixture of hydrogen and natural gas. Considering the relationship:

[0124] M = ρuA

[0125] and

[0126] p = ω 2 ρ = zRTρ

[0127] In the formula, M is the mass flow rate of the mixed gas; A is the cross - sectional area of the pipeline; ω is the speed of sound propagation in the gas, z is the compression factor of the mixed gas; R is the gas constant; T is the temperature;

[0128] Substitute ρu and ρ in the natural gas dynamic state estimation model with the above two equations to obtain the following formula:

[0129]

[0130] Hydrogen is injected into the natural gas pipeline, changing the gas operation state and physical property parameters inside the pipeline, that is, changing the density, compression factor and hydraulic friction coefficient of the mixed gas. According to the characteristics of hydrogen - mixed natural gas, the parameter values of the mixed gas are calculated. First, the compression factor z is calculated mix :

[0131]

[0132] In the formula, T mix , T 0 are the critical temperature of the mixed gas and the temperature under standard conditions respectively; p mix are the average pressures at both ends of the pipeline and the critical pressure of the mixed gas respectively; p i , p j are the pressures at node i and node j respectively; T g , T H2 are the critical temperature of natural gas and the critical temperature of hydrogen respectively; p g , p H2 are the critical pressure of natural gas and the critical pressure of hydrogen respectively; θ H2 is the volume ratio of hydrogen in the mixed gas;

[0133] Calculate the hydraulic friction coefficient λ mix :

[0134]

[0135] Among them, the value of the Reynolds number Re is:

[0136]

[0137] In the formula, K is the absolute roughness of the inner wall of natural gas, taking 0.0508 mm; Q ij is the volume flow rate of the mixed gas between node i and node j; ρ mix is the standard - condition density of the mixed gas; μ mix is the viscosity of the mixed gas;

[0138] Among them, the calculation formula for the density ρ of the mixed gas mix is:

[0139]

[0140] The Wilke semi - empirical relationship is used to calculate the viscosity μ of the mixed gas mix :

[0141]

[0142] In the formula, ρ H2 and ρ g are the hydrogen density and the natural gas density respectively; μ H2 and μ g are the hydrogen viscosity and the natural gas viscosity respectively; M H2 and M g are the hydrogen molar mass and the natural gas molar mass respectively;

[0143] Substitute the compressibility factor and the hydraulic friction coefficient calculated after considering hydrogen injection into the natural gas dynamic state estimation model, and use the implicit finite difference method to process this model. The finally discretized dynamic state estimation model of the hydrogen-mixed gas network is as follows:

[0144]

[0145]

[0146] In the formula, Δt is the unit time; Δx is the unit distance; u ij is the average flow velocity of the mixed gas between pipelines i and j; p j,t+1 is the pressure at node j at time t + 1; p i,t+1 is the pressure at node i at time t + 1; p j,t is the pressure at node j at time t; p i,t is the pressure at node i at time t; M j,t+1 is the mass flow rate of the mixed gas at node j at time t + 1; M i,t+1 is the mass flow rate of the mixed gas at node i at time t + 1; M j,t is the mass flow rate of the mixed gas at node j at time t; M i,t is the mass flow rate of the mixed gas at node i at time t;

[0147] (2.3) Hydrogen advection-diffusion model

[0148] To characterize the relationship between the hydrogen volume fraction and the flow velocity after hydrogen is injected into the natural gas pipeline, and the gas quality differences of each pipeline and node, consider the advection-diffusion equation and construct a partial differential equation for the transmission process of hydrogen in the mixed gas due to advection and diffusion.

[0149] The advection-diffusion equation considering only one-dimensional gas flow is:

[0150]

[0151] In the formula, C is the substance concentration; t is the time; u is the gas flow velocity; D is the diffusion coefficient;

[0152] When it is assumed that the gas flow velocity between pipelines ij is the average flow velocity u ij the above expression is simplified to:

[0153]

[0154] In the formula, x is the pipeline distance, and u ij has the following calculation formula:

[0155]

[0156] In the formula, M ij is the mass flow rate of the mixed gas between nodes i and j; ρ ij is the density of the mixed gas between nodes i and j; A ij is the cross-sectional area of the pipeline between nodes i and j;

[0157] Introduce the auxiliary variable m H2 , and use the hydrogen mass flow rate of the mixed gas to replace the concentration C. The equation is transformed into:

[0158]

[0159] The calculation formula of the hydrogen mass flow rate is as follows:

[0160]

[0161] Using the implicit finite difference method to process the hydrogen advection-diffusion model, the final model is as follows:

[0162]

[0163] In the formula, is the hydrogen mass flow rate at node j at time t + 1; is the hydrogen mass flow rate at node i at time t + 1; is the hydrogen mass flow rate at node j at time t; is the hydrogen mass flow rate at node i at time t;

[0164] (2.4) Coupling equipment model

[0165] The relationship between the electric energy consumed by the P2H device and the converted gas energy is as follows:

[0166] E i,t = η P2G P k,t

[0167]

[0168] In the formula, η k,i is the energy conversion efficiency of the P2G device; P k,t is the electric power consumed by the P2G device at node k of the power system at time t; E k,tThe hydrogen energy of the P2G device transferred to the hydrogen - blended natural gas network node i at time t; F k,t The hydrogen flow rate of the P2G device transferred to the hydrogen - blended gas network node i at time t; H H2 The calorific value of hydrogen;

[0169] (2.5) Energy balance equation

[0170] Considering that hydrogen injection will affect the operating characteristics of the gas network, an energy balance equation is constructed with energy flow as the variable instead of the volume flow used in the pure gas network system. The equation is constructed based on the fact that the inflow energy at each node is equal to the outflow energy, as follows:

[0171]

[0172] In the formula, G w (m) represents the set of natural gas source points connected to node m; G h (m) represents the set of power - to - hydrogen devices connected to node m; G e (m) represents the set of gas loads connected to node m; G g (m) represents the set of gas turbine units connected to node m; G(m) represents the set of natural gas pipelines connected to node m; G k (m) represents the set of compressors connected to node m; The percentage of the gas flow consumed by compressor k in the transported flow; The energy provided by natural gas source w at time t; The energy provided by the power - to - hydrogen device h at time t; The actual energy consumed by the natural gas load e at time t; The gas energy consumed by the gas turbine unit v at time t; E mn,t The gas energy in pipeline mn at time t; The gas energy passing through compressor k at time t;

[0173] There is an equation relationship between the energy flow E and the volume flow F, that is, E = FH, where H is the calorific value of the gas. Assuming that the gas flowing out of the node is completely mixed, the energy conservation equation is obeyed before and after hydrogen blending at each node:

[0174]

[0175] In the formula, F mn,t The gas volume flow rate in pipeline mn at time t; The calorific value of the mixed gas in pipeline mn at time t; The volume flow rate provided by the power - to - hydrogen device h at time t; The calorific value of hydrogen; The volumetric flow rate provided by the natural gas source w at time t; H gas The calorific value of natural gas; H m,t The mixed calorific value at node m at time t.

[0176] Furthermore, in step 3, the state estimation time of the power system is set as Δt, and the state estimation time of the hydrogen - mixed gas network is set as nΔt. When the running time is a multiple of Δt, only the power grid state estimation is performed separately, that is, the dynamic state estimation model of the power system constructed in step 2 is adopted, and the dynamic state estimation result of the power system is solved through the cubature Kalman filter algorithm; when the running time is a multiple of nΔt, then state estimations are performed for both the power grid and the hydrogen - mixed gas network, that is, the dynamic state estimation models of the power system and the hydrogen - mixed gas network in step 2 and the cubature Kalman wave algorithm are used to solve the dynamic state estimation results of the power grid and the hydrogen - mixed gas network, and the boundary information exchange between the power grid and the hydrogen - mixed gas network is carried out considering the boundary coupling relationship and the energy balance equation to realize the iterative update of the state information. The specific method is as follows:

[0177] (3.1) Multi - time - scale state estimation

[0178] The dynamic state estimations are respectively carried out for the two sub - networks by setting different time scales. As Figure 1 shown, the time scales for the dynamic state updates of the power grid and the hydrogen - mixed gas network are set as Δt and nΔt. If only the time scale of the power grid state estimation is satisfied at this moment, the boundary state of the power grid is directly determined according to the gas network state and the boundary coupling constraint conditions, and the dynamic state estimation of the power grid is carried out separately; if the time scales of both the power grid and the gas network state estimations are satisfied at this moment, then the dynamic state estimations of the power grid and the hydrogen - mixed gas network are carried out simultaneously, and information is exchanged through the coupling equipment and boundary conditions;

[0179] (3.2) Adaptive cubature Kalman filter

[0180] For the state estimations of the power grid and the hydrogen - mixed gas network at each time scale, the adaptive cubature Kalman filter algorithm is adopted. The steps of this algorithm are as follows:

[0181] 1) Prediction step: Estimate the state quantity and covariance matrix at the current moment through the previously obtained dynamic state estimation models of the power grid and the hydrogen - mixed gas network. Among them, the state quantity of the power grid dynamic state estimation is x e = [V, θ], where V is the voltage of the power grid node and θ is the phase angle of the power grid node; the state quantity of the hydrogen - mixed gas network is p is the pressure of the gas network node, is the hydrogen flow rate of the gas network node;

[0182] 2) Update step: Update the covariance matrix, gain, and state estimation result according to the measurement and the dynamic state estimation model. Among them, the power grid measurement is y e=[P, Q, V, θ], where P is the active power of the grid node and Q is the reactive power of the grid node; the measurement of the hydrogen - mixed gas network volume is M is the flow rate of the mixed gas in the gas network;

[0183] 3) Adaptation: By introducing the fading - memory exponentially weighted method, the noise covariance matrix is updated, and finally the state - estimation result of the segmented hydrogen - mixed electric - gas integrated energy system is obtained.

[0184] Furthermore, the dynamic state - estimation result of the segmented hydrogen - mixed electric - gas integrated energy system obtained by adopting different state - estimation time scales for different subsystems in step 4 is as follows:

[0185] Calculate the root - mean - square error and the measurement - estimation error respectively. If the sum of the root - mean - square error and the measurement - estimation error is less than the preset value, the state estimation meets the requirements;

[0186] 1) Introduce the root - mean - square error formula:

[0187]

[0188] In the formula, is the estimated value; is the true value; N is the total number of state - estimation samples;

[0189] 2) The formula for the ratio of the measurement - estimation error:

[0190]

[0191] Among them, S H is the measurement - error statistical value, and S M is the estimation - error statistical value.

[0192] Case study

[0193] 1) Case description

[0194] The test system of the present invention is based on the Belgian 24 - node hydrogen - mixed gas network - 20 - node power system network, as Figure 2As shown in the figure, P2H devices are installed at nodes 2, 3, 5, 6, 7, 8, and 9 in the power grid. Hydrogen is introduced into the Belgian gas network at nodes 2, 3, 4, 7, and 10. The gas turbines at grid nodes 2, 6, 8, 15, 17, and 22 are connected to the gas pipeline network at nodes 4, 6, 11, and 13. Wind turbines are connected to nodes 1 - 4 of the power system, and photovoltaic units are connected to nodes 5 - 10. The measurements for state estimation are obtained by adding Gaussian noise to the power flow calculation values. The initial standard deviations of the voltage amplitude, voltage phase angle, real power, and active power measurement errors are set to 0.005, 0.002, 0.02, and 0.02 respectively. The initial standard deviation of the measurement error for the hydrogen - mixed gas network is set to 0.01. The test cases are executed using the MATLAB 2020b platform, and the test environment includes a 2.60 GHz processor and 16.0 GB of RAM.

[0195] 2) Comparison of the algorithm of the present invention with other methods

[0196] The present invention mainly evaluates the effectiveness of the proposed model method through the following comparisons: 1) Using the WLS algorithm for state estimation of the hydrogen - mixed electricity - gas integrated energy system; 2) Based on the existing steady - state model of the natural gas network, using the ACKF algorithm for state estimation of the hydrogen - mixed electricity - gas integrated energy system; 3) Based on the dynamic model of the hydrogen - mixed gas network proposed in the present invention, using the ACKF algorithm for state estimation of the hydrogen - mixed electricity - gas integrated energy system.

[0197] 3) Analysis of comparison results

[0198] The time interval for state estimation of the hydrogen - mixed gas network is 1 h, and 24 state estimations can be performed in 24 h a day; the time for state estimation of the power grid is 15 min, and a total of 96 estimations are made in a day. Within 1 h, the power grid subsystem conducts 4 estimations, while the hydrogen - mixed gas network subsystem is assumed to be in a steady state during this time period, and each state variable remains unchanged. According to the different time intervals of the power grid and the hydrogen - mixed gas network in the above - mentioned hydrogen - mixed electricity - gas integrated energy system, the state estimation result diagrams can be obtained. Figure 3 It is the state estimation result diagram of the power grid voltage amplitude using the steady - state model - WLS algorithm and the dynamic model - ACKF algorithm respectively. Figure 4 It is the state estimation result diagram of the power grid voltage phase angle using the steady - state model - WLS algorithm and the dynamic model - ACKF algorithm. Figure 5 It is the state estimation result diagram of the node pressure of the hydrogen - mixed gas network using the steady - state model - WLS algorithm, the steady - state model - ACKF algorithm, and the dynamic model - ACKF algorithm respectively. Figure 6 It is the state estimation result diagram of the hydrogen mass flow rate of the hydrogen - mixed gas network using the dynamic model - ACKF algorithm, where Figure 3 、 Figure 4 are the state estimation results of grid node 5 within a day. Figure 5The state estimation result of the hydrogen mixing network at node 10 at 12:00 Figure 6 They are the state estimation results of the hydrogen mixing network at node 6 at 5:00, 6:00, and 7:00. It can be observed from the above four figures that the dynamic model has better estimation performance than the steady-state model, and the Adaptive Cubature Kalman Filter (ACKF) has higher estimation accuracy than the Weighted Least Squares (WLS). Therefore, the dynamic model-ACKF algorithm proposed in this project has better estimation effect.

[0199] 4) Introduce index discrimination

[0200] The root mean square error is introduced to evaluate the state estimation accuracy.

[0201]

[0202] In the formula, is the estimated value; is the true value.

[0203] The ratio of measurement estimation error is introduced to evaluate the performance of state estimation. The lower the error ratio, the stronger the ability of state estimation to maintain accuracy.

[0204] Calculation formula for the statistical value of measurement error::

[0205]

[0206] Calculation formula for the statistical value of estimation error

[0207]

[0208] Introduce the calculation formula for the ratio of measurement estimation error

[0209]

[0210] In the formula, x ture represents the true value of the state variable, x i,t represents the estimated value of the state variable, z i,t represents the measured variable, and σ i,t represents the noise standard deviation of the measured variable.

[0211] Table 1 shows the comparison of the root mean square errors of the state estimation results of the hydrogen-blended electric-gas integrated energy system calculated by three model algorithms: the steady-state model-WLS algorithm, the steady-state model-ACKF algorithm, and the dynamic model-ACKF algorithm. Table 2 shows the comparison of the ratios of the measurement estimation errors of the state estimation results of the hydrogen-blended electric-gas integrated energy system calculated by the above three model algorithms. The smaller the root mean square error, the higher the calculation accuracy of the state estimation. The smaller the ratio of the measurement estimation error, the higher the calculation reliability of the state estimation. It can be seen from the two tables that by using the dynamic model-ACKF algorithm proposed in this project for the state estimation of the hydrogen-blended electric-gas integrated energy system, more accurate and reliable state estimation results can be obtained.

[0212] Table 1 is the comparison of the root mean square errors in three cases

[0213]

[0214] Table 2 is the comparison of the ratios of the measurement estimation errors in three cases;

[0215]

Claims

1. A multi-time scale state estimation method based on a hybrid hydrogen-electricity-gas integrated energy system, characterized in that: The following steps are involved: Step 1: Obtain the power grid and hybrid hydrogen-gas network topology, coupling information and boundary conditions of the hybrid hydrogen-electricity-gas integrated energy system, and construct the overall architecture of the hybrid hydrogen-electricity-gas integrated energy system; Step 2: Establish a dynamic state estimation model of the hybrid hydrogen-electric-gas integrated energy system based on the system architecture, including a dynamic state estimation model of the power system based on the two-exponential smoothing method, a dynamic state estimation model of the hybrid hydrogen-gas network based on the implicit finite difference method, a hydrogen advection-diffusion model, a coupling device model established considering the connection between the power grid and the hybrid hydrogen-gas network, and an energy balance equation; Step 3, assuming that the state estimation time of the power system is Δt, and the state estimation time of the mixed hydrogen network is nΔt. When the operation time is a multiple of Δt, only the power grid state estimation is performed separately, that is, the power system dynamic state estimation model constructed in step 2 is used, and the dynamic state estimation result of the power system is solved by the volumetric Kalman filter algorithm; when the operation time is a multiple of nΔt, the power grid and the mixed hydrogen network both perform state estimation, that is, the power system dynamic state estimation model, the mixed hydrogen network dynamic state estimation model, and the volumetric Kalman wave algorithm in step 2 are used to solve the dynamic state estimation results of the power grid and the mixed hydrogen network, and the boundary coupling relationship and the energy balance equation are considered to exchange boundary information between the power grid and the mixed hydrogen network to realize iterative update of state information; Step 4: By adopting different state estimation time scales for different subsystems, the dynamic state estimation results of the segmented hybrid hydrogen-electricity-gas integrated energy system are obtained.

2. According to claim 1, a multi-time scale state estimation method based on a hybrid hydrogen-electricity-gas integrated energy system is characterized in that: The step 1 is to obtain the power grid and the hybrid hydrogen-gas network topology, coupling information and boundary conditions of the hybrid hydrogen-electricity-gas integrated energy system, and construct the overall architecture of the hybrid hydrogen-electricity-gas integrated energy system, as follows: The Belgian 24-node power grid and 20-node mixed hydrogen network are used as the basic system. P2H equipment is installed at nodes 2, 3, 5, 6, 7, 8 and 9 in the grid. Hydrogen is introduced into the natural gas network at nodes 2, 4, 7 and 10. The gas turbines at nodes 2, 6, 8, 15, 17 and 22 of the grid are connected to the natural gas network at nodes 4, 6, 11 and 13. Wind turbines are connected to nodes 1 to 4 of the power system, and photovoltaic units are connected to nodes 5 to 10. The measured values ​​of state estimation are obtained by flow calculation plus Gaussian noise with zero mean and standard deviation; the initial standard deviations of voltage amplitude, voltage phase angle, active power and reactive power are set to 0.005, 0.002, 0.02 and 0.02 respectively; the initial standard deviation of the measurement error of the mixed hydrogen network is set to 0.

01.

3. A multi-time scale state estimation method based on a hybrid hydrogen-electricity-gas integrated energy system according to claim 2, characterized in that: The dynamic state estimation model of the hybrid hydrogen-electric-gas integrated energy system is established based on the system architecture as described in step 2, including the dynamic state estimation model of the power system based on the two-exponential smoothing method, the dynamic state estimation model of the hybrid hydrogen-gas network based on the implicit finite difference method, the hydrogen advection-diffusion model, the coupling device model established considering the connection between the power grid and the hybrid hydrogen-gas network, and the energy balance equation process are as follows: (2.1) Power system dynamic state estimation model The power system usually uses Holt's two-parameter exponential smoothing method to construct the state space equation, redistribute the weights of the estimated value and predicted value at time k-1, and obtain the state and predicted value at time k. The specific formula is as follows: In the formula, is the predicted value of the state quantity at time k; α k-1 is the horizontal component at time k-1; b k-1 is the vertical component at time k-1; α H and β H is a smoothing parameter, with a value range of [0,1]; According to the power system flow calculation formula, the grid state estimation measurement equation can be obtained as follows: Where P i , Q i are the active power and reactive power of node i respectively; V i 、V j are the voltage amplitude of node i and node j respectively; θ i ,θ j are the voltage phase of node i and node j respectively; G ij , B ij are the conductance between lines ij and the susceptance between lines ij respectively; j∈i is all the lines connected to node i, which are denoted as lines ij; (2.2) Dynamic state estimation model of hydrogen mixing network The natural gas hydrogen mixing technology relies on the existing natural gas pipeline operation. To build the state estimation model of the hydrogen mixing network, it is necessary to first build a natural gas dynamic state estimation model, consider the flow of natural gas as a one-dimensional fluid, and make the following assumptions based on the operation characteristics of natural gas: ① The radial change of natural gas is negligible, and it is only considered to flow along the axial direction; ② The running speed and temperature of the natural gas flow remain unchanged, and the flow velocity is less than the speed of sound; ③ The pipeline is placed horizontally, and the influence of the pipeline angle is ignored; The natural gas dynamic state estimation model is constructed based on the law of conservation of mass and Newton's second law of motion. Considering the impact on pipeline pressure drop and ignoring the square term of natural gas flow rate, the following equations are obtained: Where, ρ is the natural gas density; t is the time; u is the natural gas flow rate; x is the pipeline distance; p is the pressure at each node of the natural gas network; λ is the hydraulic friction coefficient; d is the pipeline diameter; The pipeline gas is a mixture of hydrogen and natural gas. Consider the relationship: M=ρuA and p=ω 2 ρ=zRTρ Where, M is the mass flow rate of the mixed gas; A is the cross-sectional area of ​​the pipe; ω is the speed of sound propagation in the gas, z is the compressibility factor of the mixed gas; R is the gas constant; T is the temperature; Substituting ρu and ρ in the natural gas dynamic state estimation model into the above two equations, we can get the following equation: Hydrogen is injected into the natural gas pipeline to change the gas operation state and physical parameters in the pipeline, that is, to change the density, compression factor and hydraulic friction coefficient of the mixed gas. The values ​​of various parameters of the mixed gas are calculated according to the natural gas hydrogen mixing characteristics. First, the compression factor z is calculated. mix : Where, T mix , T0 are the critical temperature of the mixed gas and the temperature under standard conditions respectively; p ij 、p mix are the average pressure at both ends of the pipeline and the critical pressure of the mixed gas respectively; p i 、p j are the pressure at node i and node j respectively; T g , T H2 are the critical temperatures of natural gas and hydrogen respectively; p g 、p H2 are the critical pressures of natural gas and hydrogen respectively; θ H2 is the volume ratio of hydrogen in the mixed gas; Calculate the hydraulic friction coefficient λ mix : Among them, the value of Reynolds number Re is: Where, K is the absolute roughness of the inner wall of natural gas, which is 0.0508 mm; Q ij is the volume flow rate of the mixed gas between nodes i and j; ρ mix is the standard density of the mixed gas; μ mix is the viscosity of the mixed gas; Where, the density of the mixed gas ρ mix The calculation formula is: pmix=p H2 i H2 +ρg(1-θ H2 ) Calculation of mixed gas viscosity μ using Wilke semi-empirical relationship mix : In the formula, ρ H2 , g are the density of hydrogen and natural gas respectively; μ H2 , μ g are the viscosity of hydrogen and natural gas respectively; M H2 、M g are the molar mass of hydrogen and the molar mass of natural gas, respectively; The above-mentioned compression factor and hydraulic friction coefficient after considering hydrogen injection and calculation are brought into the natural gas dynamic state estimation model, and the implicit finite difference method is used to process the model. The final discretized dynamic state estimation model of the mixed hydrogen network is as follows: In the formula, Δt is the unit time; Δx is the unit distance; u ij is the average flow rate of the mixed gas between pipes i and j; p j,t+1 is the pressure of node j at time t+1; p i,t+1 is the pressure of node i at time t+1; p j,t is the pressure of node j at time t; p i,t is the pressure of node i at time t; M j,t+1 is the mass flow rate of the mixed gas at node j at time t+1; M i,t+1 is the mixed gas mass flow rate at node i at time t+1; M j,t is the mass flow rate of the mixed gas at node j at time t; M i,t is the mass flow rate of the mixed gas at node i at time t; (2.3) Hydrogen advection-diffusion model In order to characterize the relationship between hydrogen volume fraction and flow rate after hydrogen is injected into the natural gas pipeline, as well as the quality differences of each pipeline and node, the advection-diffusion equation is considered to construct the partial differential equation of the transmission process of hydrogen in the mixed gas due to advection and diffusion. Considering only the one-dimensional flow of gas, the advection-diffusion equation is: In the formula, C is the substance concentration; t is the time; u is the gas flow rate; D is the diffusion coefficient; When assuming that the gas flow rate between pipes ij is the average flow rate u ij When , the above expression is simplified to: Where x is the pipe distance, and u ij The calculation formula is as follows: Where M ij is the mass flow rate of the mixed gas between nodes i and j; ρ ij is the mixed gas density between nodes i and j; A ij is the cross-sectional area of ​​the pipe between nodes i and j; Introduce auxiliary variable m H2 , using the hydrogen mass flow rate of the mixed gas instead of the concentration C, the equation is transformed into: The calculation formula for hydrogen mass flow rate is as follows: The final model of the hydrogen advection-diffusion model is obtained by using the implicit finite difference method as follows: In the formula, is the hydrogen mass flow rate at node j at time t+1; is the hydrogen mass flow rate of node i at time t+1; is the hydrogen mass flow rate at node j at time t; is the hydrogen mass flow rate at node i at time t; (2.4) Coupling device model The relationship between the electrical energy consumed by the P2H equipment and the gas energy converted is as follows: E i,t =the P2G P k,t Where η k,i is the energy conversion efficiency of P2G equipment; P k,t is the electric power consumed by the P2G device at the k node of the power system at time t; E k,t F is the hydrogen energy converted from the P2G device to the hydrogen-mixed natural gas network node i at time t; k,t is the hydrogen flow rate converted from the P2G device to the mixed hydrogen network node i at time t; H H2 is the calorific value of hydrogen; (2.5) Energy balance equation Considering that hydrogen injection will affect the operating characteristics of the gas network, the energy balance equation is constructed with energy flow as the variable, rather than the volume flow used in the pure gas network system. The equation is constructed based on the inflow energy of each node being equal to the outflow energy, as follows: In the formula, G w (m) represents the set of natural gas source points connected to node m; G h (m) represents the set of power-to-hydrogen devices connected to node m; G e (m) represents the gas load set connected to node m; G g (m) represents the set of gas turbine units connected to node m; G(m) represents the set of natural gas pipelines connected to node m; G k (m) represents the set of compressors connected to node m; The percentage of gas flow consumed by compressor k to the delivery flow; is the energy provided by the natural gas source w at time t; The energy provided by the power-to-hydrogen device h at time t; is the actual energy consumed by the natural gas load e at time t; is the gas energy consumed by the gas turbine unit v at time t; E mn,t is the gas energy in pipe mn at time t; is the gas energy passing through compressor k at time t; There is an equation relationship between the energy flow E and the volume flow F, that is, E = FH, where H is the calorific value of the gas. Assuming that the gas flowing out of the node is completely mixed, the energy conservation equation before and after hydrogen mixing at each node is obtained: In the formula, F mn,t is the gas volume flow rate in pipeline mn at time t; is the calorific value of the mixed gas in pipeline mn at time t; The volume flow rate provided by the power-to-hydrogen device h at time t; is the calorific value of hydrogen; is the volume flow rate provided by the natural gas source w at time t; H gas is the calorific value of natural gas; H m,t is the mixed heat value of node m at time t.

4. A multi-time scale state estimation method based on a hybrid hydrogen-electricity-gas integrated energy system according to claim 3, characterized in that: In step 3, the state estimation time of the power system is assumed to be Δt, and the state estimation time of the mixed hydrogen network is assumed to be nΔt. When the operation time is a multiple of Δt, only the power grid state estimation is performed separately, that is, the power system dynamic state estimation model constructed in step 2 is adopted, and the dynamic state estimation result of the power system is solved by the volumetric Kalman filter algorithm; when the operation time is a multiple of nΔt, the power grid and the mixed hydrogen network both perform state estimation, that is, the power system dynamic state estimation model, the mixed hydrogen network dynamic state estimation model, and the volumetric Kalman wave algorithm in step 2 are adopted to solve the dynamic state estimation results of the power grid and the mixed hydrogen network, and the boundary coupling relationship and the energy balance equation are considered to exchange boundary information between the power grid and the mixed hydrogen network to realize iterative update of state information. The specific method is as follows: (3.1) Multi-time scale state estimation Different time scales are set for the two sub-networks to perform dynamic state estimation. The time scales for the dynamic state update of the power grid and the mixed hydrogen gas network are assumed to be Δt and nΔt. If only the power grid state estimation time scale is met at this moment, the boundary state of the power grid is directly determined according to the gas grid state and the boundary coupling constraint conditions, and the dynamic state estimation of the power grid is performed separately; if the time scales for the state estimation of both the power grid and the gas grid are met at this moment, the dynamic state estimation of the power grid and the mixed hydrogen gas network is performed simultaneously, and information is exchanged through coupling devices and boundary conditions; (3.2) Adaptive cubature Kalman filter The state estimation of the power grid and the mixed hydrogen network at each time scale uses the adaptive volumetric Kalman filter algorithm. The steps of the algorithm are as follows: 1) Prediction step: The state quantity and covariance matrix of the current moment are estimated by using the previously obtained dynamic state estimation model of the power grid and the mixed hydrogen network, where the state quantity of the power grid dynamic state estimation is x e =[V,θ], V is the grid node voltage, θ is the grid node phase angle; the state quantity of the mixed hydrogen network is x g =[p,m H2 ], p is the gas network node pressure, is the hydrogen flow rate at the gas grid node; 2) Update step: Update the covariance matrix, gain and state estimation results according to the quantity measurement and dynamic state estimation model, where the power grid quantity measurement is y e =[P,Q,V,θ], P is the active power of the grid node, Q is the reactive power of the grid node; the mixed hydrogen network quantity is measured as y g =[M,p,m H2 ], M is the mixed gas flow rate of the gas network; 3) Adaptation: By introducing the gradually fading memory exponential weighting method, the noise covariance matrix is ​​updated, and finally the state estimation result of the segmented hybrid hydrogen-electric-gas integrated energy system is obtained.

5. A multi-time scale state estimation method based on a hybrid hydrogen-electricity-gas integrated energy system according to claim 4, characterized in that: In step 4, by using different state estimation time scales for different subsystems, the dynamic state estimation results of the segmented hybrid hydrogen-electricity-gas integrated energy system are obtained, as follows: Calculate the root mean square error and the measurement estimation error respectively. If the sum of the root mean square error and the measurement estimation error is less than the preset value, the state estimation meets the requirements. 1) Introduce the root mean square error formula: In the formula, is an estimated value; is the true value; N is the total number of state estimation samples; 2) Measurement estimation error ratio formula: Among them, S H is the measurement error statistic, S M is the estimated error statistic.

Images