A photovoltaic power generation hydrogen production cluster random optimization scheduling method based on It? process

Abstract

The application discloses a photovoltaic power generation hydrogen production cluster random optimization scheduling method based on an It process, and first establishes a photovoltaic power generation hydrogen production system model, including an It process model of photovoltaic output, an electrolytic cell cluster scheduling model and a hydrogen storage tank model; then taking the highest expected economic benefit and the highest energy utilization rate under the influence of photovoltaic output prediction error uncertainty as an objective function, designing a control law of the system based on the photovoltaic power generation hydrogen production system model and using an affine strategy, and establishing a random optimization scheduling model of the electrolytic cell cluster; finally, based on trajectory sensitivity decomposition, the random optimization control model is transformed into a deterministic optimization problem, and a model predictive control form is used for rolling optimization solution. The application can dynamically adjust the running state and power distribution of each electrolytic cell, effectively realizes efficient hydrogen production by electricity under the condition of photovoltaic output uncertainty, and fully absorbs fluctuating photovoltaic output.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of electrolytic cell cluster scheduling, in particular to a photovoltaic power generation hydrogen production cluster random optimization scheduling method based on an It process. BACKGROUND

[0002] Photovoltaic power generation hydrogen production is one of important technical routes for constructing a clean and low-carbon society in China. Due to the limitation of single machine capacity, a cluster composed of several to dozens of large hydrogen production devices is needed to meet the hydrogen production demand. In addition, the uncertainty of renewable energy output will further affect the production efficiency and reliability of electrolytic hydrogen production. Therefore, it is of great practical significance to study the random optimization scheduling method of electrolytic cell cluster considering the uncertainty of photovoltaic output.

[0003] In the existing research on electrolytic cell cluster scheduling, Shen Xiaojun et al. proposed an off-grid wind power hydrogen production alkaline electrolytic cell array optimization control strategy based on the thermal characteristics, regulation characteristics and other constraints of electrolytic cells, and proposed a coordinated control strategy for electrolytic cell array rotation in a wind power hydrogen production system. Qiu Y et al. proposed a variable load control method for hydrogen production cluster to consume wind and solar power in the paper "Extending the load flexibility of industrial P2H plants: process constraint-aware scheduling method", which takes into account the temperature of hydrogen production machine and the accumulation effect of hydrogen impurities in oxygen. Li Y et al. proposed a wind power hydrogen production system electrolytic cell cycle rotation strategy to balance the working time of each electrolytic cell in the paper "Configuration and operation law of large-scale alkaline water hydrogen production system multi-electrolytic cell hybrid system". Niu Meng et al. proposed a modular hydrogen production control strategy to alleviate the impact of renewable energy on hydrogen energy storage system in the paper "Influence of renewable energy on hydrogen energy storage system and control strategy". Yuan Tiejiang et al. proposed a hydrogen production system day-ahead output planning model considering the start-stop characteristics of electrolytic cells based on the operating state transition relationship of electrolytic cells.

[0004] However, the above-mentioned researches are all based on the deterministic optimization of electricity price or renewable energy output prediction, and there is still a lack of research on the random optimization scheduling of electrolytic cell cluster under the condition of photovoltaic output uncertainty. SUMMARY

[0005] In view of the above problems, the purpose of the present application is to provide an electrolytic cell cluster random optimization scheduling method for a photovoltaic power generation hydrogen production system, which can realize the random optimization scheduling of electrolytic cell cluster under the condition of considering the uncertainty of photovoltaic output, realize high-efficiency hydrogen production, and improve the ability to consume fluctuating photovoltaic power based on the It process modeling of the uncertainty of photovoltaic output.

[0006] The technical scheme is as follows:

[0007] A stochastic optimization scheduling method for photovoltaic power generation hydrogen production clusters based on the Ito process includes the following steps:

[0008] Step 1: Establish a photovoltaic power generation hydrogen production system model, including the Itoh process model of photovoltaic output, the electrolyzer cluster scheduling model, and the hydrogen storage tank model;

[0009] Step 2: Taking the highest expected economic benefits and highest energy utilization rate under the influence of uncertainties in photovoltaic output prediction errors as the objective function, based on the photovoltaic power generation hydrogen production system model and using affine strategy to design the system control law, a stochastic optimization scheduling model of the electrolyzer cluster is established.

[0010] Step 3: Based on trajectory sensitivity decomposition, the stochastic optimization control model is transformed into a deterministic optimization problem, and the model predictive control is used to solve it through rolling optimization.

[0011] Furthermore, the Itoh process model for photovoltaic power output is as follows:

[0012]

[0013] In the formula: P t PV It contributes to the actual photovoltaic power generation; P t PV,pred Forecast value of photovoltaic power output; This represents the prediction error;

[0014] The Itoh process model for prediction error is as follows:

[0015]

[0016]

[0017]

[0018] In the formula: W t Represents the standard Wiener process; For drift term; t is the diffusion term; a is the time constant of cloud shadow change; b is the mean recovery target of the drift term; r t s These represent sunrise and sunset times, respectively; the sine function represents the solar altitude angle; c represents the uncertainty intensity; and d and e represent the range of irradiance variation.

[0019] Furthermore, the electrolytic cell cluster scheduling model includes:

[0020] Logical constraints for the transition of operating status of a single electrolytic cell:

[0021]

[0022]

[0023]

[0024]

[0025] In the formula: n represents the nth electrolytic cell in the electrolytic cell cluster; and These represent the three operating states of the electrolytic cell: production, standby, and shutdown. and These are the start-up and shutdown operations for the electrolytic cell. This indicates the transition of the electrolytic cell from standby to operating state; all variables are represented in binary; the subscripts t-1 and t-2 represent the previous scheduling period and the scheduling period before that when the scheduling period is t, respectively;

[0026] The physical constraints of the electrolytic cell cluster itself:

[0027]

[0028]

[0029]

[0030] Where: N is the total number of electrolytic cells; P n,t P is the power of the nth electrolytic cell at time t; max P min These represent the upper and lower power limits for a single electrolytic cell under production conditions; P SB F represents the constant power of the auxiliary equipment in standby mode for a single electrolytic cell; n,t Let A1, A2, and A3 be the hydrogen production of the nth electrolyzer; A1, A2, and A3 are constant coefficients determined by the characteristics of the electrolyzer.

[0031] Furthermore, the hydrogen storage tank model includes:

[0032] The hydrogen buffer tank storage capacity and its constraints are as follows:

[0033]

[0034]

[0035]

[0036]

[0037] In the formula: F represents the hydrogen buffer tank capacity at time t. t outThe outlet flow rate of the hydrogen buffer tank is used to supply downstream hydrogen applications; and These are the upper and lower limits of the available storage capacity range for the hydrogen buffer tank; Δt represents the ramp rate of hydrogen emission; Δt represents the scheduling step size. These represent the upper and lower limits of the ramp rate for hydrogen emission.

[0038] Furthermore, step 2 specifically includes:

[0039] Step 2.1: The objective function is:

[0040]

[0041] In the formula: This represents the initial state x0 of the system and the initial value of the prediction error. Expected conditions under the following conditions; and C PV These represent the price of hydrogen and the cost per kilowatt-hour of electricity purchased from photovoltaic power plants, respectively; C SU and C SD These represent the operating costs for starting up and shutting down the electrolyzer, respectively; α and β are cost conversion factors, the quadratic term corresponding to α represents maximizing energy utilization, and the quadratic term corresponding to β represents ensuring that the hydrogen buffer tank storage is as close as possible to the initial value after the entire scheduling cycle; and These represent the hydrogen buffer tank capacity at the end time T and the initial time, respectively. For augmented state variables, where: Represents state variables, Represents the control variables; Q is a constant coefficient matrix; T is the scheduling period, x T Let T be the state variable at the final time point.

[0042] Step 2.2: Express the inequality constraints in the photovoltaic power generation hydrogen production system in probabilistic form, construct chance constraints, and represent them in vector form:

[0043]

[0044] In the formula: γ represents the tolerance size; Pr[·] represents the probability; This represents the coefficient vector of the inequality constraints. Ω represents the upper limit of the inequality constraint. C The set consisting of all constraints;

[0045] Step 2.3: Design the system's control law using an affine strategy, parameterizing the control commands as an affine function of the photovoltaic power output prediction error:

[0046]

[0047] In the formula: K is a constant term. t This is the gain coefficient matrix; therefore, Simplify to

[0048] Step 2.4: Combining the established objective function, constraints, and control function, the stochastic optimization scheduling model for the electrolytic cell cluster is shown below:

[0049]

[0050] st

[0051]

[0052]

[0053]

[0054]

[0055]

[0056] In the formula: φ is an integer variable. i The coefficients of integer variables; , which represents the upper limit of the constraint; A, B, C, and D are the coefficient matrices of the state variable, control variable, random variable, and integer variable, respectively.

[0057] Furthermore, step 3 specifically includes:

[0058] Step 3.1: Transform the stochastic optimization scheduling model of the electrolyzer cluster into a deterministic optimization model by using trajectory sensitivity decomposition of augmented state variables; specifically, this includes:

[0059] Step 3.1.1: Variables in the stochastic process Decomposed into baseline trajectory variables and trajectory sensitivity variables As shown below:

[0060]

[0061]

[0062] In the formula: and These are the baseline values ​​of the state variables, the trajectory variables, and the trajectory sensitivity variables, respectively. is the step size of the series expansion terms; o(·) denotes higher-order infinitesimal terms; and These are the baseline trajectory variable and the trajectory sensitivity variable of the random variable, respectively.

[0063] Step 3.1.2: Decompose equations (20)-(21) in the stochastic optimization scheduling model into the baseline trajectory equation and trajectory sensitivity equation as shown below;

[0064] Baseline value trajectory equation:

[0065]

[0066]

[0067] Where: the initial condition is

[0068] Trajectory sensitivity equation:

[0069]

[0070]

[0071] Where: the initial condition is This refers to the gain coefficient in the system control law;

[0072] Step 3.1.3: Based on M t The trajectory sensitivity decomposition decomposes the objective function (19) as follows:

[0073] J = J0 + J1 (31)

[0074]

[0075]

[0076] In the formula: The baseline value of the state variable at time T is the trajectory variable. The trajectory sensitivity variable is in the form of a set of variables; This indicates taking the derivative with respect to each variable in the function;

[0077] Step 3.1.4: To ensure the optimized scheduling model is convex, the following relaxation process is performed:

[0078]

[0079]

[0080] In the formula: The baseline value trajectory variable form of the variable set;

[0081] Opportunity constraint (23) expression:

[0082]

[0083] In the formula: κ γ For constant terms, the quantiles of γ in the standard normal distribution are represented.

[0084] Step 3.1.5: Express equations (19)-(24) in the stochastic optimization scheduling model as the following deterministic optimization problem:

[0085] min J=J0+J1 (37)

[0086] st

[0087]

[0088]

[0089]

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096] Step 3.2: Use model predictive control to solve the stochastic optimization scheduling model (37)-(46) in a rolling manner;

[0097] Step 3.2.1: Given the initial state x of the photovoltaic power generation hydrogen production system at initial time t. t , and z t Given a control period T; where, This represents the prediction error at the initial time.

[0098] Step 3.2.2: Calculate the predicted photovoltaic output value at time t+1. As input to model predictive control, the stochastic optimization scheduling model shown in equations (37)-(46) is solved to obtain the system state x at time t+1. t+1 , and z t+1 and control laws

[0099] Step 3.2.3: Repeat step 3.2.2 until t>T; output the optimal scheduling result x of the photovoltaic power generation hydrogen production system.T , z T And the objective function J.

[0100] The beneficial effects of this invention are:

[0101] 1) This invention proposes to model the uncertainty of photovoltaic output based on the Ito process model, and realize the unified modeling, analysis and control of the cluster scheduling model of electrolyzers under the stochastic dynamics framework. It establishes a stochastic optimization scheduling model for electrolyzer clusters that considers the uncertainty of photovoltaic output; it can dynamically adjust the operating status and power allocation of each electrolyzer, realize efficient electro-hydrogen production under the condition of photovoltaic output uncertainty, and improve its ability to absorb fluctuating photovoltaic power.

[0102] 2) Based on dynamic trajectory sensitivity decomposition, this invention transforms the stochastic optimization scheduling problem into deterministic optimization. It decomposes the objective function and chance constraints of the stochastic optimization scheduling model, transforms the stochastic dynamic optimization problem into deterministic second-order cone programming, and adopts model predictive control rolling solution, which effectively avoids the shortcomings of high computational complexity and low solution efficiency of traditional random sampling simulation methods. Attached Figure Description

[0103] Figure 1 This is a schematic diagram of the stochastic optimization scheduling method for the electrolyzer in the photovoltaic hydrogen production system based on the Ito process of the present invention.

[0104] Figure 2 The overall structure of a photovoltaic power generation hydrogen production system.

[0105] Figure 3 The diagram shows the operating status of each electrolytic cell during 96 time periods in the embodiment.

[0106] Figure 4 The photovoltaic output and electrolytic cell cluster power are shown in the examples. Detailed Implementation

[0107] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The present invention models the uncertainty of photovoltaic output based on the Iton process, enabling stochastic optimization scheduling of electrolyzer clusters under the condition of photovoltaic output uncertainty, achieving efficient electro-hydrogen production, and improving its ability to absorb fluctuating photovoltaic power. First, a photovoltaic power generation hydrogen production system model is established, including the Iton process model of photovoltaic output, the electrolyzer cluster scheduling model, and the hydrogen storage tank model. Then, with the objective function of maximizing the expected economic benefits and energy utilization rate under the influence of photovoltaic output prediction error uncertainty, and using an affine strategy to design the system's control law, a stochastic optimization scheduling model for the electrolyzer cluster is established based on the above model. Finally, based on trajectory sensitivity decomposition, the stochastic optimization control model is transformed into a deterministic optimization problem, and a rolling optimization solution is obtained using model predictive control. Specifically, as follows... Figure 1As shown, the detailed process is as follows:

[0108] Step 1: Establish a photovoltaic power generation hydrogen production system model, including the Itoh process model of photovoltaic output, the electrolyzer cluster scheduling model, and the hydrogen storage tank model.

[0109] A photovoltaic hydrogen production system consists of photovoltaic power generation, an electrolyzer cluster, and a hydrogen buffer tank, among other units. The overall structure is as described above. Figure 2 As shown.

[0110] 1.1 Itoh Process Model for Photovoltaic Output

[0111]

[0112] In the formula: P t PV It contributes to the actual photovoltaic power generation; P t PV,pred Forecast value of photovoltaic power output; This represents the prediction error.

[0113] The Itoh process model for prediction error is as follows:

[0114]

[0115]

[0116]

[0117] In the formula: W t Represents the standard Wiener process; For drift term; t is the diffusion term; a is the time constant of cloud shadow change; b is the mean recovery target of the drift term; r t s These represent sunrise and sunset times, respectively; the sine function represents the solar altitude angle; c represents the uncertainty intensity; and d and e represent the range of irradiance variation.

[0118] 1.2 Electrolytic Cell Cluster Scheduling Model

[0119] Logical constraints for the transition of operating status of a single electrolytic cell:

[0120]

[0121]

[0122]

[0123]

[0124] In the formula: n represents the nth electrolytic cell in the electrolytic cell cluster; These represent the three operating states of the electrolytic cell: production, standby, and shutdown. These are the start-up and shutdown operations for the electrolytic cell. This indicates the transition of the electrolytic cell from standby to operating state; all the above variables are represented in binary; the subscripts t-1 and t-2 represent the previous scheduling period and the scheduling period before that when the scheduling period is t, respectively.

[0125] The physical constraints of the electrolytic cell cluster itself:

[0126]

[0127]

[0128]

[0129] Where: N is the total number of electrolytic cells; P n,t P is the power of the nth electrolytic cell at time t; max P min These represent the upper and lower power limits for a single electrolytic cell under production conditions; P SB F represents the constant power of the auxiliary equipment in standby mode for a single electrolytic cell; n,t Let A1, A2, and A3 be the hydrogen production of the nth electrolyzer; A1, A2, and A3 are constant coefficients determined by the characteristics of the electrolyzer.

[0130] 1.3 The storage capacity and capacity constraints of the hydrogen buffer tank are as follows:

[0131]

[0132]

[0133]

[0134]

[0135] In the formula: For hydrogen buffer tank capacity; F t out The outlet flow rate of the hydrogen buffer tank is used to supply downstream hydrogen applications; and These are the upper and lower limits of the available storage capacity range for the hydrogen buffer tank; Δt represents the ramp rate of hydrogen emission; Δt represents the scheduling step size. These represent the upper and lower limits of the ramp rate for hydrogen emission.

[0136] Step 2: Taking the highest expected economic benefits and highest energy utilization rate under the influence of uncertainties in photovoltaic output prediction errors as the objective function, a stochastic optimization scheduling model for the electrolyzer cluster is established based on the photovoltaic power generation hydrogen production system model and by using affine strategy to design the system's control law.

[0137] 2.1 The objective function is to maximize the expected economic benefits and energy utilization efficiency of the electrolytic cell cluster under the influence of uncertainties in photovoltaic output prediction errors:

[0138]

[0139] In the formula: This represents the initial state x0 of the system and the initial value of the prediction error. Expected conditions under the following conditions; C PV These represent the price of hydrogen and the cost per kilowatt-hour of electricity purchased from photovoltaic power plants, respectively; C SU C SD These represent the operating costs for starting up and shutting down the electrolyzer, respectively; α and β are cost conversion factors, the quadratic term corresponding to α represents maximizing energy utilization, and the quadratic term corresponding to β represents ensuring that the hydrogen buffer tank storage is as close as possible to the initial value after the entire scheduling cycle; For augmented state variables, where: Represents state variables, This represents the control variables; Q is a matrix of constant coefficients.

[0140] 2.2 The inequality constraints in the photovoltaic power generation hydrogen production system are expressed in probabilistic form, and chance constraints are constructed and represented in vector form:

[0141]

[0142] In the formula: γ represents the tolerance size; Pr[·] represents the probability; This represents the coefficient vector of the inequality constraints. Ω represents the upper limit of the inequality constraint. C It is the set of all constraints.

[0143] 2.3 The control law of the system is designed using an affine strategy, and the control command is parameterized as an affine function of the photovoltaic power output prediction error:

[0144]

[0145] In the formula: K is a constant term. t This is the gain coefficient matrix.

[0146] 2.4 Combining the established objective function, constraints, and control function, the stochastic optimization scheduling model for the electrolytic cell cluster is shown below:

[0147]

[0148] st

[0149]

[0150]

[0151]

[0152]

[0153]

[0154] In the formula: φ is an integer variable. i The coefficients of integer variables; To constrain the upper limit.

[0155] Step 3: Based on trajectory sensitivity decomposition, the stochastic optimization scheduling model is transformed into a deterministic optimization problem, and a rolling solution is adopted in the form of MPC, as follows:

[0156] 3.1 The stochastic optimization scheduling model (19)-(24) is transformed into a deterministic optimization model by trajectory sensitivity decomposition of augmented state variables.

[0157] First, the variables in the stochastic process Decomposed into baseline trajectory variables and trajectory sensitivity variables As shown below:

[0158]

[0159]

[0160] Therefore, equations (20)-(21) in the stochastic optimization scheduling model can be decomposed into the baseline trajectory equations (27)-(28) and trajectory sensitivity equations (29)-(30) as shown below.

[0161] The baseline trajectory equations of equations (20)-(21) are:

[0162]

[0163]

[0164] Where: the initial condition is

[0165] Trajectory sensitivity equations of equations (20)-(21):

[0166]

[0167]

[0168] Where: the initial condition is This is the gain coefficient in the system control law.

[0169] Then, based on M t The trajectory sensitivity decomposition decomposes the objective function (19) as follows:

[0170] J = J0 + J1 (31)

[0171]

[0172]

[0173] In the formula:

[0174] To ensure that the optimized scheduling model is convex, the following relaxation process is performed:

[0175]

[0176]

[0177] And the opportunity constraint (23) can be expressed as:

[0178]

[0179] Thus, the stochastic optimization scheduling model (19)-(24) can be expressed as the following deterministic optimization problem:

[0180] min J=J0+J1 (37)

[0181] st

[0182]

[0183]

[0184]

[0185]

[0186]

[0187]

[0188]

[0189]

[0190]

[0191] 3.2 The stochastic optimal scheduling model is solved in a rolling manner using model predictive control (37)-(46).

[0192] Step 1): Given the initial state x of the photovoltaic power generation hydrogen production system at initial time t. t , z t Given a control period T.

[0193] Step 2): Calculate the predicted photovoltaic power output at time t+1. As input to model predictive control, the stochastic optimization scheduling model shown in equations (37)-(46) is solved to obtain the system state x at time t+1. t+1 , z t+1 and control laws

[0194] Step 3): Repeat step 2 until t > T. Output the optimal scheduling result x of the photovoltaic power generation hydrogen production system. T , z T And the objective function J.

[0195] 4. Case Analysis

[0196] This example selects a photovoltaic power station in a region of southwestern Sichuan. The hydrogen production plant is equipped with a cluster of 4 alkaline electrolyzers and is subject to 24-hour optimized scheduling with a scheduling step size of 15 minutes.

[0197] This example builds a stochastic optimization scheduling model for electrolytic cells based on the Wolfram Mathematica platform and solves it using the Mosek solver.

[0198] The operating status of each electrolytic cell during the 96-hour period in the example is as follows: Figure 3 Examples of photovoltaic power output and electrolytic cell cluster power are as follows: Figure 4 As shown.

[0199] As can be seen, the stochastic optimization scheduling method for the electrolyzer cluster of the photovoltaic hydrogen production system based on the Ito process proposed in this invention can dynamically adjust the operating status of each electrolyzer, effectively realize efficient electro-hydrogen production under uncertain photovoltaic output conditions, and fully absorb fluctuating photovoltaic output.

Claims

1. A stochastic optimization scheduling method for photovoltaic power generation hydrogen production clusters based on the Itō process, characterized in that, Includes the following steps: Step 1: Establish a photovoltaic power generation hydrogen production system model, including the Itoh process model of photovoltaic output, the electrolyzer cluster scheduling model, and the hydrogen storage tank model; Step 2: Taking the highest expected economic benefits and highest energy utilization rate under the influence of uncertainties in photovoltaic output prediction errors as the objective function, based on the photovoltaic power generation hydrogen production system model and using affine strategy to design the system control law, a stochastic optimization scheduling model of the electrolyzer cluster is established. Step 3: Based on trajectory sensitivity decomposition, the stochastic optimization control model is transformed into a deterministic optimization problem, and the model predictive control is used to solve it through rolling optimization. Step 2 specifically involves: Step 2.1: The objective function is: （16）； In the formula: Indicates the initial state of the system and initial value of prediction error Expected conditions under the following conditions; and These are the price of hydrogen and the cost per kilowatt-hour of electricity purchased from photovoltaic power plants, respectively. and These are the operating costs for starting up and shutting down the electrolytic cell, respectively. , All are cost conversion factors, and The corresponding quadratic term represents maximizing energy utilization, and The corresponding quadratic term indicates that the hydrogen buffer tank storage is as close as possible to the initial value after the entire scheduling cycle. and These represent the hydrogen buffer tank capacity at time T and the initial hydrogen buffer tank capacity, respectively. For augmented state variables, where: Represents state variables, Indicates control variables; A constant coefficient matrix; T represents the scheduling period. Here, N represents the state variable at time T; N is the total number of electrolytic cells. for Time of the first The power of the Taiwan electrolytic cell; For the first Hydrogen production of the Taiwan electrolyzer; and These represent the start-up and shutdown operations of the electrolytic cell, respectively, and are represented by binary variables. This represents the prediction error; The outlet flow rate of the hydrogen buffer tank is used to supply downstream hydrogen applications; The rate at which hydrogen is discharged from the slope; Step 2.2: Express the inequality constraints in the photovoltaic power generation hydrogen production system in probabilistic form, construct chance constraints, and represent them in vector form: （17）； In the formula: Indicates the size of the tolerance; Represents probability; This represents the coefficient vector of the inequality constraints. This represents the upper limit of the inequality constraint. The set consisting of all constraints; Step 2.3: Design the system's control law using an affine strategy, parameterizing the control commands as an affine function of the photovoltaic power output prediction error: （18）； In the formula: For constant terms, This is the gain coefficient matrix; therefore, Simplify to ; Step 2.4: Combining the established objective function, constraints, and control function, the stochastic optimization scheduling model for the electrolytic cell cluster is shown below: （19）； st （20）； （21）； （22）； （23）； （24）； In the formula: It is an integer variable; The coefficients of integer variables; To constrain the upper limit; , , and These are the coefficient matrices for state variables, control variables, random variables, and integer variables, respectively. , and These represent the three operating states of the electrolytic cell: production, standby, and shutdown, respectively, and are represented by binary variables. The transition of the electrolytic cell from standby to operating state is represented by binary variables.

2. The stochastic optimization scheduling method for photovoltaic power generation hydrogen production clusters based on the Ito process according to claim 1, characterized in that, The Itoh process model for photovoltaic power output is as follows: （1） In the formula: To contribute to the actual development of photovoltaics; Forecast value of photovoltaic power output; The Itoh process model for prediction error is as follows: （2）； （3）； （4）； In the formula: Represents the standard Wiener process; For drift term; denoted as the diffusion term; a is the time constant of cloud shadow change; b is the mean recovery target of the drift term; t r , t s These represent sunrise and sunset times, respectively; the sine function represents the solar altitude angle; c represents the uncertainty intensity; and d and e represent the range of irradiance variation.

3. The stochastic optimization scheduling method for photovoltaic power generation hydrogen production clusters based on the Ito process according to claim 2, characterized in that, The electrolytic cell cluster scheduling model includes: Logical constraints for the transition of operating status of a single electrolytic cell: （5）； （6）； （7）； （8）； In the formula: n represents the nth electrolytic cell in the electrolytic cell cluster; the subscripts t-1 and t-2 represent the previous scheduling period and the scheduling period before that when the scheduling period is t, respectively; The physical constraints of the electrolytic cell cluster itself: （9）； （10）； （11）； In the formula: , These are the upper and lower limits of power constraints for a single electrolytic cell in production mode; This refers to the constant power of the auxiliary equipment in standby mode for a single electrolytic cell; , , It is a constant coefficient, determined by the characteristics of the electrolytic cell.

4. The stochastic optimization scheduling method for photovoltaic power generation hydrogen production clusters based on the Ito process according to claim 3, characterized in that, The hydrogen storage tank model includes: The hydrogen buffer tank storage capacity and its constraints are as follows: （12）； （13）； （14）； （15）； In the formula: for Hydrogen buffer tank capacity at any given time; and These are the upper and lower limits of the available storage capacity range for the hydrogen buffer tank; For scheduling step size; , These represent the upper and lower limits of the ramp rate for hydrogen emission.

5. The stochastic optimization scheduling method for photovoltaic power generation hydrogen production clusters based on the Ito process according to claim 4, characterized in that, Step 3 specifically involves: Step 3.1: Transform the stochastic optimization scheduling model of the electrolyzer cluster into a deterministic optimization model by using trajectory sensitivity decomposition of augmented state variables; specifically, this includes: Step 3.1.1: Variables in the stochastic process Decomposed into baseline trajectory variables and trajectory sensitivity variables As shown below: （25）； （26）； In the formula: and These are the baseline values ​​of the state variables, the trajectory variables, and the trajectory sensitivity variables, respectively. This represents the step size of the series expansion terms; Represents higher-order infinitesimal terms; and These are the baseline trajectory variable and the trajectory sensitivity variable of the random variable, respectively. Step 3.1.2: Decompose equations (20)-(21) in the stochastic optimization scheduling model into the baseline trajectory equation and trajectory sensitivity equation as shown below; Baseline value trajectory equation: （27）； （28）； Where: the initial condition is , ; Trajectory sensitivity equation: （29）； （30）； Where: the initial condition is , ; , where is the gain coefficient in the system control law; Step 3.1.3: Based on the The trajectory sensitivity decomposition decomposes the objective function (19) as follows: （31）； （32）； （33）； In the formula: The baseline value of the state variable at time T is the trajectory variable. ; The trajectory sensitivity variable is in the form of a set of variables; This indicates taking the derivative with respect to each variable in the function; Step 3.1.4: To ensure the optimized scheduling model is convex, the following relaxation process is performed: （34）； （35）； In the formula: The baseline value trajectory variable form of the variable set; Opportunity constraint (23) expression: （36）； In the formula: The constant term corresponds to the standard normal distribution. quantiles; Step 3.1.5: Express equations (19)-(24) in the stochastic optimization scheduling model as the following deterministic optimization problem: （37）； st （38）； （39）； （40）； （41）； （42）； （43）； （44）； （45）； （46） Step 3.2: Use model predictive control to solve the stochastic optimization scheduling model (37)-(46) in a rolling manner; Step 3.2.1: Given the initial state of the photovoltaic power generation hydrogen production system at initial time t , and Given a control period T; where, This represents the prediction error at the initial time. Step 3.2.2: Calculate the predicted photovoltaic output value at time t+1. As input to model predictive control, the stochastic optimal scheduling model shown in equations (37)-(46) is solved to obtain the system state at time t+1. , and and control laws ; Step 3.2.3: Repeat step 3.2.2 until t>T; output the optimal scheduling result of the photovoltaic power generation hydrogen production system. , And the objective function J.

Images