A scheduling method for electrolytic water hydrogen synthesis ammonia with daily flexibility adjustment

Abstract

The present application relates to the configuration and operation optimization technology of the electric hydrogen ammonia coupling system, and aims to provide a scheduling method for the electrolytic water hydrogen synthesis ammonia with daily flexible adjustment. The method comprises the following steps: slicing the annual hourly data by month and layering in the rank space, and selecting a representative month to construct a characteristic matrix; introducing a daily flexible adjustment mechanism, and describing the synthetic ammonia load as a continuous piecewise linear operation curve; constructing a unit ammonia production power consumption function with the synthetic ammonia load as the independent variable and a linearly solvable synthetic ammonia segment power model; and constructing a system full-chain mixed integer linear programming model, taking the minimum annual total cost as the target, and performing annualized weighted solution according to the representative month weight to obtain the optimal system configuration and operation scheduling result of the whole year. The present application can realize the sparse distribution of the adjustment section, effectively suppress the sawtooth fluctuation of the load curve and the invalid micro-amplitude adjustment, reduce the synthetic ammonia load climbing frequency, and improve the operation executability and equipment life friendliness.

Description

Technical Field

[0001] This invention relates to the field of configuration and operation optimization of electro-hydrogen-ammonia coupling systems, and specifically to a scheduling method for an intraday flexible-adjustment electrolysis water-to-hydrogen-to-ammonia synthesis system. Background Technology

[0002] With the accelerated low-carbon transformation of the energy structure, renewable energy sources such as wind power and photovoltaics are being integrated into the power system on a large scale. However, renewable energy is characterized by randomness, intermittency, and uncertainty, which places higher demands on the safe and stable operation of the power system. Electrolysis of water to produce hydrogen and synthesize ammonia is an important pathway to promote the consumption of renewable energy and drive the green transformation of industry. Its configuration and operation optimization are directly related to energy utilization efficiency and system economy.

[0003] Wind and solar power output exhibits significant seasonality and intraday multi-scale fluctuations. Meanwhile, key state parameters such as battery state of charge and hydrogen storage tank inventory show strong inter-day correlations, resulting in strong temporal coupling characteristics in the configuration and operation of the water electrolysis hydrogen production and ammonia synthesis system. Directly using year-round data for modeling would drastically increase the number of variables and constraints, leading to a large-scale mixed-integer optimization problem with high computational costs and insufficient convergence stability. Conversely, dimensionality reduction based on typical days or weeks makes it difficult to preserve the continuous temporal structure over the entire month, thus reducing the reliability of capacity configuration and intraday scheduling conclusions.

[0004] In terms of modeling and solving for the electrical power in the ammonia synthesis stage, existing studies often approximate the electrical power in this stage as a linear relationship with a fixed coefficient to the ammonia production flow rate. While this approach facilitates modeling and solving, it neglects the differences in unit power consumption under different ammonia synthesis load levels, which can easily lead to systematic biases in energy consumption assessment and economic optimization. In actual operation, the electrical power of ammonia synthesis exhibits nonlinear characteristics as the load level changes; however, directly using a nonlinear expression introduces product terms and non-convex structures, significantly increasing the scale and difficulty of solving the optimization problem.

[0005] Regarding the handling of hydrogen flow direction, existing studies mostly adopt a single path: directly supplying hydrogen produced from the electrolyzer to the ammonia synthesis unit to simplify material coupling; or sending all produced hydrogen to a hydrogen storage tank, which then supplies hydrogen to the ammonia synthesis section. This fixed-destination approach ignores the synergistic adjustment space between direct supply and resupply via the storage tank, and the resulting comprehensive impact on unit capacity configuration and operational economy, making it difficult to reflect the optimal configuration of the hydrogen storage tank.

[0006] At the level of ammonia synthesis load dispatching strategies, existing methods can be mainly divided into two categories: the first is to continuously track renewable energy output hourly. Although this method can improve the local consumption rate of renewable energy in the short term, it is prone to forming sawtooth loads, inducing frequent thermomechanical cycles in ammonia synthesis units and accelerating equipment aging, resulting in insufficient feasibility and safety margin. The second is to operate with approximately constant load over a longer period. This method is simple to implement but ignores the time-varying nature of output and the ramp-up process, resulting in low wind-solar matching.

[0007] Furthermore, most current studies model the capacity configuration and intraday scheduling of the electro-hydrogen-ammonia coupled system separately, and research on integrated joint optimization of system capacity configuration and intraday scheduling under a unified modeling framework is still relatively limited.

[0008] Therefore, it is necessary to propose new solutions to address the aforementioned problems. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a scheduling method for an intraday flexible adjustment system for electrolytic water hydrogen production and ammonia synthesis.

[0010] To solve the technical problem, the solution of the present invention is:

[0011] A scheduling method for a water electrolysis hydrogen production and ammonia synthesis system with intraday flexible regulation is provided, which involves connecting wind and solar renewable energy sources into the power system for water electrolysis hydrogen production and ammonia synthesis; including:

[0012] (1) The hourly wind power and photovoltaic data of the whole year are sliced ​​by month. After calculating the monthly wind and solar energy, the data is divided into four quadrants in the Copula rank space. Multi-scale fluctuation features are extracted and the joint distance between months is constructed. A representative month is selected from each of the four quadrants to construct a feature matrix. Weights are assigned to the four selected representative months.

[0013] (2) Introduce a flexible adjustment mechanism within the time domain of the representative month, and describe the synthetic ammonia load as a continuous piecewise linear operating curve; by solving the annualized total cost objective function, concentrate the synthetic ammonia load change in a few preset controllable sections to match the fluctuation of wind and solar power and reduce the adjustment frequency of synthetic ammonia equipment.

[0014] (3) Construct a unit ammonia production power consumption function with the per-unit load of ammonia synthesis as the independent variable, and perform equivalent linearization; establish a linearly solvable ammonia synthesis section power model, and incorporate it into the mixed integer optimization framework after superimposing it with the power of the air separation unit.

[0015] (4) Construct a full-chain mixed integer linear programming model for the electrolysis of water to produce hydrogen and synthesize ammonia. With the goal of minimizing the annual total cost of the system, perform joint optimization on four representative months and solve the annual weighted solution according to the weight of the representative month to obtain the optimal system configuration and operation scheduling results for the whole year.

[0016] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0017] 1. This invention is based on the combination of rank space hierarchical structure and constraint D-optimal. It selects a small number of representative months from four typical scenarios and assigns weights to them. This preserves the frequency of wind and light joint distribution and extreme cases, while maintaining the temporal structure of the whole month, which significantly reduces the scale and computational cost of hourly modeling throughout the year.

[0018] 2. This invention introduces an intraday flexible adjustment mechanism. The ammonia load remains constant in the non-adjustment zone, while the load in the adjustment zone changes at a constant ramp rate. By introducing an ammonia adjustment cost term into the annualized total cost objective function to penalize the magnitude and frequency of load changes and the number of adjustment zones, a sparse distribution of adjustment zones is achieved. This effectively suppresses sawtooth fluctuations and ineffective micro-adjustments in the load curve, reduces the frequency of ammonia load ramp-ups, and improves operational feasibility and equipment lifespan friendliness.

[0019] 3. This invention establishes a unit power consumption function for ammonia synthesis using the per-unit load of ammonia synthesis as the independent variable and performs equivalent linearization. This improves the characterization accuracy of the ammonia synthesis power model without significantly increasing the difficulty of solving the problem, making it compatible with mixed integer optimization frameworks.

[0020] 4. In this invention, the hydrogen generated by water electrolysis is dynamically allocated between two paths: direct supply to ammonia synthesis and supply via storage. This is based on the optimization solution results, avoiding capacity configuration deviations caused by a single-path assumption, and taking into account energy utilization, safety, and economy.

[0021] 5. This invention aims to minimize the total annual cost of the system. It simultaneously optimizes the device scale and intraday operation strategy in each representative month and aggregates them annually by weight. This avoids deviations caused by simply configuring or scheduling, and improves the consistency and reliability of the system's decisions throughout the year. Attached Figure Description

[0022] Figure 1 This represents the selection results of typical months and candidate months in the four quadrants of this invention.

[0023] Figure 2 This is a diagram showing the power output and load demand of wind and solar power in four typical months according to the present invention.

[0024] Figure 3 The figure shows the unit ammonia production power and efficiency curve of this invention.

[0025] Figure 4 This is the power and efficiency curve of ammonia production per unit area under strong wind and strong light conditions according to the present invention.

[0026] Figure 5 This is a schematic diagram of the flexible time window under strong wind and strong light conditions according to the present invention.

[0027] Figure 6 This is a diagram showing the ammonia synthesis flow rate and ramp rate under strong wind and strong light conditions according to the present invention.

[0028] Figure 7 This is a hydrogen production diversion diagram under strong wind and strong light conditions according to the present invention. Detailed Implementation

[0029] The specific implementation process of the present invention will be described in detail below with reference to the accompanying drawings.

[0030] This invention proposes a system optimization scheduling method that balances capacity allocation and operational optimization based on hourly data of wind power, photovoltaic power, and load throughout the year. First, the wind and solar energy for each month is stratified in rank space, and four representative months are selected and assigned weights to maintain annual statistical representativeness and monthly temporal continuity. An intraday flexible adjustment mechanism is introduced in the time domain of the representative months, stipulating that the ramp rate of ammonia load in the non-adjustment zone is zero, while the ramp rate in the adjustment zone is constant. High-frequency fluctuations are replaced by a small number of linearly varying load segments to achieve stable and controllable adjustment of the ammonia load. A unit power consumption function with the per-unit load of ammonia as the independent variable is established, and a linearized expression of ammonia power is obtained through an equivalent linearization method. This expression is then integrated with the power of the air separation unit into a mixed integer optimization framework. Allocation constraints for electrolytic hydrogen production are established for two paths: direct supply and storage-re-supply, describing the balance of hydrogen flow within the system. An integrated mixed integer programming model for capacity allocation and operational scheduling is constructed, with the goal of minimizing the annualized total cost. The optimal system configuration and scheduling results are output, thereby improving the renewable energy absorption capacity while considering equipment lifespan and economic efficiency.

[0031] Specifically, the present invention includes the following steps:

[0032] 1. The hourly wind and solar power data for the whole year are sliced ​​into monthly slices. The monthly wind and solar energy are calculated and layered into four quadrants in the Copula rank space: strong wind and strong solar, strong wind and weak solar, weak wind and strong solar, and weak wind and weak solar. Multi-scale fluctuation features are extracted and joint distance between months is constructed. Under the constraint of taking one month from each of the four quadrants, four representative months are selected by constrained D-optimal and assigned weights.

[0033] (1) To represent the monthly total, the annual hourly series is first sliced ​​by natural month.

[0034] The specific formulas for calculating monthly wind energy and monthly solar energy are as follows:

[0035]

[0036]

[0037] in, , These represent the total wind power energy and total solar power energy in month m, respectively. , Let t represent the wind power and solar power outputs respectively in hour t. For the hour of month m.

[0038] Will and Perform an empirical distribution rank transformation to obtain the rank coordinates. Specifically,

[0039] Monthly wind and solar energy are mapped to the (0,1) interval using rank transformation to form corresponding Copula coordinates, and then thresholded to... , The 12 months are divided into four quadrants; when the wind energy order or light energy rank If the value is not lower than the threshold, it is judged as strong wind or strong light respectively; otherwise, it is weak wind or weak light. If any quadrant is empty, the corresponding threshold is shrunk to 0.5 until all four quadrants are not empty or the threshold returns to 0.5. The weight of each representative month is equal to the proportion of the month in that quadrant. Parallel ranks are processed by the average method.

[0040] The threshold pair ( , () refers to two threshold parameters used to distinguish between wind energy rank values ​​and solar energy rank values, where , These are the threshold values ​​for the rank values ​​of wind and solar energy, respectively.

[0041] The Copula coordinates are calculated using the following formula:

[0042]

[0043]

[0044] in, , , respectively, represent the normalized rank positions of month m in the wind and light dimensions; To obtain the ascending rank and average the results of the tied items; , These represent the total wind power and total solar power in month m, respectively; M is the total number of months in a year.

[0045] To maintain the combined wind and solar structure, the Copula rank space is divided into four quadrants: strong wind and strong light, strong wind and weak light, weak wind and strong light, and weak wind and weak light, and each quadrant is guaranteed to be non-empty.

[0046]

[0047]

[0048]

[0049]

[0050] in, , , , These correspond to strong wind and strong light, strong wind and weak light, weak wind and strong light, and weak wind and weak light, respectively. , This is the threshold value for the rank space, initially set to 0.5; if an empty group is found, then... The system automatically adjusts to ensure that each group is not empty.

[0051] (2) Monthly total data alone cannot reflect the multi-scale fluctuation characteristics of wind and solar energy. Therefore, multi-dimensional features are constructed for each month to characterize these multi-scale fluctuations. These multi-dimensional features specifically include the intensity of wind and solar energy fluctuations, intraday amplitude of wind and solar energy, and the intraday fundamental frequency energy percentage.

[0052] The formulas for measuring the intensity of wind and solar radiation are as follows:

[0053]

[0054] in, This represents the total variation in wind or light in the m-th month sequence; a larger index indicates greater fluctuations. This refers to the hourly sequence of wind and solar power output in month m. The solar power output value for hour t.

[0055] The formula for calculating the intraday amplitude of scenery is as follows:

[0056]

[0057] in, An indicator for measuring the daily range of scenic views; for Quantile operator.

[0058] The formula for calculating the proportion of fundamental frequency energy during the day is as follows:

[0059]

[0060] in, The percentage of spectral energy in the fundamental frequency neighborhood of the m-th month sequence over a 24-hour period; This represents the spectral power obtained by performing a discrete Fourier transform on the mean-reduced hourly sequence of wind power or photovoltaic power; k is the index of the corresponding frequency component. This is the set of indices near the 24-hour base frequency.

[0061] (3) In order to simultaneously characterize the distribution pattern and temporal fluctuation of wind and solar power output, for any month i , j The joint distance is defined as the weighted Euclidean sum of the vector differences of the wind power quantile curves, the vector differences of the photovoltaic quantile curves, and the vector differences of the multi-scale eigenvectors; the joint distance is calculated using the following formula:

[0062]

[0063] in, Indicates month i With Month j The joint distance, the smaller the value, the more similar the two months are; by month i For example, , These are the equidistant quantile curve vectors for wind power and photovoltaic power, respectively. Multi-scale feature row vectors; It is the Euclidean norm; α , β , γ For weights.

[0064] (4) To improve information coverage and recognizability, candidate month sets are formed in each of the four quadrants, and representative months are selected using the constrained D-optimal criterion; specifically, the following steps are included:

[0065] (4.1) Object Limitation s Month set Calculate the intragroup medoid score:

[0066]

[0067] in, For month m in quadrant s The medoid score within.

[0068] (4.2) Sort in ascending order and select the top K to form a candidate set:

[0069]

[0070] in, Quadrant s The set of candidate months; Quadrant s The set of months; Indicates in set Internal The ascending rank is used; if there are ties, the average rank is taken; K is the upper limit of the number of candidates.

[0071] (4.3) From the four quadrants respectively Select January to form candidate combinations And construct the characteristic matrix X (c), as shown in the following formula:

[0072]

[0073] in, For month m, the multi-scale feature row vector; Quadrant s The month index of the selected representative month.

[0074] (4.4) Using the constrained D-optimal criterion as the objective, calculate the value of the objective function representing the month selection for each candidate combination, and select the combination that maximizes the objective function; if multiple combinations achieve the same maximum objective function value, then the combination that maximizes the objective function value is selected first. The smallest combination. It refers to the sum of the medoid scores of each month in the selected month combination.

[0075] The objective function for selecting the representative month is:

[0076]

[0077] in, An objective function is selected to represent the months, measuring the overall information content of different combinations of candidate months. A larger objective function value indicates higher coverage and stronger diversity in the feature space, making it more representative of the year's operational characteristics. ε > 0 is a diagonal regularization coefficient used to enhance numerical stability. It is the identity matrix; It refers to the determinant of a matrix; Refers to the candidate month feature matrix X The Gram matrix reflects the correlation structure of candidate months in the feature space.

[0078] (4.5) Determine the weights for the four selected representative months as shown in the following formula:

[0079]

[0080] in, Quadrant s The representative monthly weight; The number of months in quadrant s; This represents the total number of months in a year.

[0081] 2. A flexible intraday adjustment mechanism is introduced within the representative monthly time domain, describing the ammonia synthesis load as a continuous piecewise linear operating curve, divided into a stable operating segment and a preset flexible adjustment segment. To avoid frequent fluctuations in the ammonia synthesis load leading to drastic changes in pressure and temperature of the ammonia synthesis equipment, which could damage the catalyst and pose safety risks, the ammonia synthesis load is only varied at a constant ramp rate within the permissible flexible adjustment range, while the ammonia synthesis load remains unchanged at other times.

[0082] By introducing a synthetic ammonia adjustment cost term into the annualized total cost objective function, penalties can be imposed on the changes in synthetic ammonia load and the number of flexible adjustment intervals, thereby achieving sparsity of adjustment segments. The number, start and end times, and duration of these segments are automatically determined through the optimization solution of the annualized total cost objective function, which can concentrate synthetic ammonia load changes in a small number of controllable segments to match wind and solar power fluctuations and reduce the adjustment frequency of synthetic ammonia equipment.

[0083] The flexible adjustment modeling constraints corresponding to the annualized total cost objective function include the following:

[0084] (1) The hourly variation of ammonia synthesis load is related to the maximum ammonia synthesis yield and the ramp-up rate, as shown in the following formula:

[0085]

[0086] in, Let t be the ammonia synthesis yield; This represents the maximum yield of synthesized ammonia. The ammonia synthesis load ramp-up rate; Represents the total number of hours in a month.

[0087] (2) The state of the flexible adjustment range is described by activating an indicator variable. When the indicator variable is 1, the ramp rate is constrained by its upper and lower limits. When the indicator variable is 0, the ramp rate is 0, thus limiting the ammonia synthesis load to change only within the selected adjustment range, while remaining stable in other periods. The specific constraint form is as follows:

[0088]

[0089] in, The interval indicator variable is a binary variable that indicates whether the interval is in an active state. The ammonia synthesis load ramp-up rate; , These represent the lower and upper limits of the ammonia synthesis ramp-up rate.

[0090] (3) To ensure smooth cross-day connections, the daily 24-hour ramp rate is set to zero, and a condition representing the first and last months of the month with equal values ​​is applied to facilitate annualized aggregation:

[0091]

[0092]

[0093] in, For day sequence number; This represents the total number of hours in a month.

[0094] (4) Furthermore, to ensure the executability of flexible adjustment section operation and obtain sparse solutions, a logical consistency constraint is established between the adjustment section state indicator variable and the section state switching event variable to ensure that the synthetic ammonia load ramp-up rate remains constant within each adjustment section, and a synthetic ammonia load adjustment penalty term is introduced into the annualized total cost objective function. Specifically:

[0095] (4.1) To accurately describe the start and end changes of the adjustment segment and to ensure logical consistency between the event variable and the state indicator, a start event of the adjustment segment is defined. s ( t ) and the end of the adjustment section event e ( t )as follows:

[0096]

[0097] (4.2) To ensure a constant ammonia loading ramp rate within the same regulation zone and to achieve a continuous and smooth transition of ammonia loading between the inside and outside of the zone, a consistent ramp rate constraint between adjacent time points needs to be applied:

[0098]

[0099] in, For a sufficiently large constant; when When This ensures that the gradient rate remains constant within each adjustment range.

[0100] (4.3) To suppress ineffective regulation and constrain the number of regulation sections, a penalty cost for opening regulation sections is introduced into the regulation cost of ammonia synthesis, as shown in the following formula:

[0101]

[0102] in, The penalty cost for activating the adjustment zone; This is the penalty coefficient.

[0103] (4.4) To characterize the adjustment scheme in the optimization results, the number of adjustment segments, the start time, and the duration are all determined by the optimization solution:

[0104]

[0105]

[0106]

[0107] in, To adjust the number of sections; Represents the total number of hours in a month; The set representing the starting times of the adjustment segment; For the first The duration of each adjustment segment; For the first The start time of each adjustment segment; To adjust the candidate length of the segment.

[0108] 3. Construct a unit ammonia production power consumption function with the per-unit load of ammonia synthesis as the independent variable. Use SOS2 convex combination to perform piecewise linear characterization of the ammonia synthesis energy consumption curve. Use McCormick envelope linearization to process the bilinear terms introduced by the convex combination and establish a linearly solvable ammonia synthesis section power model. After superimposing it with the power of the air separation unit, incorporate it into the mixed integer optimization framework.

[0109] (1) Construct an electrical power model for the ammonia synthesis section (Haber-Bosch section). This model uses a unit power consumption function with the ammonia synthesis per-unit load level as the independent variable. Specifically, it is expressed as the formula for ammonia synthesis power minus ammonia production flow rate based on the ammonia synthesis per-unit load:

[0110]

[0111]

[0112]

[0113] in, Per unit load for ammonia synthesis; Electricity consumption per unit of ammonia production; This represents the electrical power of the non-linearized ammonia synthesis section. For the first Hourly ammonia production flow rate; This refers to the rated maximum ammonia production flow rate of the ammonia synthesis unit. This represents the minimum per-unit load level for ammonia synthesis that is permissible for the ammonia synthesis unit. This is a function of electricity consumption per unit of ammonia production; , , , The polynomial coefficients are the power consumption function per unit ammonia production.

[0114] (2) The above formula is equivalently linearized. SOS2 convex combination is used on the set of breakpoints and the product term is processed with McCormick envelope to obtain the linearized expression of the power of ammonia synthesis.

[0115] First, a set of monotonic breakpoints is constructed within the per-unit load value range, and piecewise convex combination interpolation of the per-unit load is performed using SOS2 weights, as shown in the following formula:

[0116]

[0117]

[0118] in, SOS2 weights; Per unit load for ammonia synthesis; This represents the per-unit load break point for ammonia synthesis.

[0119] Then, to transform the product of weights and ammonia production into a linear form, variables are introduced. This represents the product of the weight and the synthetic ammonia production, and is defined by the upper and lower bounds of the production capacity. and Under known conditions, the McCormick envelope is applied as follows:

[0120]

[0121] in, This is the lower bound of the ammonia production flow rate in ammonia synthesis. This is the upper limit of the ammonia production flow rate in ammonia synthesis. For McCormick, it is an auxiliary variable representing the product of weight and ammonia production.

[0122] Based on the above formula, the ammonia production flow rate and the electrical power of the ammonia synthesis section can be uniformly expressed as a linear weighted sum, as shown in the following expression:

[0123]

[0124] (Calculation function of the electrical power model for the ammonia synthesis stage)

[0125]

[0126] in, This refers to the electrical power of the ammonia synthesis section (Haber–Bosch section); is a pre-calculated constant at the breakpoint, used to write the electric power as a linear combination.

[0127] The electrical power model of the ammonia synthesis section is used to describe the energy consumption variation of the ammonia synthesis unit under different ammonia synthesis loads. It is linearized by the SOS2 and McCormick methods to form an electrical power expression of the ammonia synthesis section that can be directly called in mixed integer optimization, serving as the basic input for system optimization.

[0128] (3) Based on the linear relationship between air demand and ammonia production in the ammonia synthesis process, the electrical power of the air separation unit can be expressed as:

[0129]

[0130] in, This refers to the electrical power of the air separation device; This represents the power coefficient of the air separation device.

[0131] The electrical power of the air separation unit and the electrical power of the ammonia synthesis section will be superimposed and incorporated into the mixed integer optimization framework to construct a mixed integer programming model.

[0132] (4) The total power of the ammonia synthesis section is obtained by superimposing the power of the ammonia synthesis section and the air separation unit. The calculation formula is as follows:

[0133]

[0134] in, This represents the total electrical power of the ammonia synthesis section.

[0135] (5) Based on the stoichiometric relationship between hydrogen and ammonia, the ammonia synthesis stage is linked to the ammonia production rate using the following formula:

[0136]

[0137] in, Directly supplying flow to the synthesis section for electrolytic hydrogen; This refers to the hydrogen flow rate from the hydrogen storage tank. is the stoichiometric coefficient from hydrogen to ammonia.

[0138] 4. Construct a full-chain hybrid integer programming model for the water electrolysis hydrogen production and ammonia synthesis system. With the goal of minimizing the annualized total cost of the system, perform joint optimization based on four representative months. Then, perform an annualized weighted solution according to the weight of the representative months to obtain the optimal system configuration and operation scheduling results for the whole year.

[0139] (1) Based on the influencing factors during system operation, determine the constraints corresponding to the full-chain mixed integer programming model; specifically as follows:

[0140] The actual output of wind power cannot exceed the available power. The wind power output constraints are as follows:

[0141]

[0142] in, To actually contribute to wind power; This represents the available power of wind power.

[0143] The actual output of photovoltaic power cannot exceed the available photovoltaic power at that time. The photovoltaic output constraints are as follows:

[0144]

[0145] in, To contribute to the actual development of photovoltaics; This is the usable power of photovoltaics.

[0146] The start-up and shutdown constraints for the electrolytic cell are as follows:

[0147]

[0148]

[0149] in, Indicates the electrolytic cell's operating status; express The electrolytic cell is in constant operation status. and They represent Electrolytic cell startup status at all times The electrolytic cell is in constant operation status.

[0150] The upper and lower limits of the electrolytic cell's electrical power are constrained as follows:

[0151]

[0152]

[0153] in, This represents the minimum electrolysis power. This refers to the power of the electrolytic cell; This is the rated power of the electrolytic cell.

[0154] The formula for calculating the electrical power of an electrolytic cell is as follows:

[0155]

[0156] in, This refers to the electrical power of the electrolytic cell; For hydrogen production flow rate; Electricity consumption for hydrogen production.

[0157] The formula for calculating the compressor's electrical power is:

[0158]

[0159] in, This refers to the electrical power of the compressor. This is the compressor power factor.

[0160] The formula for calculating the electrical power of other auxiliary equipment is as follows:

[0161]

[0162] in, For the electrical power of other auxiliary equipment (such as cooling water pumps, circulating fans, or control systems, etc.); To maintain a fixed electrical power after startup; This is the auxiliary power coefficient.

[0163] The charging and discharging power of energy storage is constrained by a power limit, and simultaneous charging and discharging are not allowed. The constraints are as follows:

[0164]

[0165]

[0166]

[0167] in, , These are the charging and discharging power, respectively. These represent the charging and discharging states, respectively. This is the rated power of the energy storage.

[0168] Energy storage SOC constraints are as follows:

[0169]

[0170]

[0171]

[0172]

[0173] in, This represents the actual energy storage capacity. This represents the minimum energy storage capacity. This refers to the rated capacity of the energy storage. For efficiency.

[0174] The constraints of the hydrogen storage tank are as follows:

[0175]

[0176]

[0177]

[0178]

[0179] in, This represents the lower limit of hydrogen storage energy. Energy for hydrogen storage tanks; The rated capacity of the hydrogen storage tank; This refers to the calorific value of hydrogen. The flow rate of hydrogen entering the hydrogen storage tank; This refers to the hydrogen flow rate from the hydrogen storage tank.

[0180] The upper and lower limits of ammonia synthesis flow rate are constrained as follows:

[0181]

[0182] in, This is the rated flow rate for ammonia synthesis. This represents the flow rate for ammonia synthesis.

[0183] The power balance constraint is:

[0184]

[0185] in, To actually contribute to wind power; To contribute to the actual development of photovoltaics; This refers to the energy storage discharge power; This refers to the power of the electrolytic cell; Power for energy storage charging; To actually meet the electrical load; For auxiliary equipment power; This refers to the electrical power of the compressor. This represents the total electrical power of the ammonia synthesis section.

[0186] (2) With the goal of minimizing the annualized total cost, a full-chain (integrated) mixed integer linear programming model for system configuration and scheduling is established.

[0187] An annualized total cost objective function is constructed with the goal of minimizing the annualized total cost. This function is then used as the objective function of a mixed-integer linear programming model to comprehensively optimize resource allocation and operational coordination, thereby achieving optimal annual economic performance of the system.

[0188] The objective function is shown in the following formula:

[0189]

[0190] in, This represents the annualized total cost target value. The cost of curtailing wind power; The cost of penalties for abandoning light; Penalty costs for abandoned power loads; Annual investment cost; For equipment operation and maintenance costs; The replacement cost of the equipment during its lifespan; For revenue from ammonia sales; To adjust the cost of ammonia synthesis.

[0191] The formula for calculating the cost of wind curtailment penalty is as follows:

[0192]

[0193] in, This represents the wind curtailment penalty coefficient.

[0194] The formula for calculating the cost of light skipping penalty is as follows:

[0195]

[0196] in, This represents the penalty coefficient for discarded light.

[0197] The formula for calculating the penalty cost of abandoned power load is as follows:

[0198]

[0199] in, This is the penalty factor for abandoned power load; For electrical load demand; To actually meet the electrical load.

[0200] The formula for calculating annual investment costs is as follows:

[0201]

[0202] in, This is the annualized coefficient; , , , , , , The unit costs are for the electrolyzer, energy storage, hydrogen storage tank, compressor, auxiliary equipment, air separation unit, and ammonia synthesis equipment, respectively. This refers to the rated installed capacity of the compressor; This refers to the rated installed capacity of auxiliary equipment.

[0203] The formula for calculating annual equipment operation and maintenance costs is as follows:

[0204] in, , , , , , , These are the operation and maintenance cost coefficients for electrolyzers, energy storage, hydrogen storage tanks, compressors, auxiliary equipment, air separation, and ammonia synthesis equipment, respectively.

[0205] The formula for calculating equipment (battery) replacement cost is as follows:

[0206]

[0207] in, This represents the battery replacement cost coefficient.

[0208] The formula for calculating revenue from ammonia sales is as follows:

[0209]

[0210] in, This refers to the unit price of ammonia sold.

[0211] The cost of regulating ammonia synthesis consists of the change in ammonia synthesis load amplitude, the frequency of ammonia synthesis regulation, and the penalty for opening the regulation range. Its expression is:

[0212]

[0213] in, To adjust the cost of ammonia synthesis; , These are the penalty coefficients for changes in the ammonia synthesis load amplitude and the penalty coefficients for the number of adjustments, respectively. This indicates whether the ammonia synthesis load has changed; The penalty cost for activating the adjustment zone.

[0214] The mixed-integer linear programming model is the overall model for achieving system optimization in this invention. It unifies the modeling of processes such as water electrolysis for hydrogen production, energy storage, hydrogen storage, air separation, and ammonia synthesis, and aims to minimize the annual total cost of the system. By jointly optimizing on four representative months and performing annual weighted solutions according to the weights of the representative months, the optimal system configuration and operation scheduling results for the whole year can be obtained.

[0215] The annualized total cost objective function incorporates a synthetic ammonia regulation cost term. By penalizing the magnitude of load changes and the number of regulation segments, the optimization process automatically suppresses frequent load fluctuations, reduces the number of regulation segments, and concentrates necessary load changes into at least a few controllable regulation segments, thereby obtaining a stable, sparse, and flexible regulation operation curve that matches wind and solar fluctuations. The synthetic ammonia regulation cost term drives the optimization to automatically select whether to activate regulation segments, and determines the boundaries of regulation segments through relevant constraints. This ensures that each regulation segment maintains a constant ramp rate, while the load in non-regulation segments remains stable, thus forming an executable piecewise linear flexible regulation curve. The objective function introduces a synthetic ammonia regulation cost term, which, by penalizing regulation segment activation events, makes the optimization process tend to generate fewer regulation segments, thereby obtaining a sparse, stable, and engineering-feasible flexible regulation scheme; and together with segment switching event variables and consistency constraints, determines the number, start and end times, and duration of regulation segments.

[0216] A concrete implementation case:

[0217] In this embodiment, the simulation is carried out using wind power, photovoltaic and load data of a certain place in China. A mixed integer optimization model integrating capacity configuration and operation scheduling of the electric hydrogen ammonia system based on intraday flexible regulation is constructed on the MATLAB simulation platform (system), and the optimization calculation of the isolated power system of wind, solar and water electrolysis to produce hydrogen and synthesize ammonia is carried out.

[0218] Specific input conditions: The selection results of typical months and candidate months from the four quadrants obtained based on the year-round data of this location are as follows: Figure 1 As shown, the power output and load diagrams for four typical months are as follows: Figure 2 As shown in the diagram. The electrical power model diagram for the ammonia synthesis stage is as follows. Figure 3 As shown. The electrolyzer adjustment range is 25%~100%; the upper and lower limits of the energy storage SOC are 10%~90%; the energy storage charge and discharge efficiency is 0.95; the upper and lower limits of the hydrogen storage tank are 10%~100%; the synthetic ammonia ramp rate range is -20%~20%; the upper and lower limits of the synthetic ammonia load are 20%~110%; the initial wind power installed capacity is 7MW; the initial photovoltaic installed capacity is 12MW; and the system operation cycle is 20 years.

[0219] The above input parameters are fed into a mixed-integer optimization model integrating capacity configuration and operation scheduling of an electro-hydrogen-ammonia system based on intraday flexible adjustment. The model is solved simultaneously over four representative months and then aggregated annually according to weights. Iterative optimization calculations are performed with the objective of minimizing the annualized total cost, and the optimal system configuration results are output, as shown in Table 1. The operation scheduling results are also output, as shown in Table 1. Figure 4-7 As shown.

[0220] Table 1 Optimal System Configuration Results

[0221]

[0222] From Table 1 and Figure 4-7 As can be seen from the data, compared with traditional methods, the proposed method for optimizing the configuration and operation of the hydro-hydrogen-ammonia system based on intraday flexible regulation in this embodiment, while ensuring the solvability of the optimization problem, controls the wind and solar curtailment rate to within 3%, effectively improves the matching degree of wind and solar output, reduces energy curtailment, and achieves an annual window of 513, significantly reducing fatigue losses caused by frequent changes in ammonia synthesis load. This provides technical support and an optimization path for the rational configuration and safe operation of wind-solar hydrogen-to-ammonia synthesis units in isolated power systems.

[0223] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various modifications or variations within the scope of the claims, which do not affect the essence of the present invention.

Claims

1. A scheduling method for a water electrolysis hydrogen production and ammonia synthesis system with intraday flexible regulation, characterized in that, Integrating wind and solar renewable energy sources into the power system for hydrogen production through water electrolysis and ammonia synthesis involves the following steps: (1) The hourly wind power and photovoltaic data of the whole year are sliced ​​by month. After calculating the monthly wind and solar energy, the data is divided into four quadrants in the Copula rank space. Multi-scale fluctuation features are extracted and the joint distance between months is constructed. A representative month is selected from each of the four quadrants to construct a feature matrix. Weights are assigned to the four selected representative months. (2) Introduce a flexible adjustment mechanism within the time domain of the representative month, and describe the synthetic ammonia load as a continuous piecewise linear operating curve; by solving the annualized total cost objective function, concentrate the synthetic ammonia load change in a few preset controllable sections to match the fluctuation of wind and solar power and reduce the adjustment frequency of synthetic ammonia equipment. (3) Construct a unit ammonia production power consumption function with the per-unit load of synthetic ammonia as the independent variable, and perform equivalent linearization; A linearly solvable electrical power model for the ammonia synthesis stage was established and incorporated into a mixed-integer optimization framework after being superimposed with the electrical power of the air separation unit. (4) Construct a full-chain mixed integer linear programming model for the electrolysis of water to produce hydrogen and synthesize ammonia. With the goal of minimizing the annual total cost of the system, perform joint optimization on four representative months and solve the annual weighted solution according to the weight of the representative month to obtain the optimal system configuration and operation scheduling results for the whole year.

2. The method according to claim 1, characterized in that, In step (1), the wind and solar energy of each month are mapped to the (0,1) interval by rank transformation to form the corresponding Copula coordinates, and a threshold is used to set the (0,1) interval to the (0,1) interval. , The 12 months are divided into four quadrants: strong wind and strong light, strong wind and weak light, weak wind and strong light, and weak wind and weak light; when the wind energy rank or light energy rank If the value is not lower than the threshold, it is judged as strong wind or strong light respectively; otherwise, it is weak wind or weak light. If any quadrant is empty, the corresponding threshold is shrunk to 0.5 until all four quadrants are not empty or the threshold returns to 0.

5. The weight of each representative month is equal to its proportion of the month in its respective quadrant. Parallel ranks are processed by the average method.

3. The method according to claim 1, characterized in that, In step (1), the multi-scale fluctuation characteristics refer to the multi-dimensional characteristics constructed for each month's wind power and photovoltaic data, specifically including the intensity of wind and solar fluctuations, the intraday amplitude of wind and solar power, and the intraday fundamental frequency energy ratio.

4. The method according to claim 1, characterized in that, In step (1), for any month The joint distance is defined as the weighted Euclidean sum of the differences in wind power quantile curve vectors, photovoltaic quantile curve vectors, and multi-scale eigenvectors, and is used to simultaneously characterize the distribution pattern and temporal fluctuations of wind and solar power output. The joint distance is calculated using the following formula: ； in, Indicates month With Month The joint distance, the smaller the value, the more similar the two months are; by month For example, , These are the equidistant quantile curve vectors for wind power and photovoltaic power, respectively. These are multi-scale feature row vectors; It is the Euclidean norm; For weights.

5. The method according to claim 1, characterized in that, In step (1), the representative month for constructing the feature matrix is ​​selected using the constrained D-optimal criterion, specifically including: a. Calculate the medoid score within each group for the monthly set in each quadrant, then sort them in ascending order and take the top K to form a candidate set; b. Select one month from the candidate sets of each of the four quadrants to form candidate combinations, and construct the feature matrix. ; c. Using the constrained D-optimal criterion as the objective, calculate the value of the objective function for each candidate combination, and select the combination that maximizes the objective function. If multiple combinations achieve the same maximum objective function value, prioritize the combination with the smallest sum of medoid scores for each month among the selected month combinations. The objective function for selecting the representative month is: ； in, Choose an objective function to represent the month; It refers to the determinant of a matrix; Refers to the candidate month feature matrix Gram matrix; These are the diagonal regularity coefficients; It is an identity matrix.

6. The method according to claim 1, characterized in that, In step (2), the linear operating curve is divided into a stable operating section and a preset flexible adjustment section. In the flexible adjustment section, the ammonia synthesis load changes at a constant ramp rate, while the others are stable operating sections where the load remains unchanged. The number, start and end times, and duration of the flexible adjustment sections are automatically determined by the optimization solution process of the annualized total cost objective function. The flexible adjustment modeling constraints corresponding to the annualized total cost objective function include: a. The hourly variation of ammonia synthesis load is related to the maximum ammonia synthesis yield and the ramp-up rate; b. Flexible adjustment of interval states can be described by activating indicator variables; c. Set the daily 24-hour ramp rate to zero, and apply a condition that represents the first and last months of the month being equal in value.

7. The method according to claim 1, characterized in that, In step (2), to ensure the executability of the flexible adjustment section operation and obtain a sparse solution, a logical consistency constraint is established between the flexible adjustment section state indicator variable and the section state switching event variable to ensure that the synthetic ammonia load ramp rate in each flexible adjustment section remains constant, and a penalty term for load changes and the number of flexible adjustment sections is introduced into the annualized total cost objective function; specifically, the following content is included: a. Accurately describe the start and end changes of the flexible adjustment section, and keep the event variables and status indicators logically consistent; b. Apply a consistent ramp rate constraint between adjacent time points to ensure that the ramp rate of ammonia synthesis load is constant within the same flexible adjustment section and to achieve a continuous and smooth transition of ammonia synthesis load inside and outside the section; c. Introduce a penalty cost for enabling flexible adjustment segments into the objective function to suppress ineffective adjustment and constrain the number of flexible adjustment segments; d. The number, start time, and duration of the flexible adjustment segments are all determined by the optimization solution, representing the adjustment scheme in the optimization results.

8. The method according to claim 1, characterized in that, Step (3) specifically includes: a. Construct an electrical power model for the ammonia synthesis section, in which a unit power consumption function with the per-unit load level of ammonia synthesis as the independent variable is used; b. The unit power consumption function is equivalently linearized by using SOS2 convex combination on the set of breakpoints and processing the product terms with McCormick envelope to obtain the linearized expression of the power of ammonia synthesis. c. Based on the linear relationship between air demand and ammonia production in the ammonia synthesis process, construct the power function of the air separation unit; d. The total electrical power of the ammonia synthesis section is obtained by superimposing the electrical power of the ammonia synthesis section and the air separation unit; e. Based on the stoichiometric relationship between hydrogen and ammonia, establish the connection between the amount of hydrogen used in the ammonia synthesis stage and the amount of ammonia produced.

9. The method according to claim 1, characterized in that, In step (4), a full-chain hybrid integer programming model is used to achieve integrated optimization of system configuration and scheduling; wherein, a. Based on the influencing factors during system operation, determine the constraints corresponding to the full-chain mixed integer programming model; the constraints include: wind power output constraints, photovoltaic power output constraints, electrolyzer start-up and shutdown constraints, electrolyzer power upper and lower limit constraints, energy storage charging and discharging power constraints, energy storage SOC constraints, hydrogen storage tank constraints, synthetic ammonia flow upper and lower limit constraints, and power balance constraints. b. Construct an annualized total cost objective function with the goal of minimizing the annualized total cost, and use it as the objective function of the mixed-integer linear programming model, as shown in the following formula: ； in, This represents the annualized total cost target value. The cost of curtailing wind power; The cost of penalties for abandoning light; Penalty costs for abandoned power loads; Annual investment cost; For equipment operation and maintenance costs; The replacement cost of the equipment during its lifespan; For revenue from ammonia sales; To adjust the cost of ammonia synthesis.

Images