Hydrogen network optimization method and device, computer equipment and storage medium

Abstract

This application relates to a hydrogen network optimization method, apparatus, computer equipment, and storage medium. The hydrogen network optimization method includes: acquiring parameter information of a pressure swing adsorption (PSA) unit and constructing a PSA mechanism model based on the parameter information; generating multiple sets of input-output parameter sets according to the PSA mechanism model; each set of input-output parameter sets includes the input parameters and corresponding output parameters of the PSA mechanism model; using each set of input-output parameter sets as constraints, optimizing the hydrogen network according to the hydrogen network optimization model to obtain the hydrogen network and economic indicators corresponding to each set of input-output parameter sets; constructing a sample set based on the hydrogen network and economic indicators corresponding to each set of input-output parameter sets; iteratively optimizing the sample set using a Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter set, thereby optimizing the hydrogen network, improving the reliability of the hydrogen network, and thus improving the utilization rate of hydrogen.

Description

Technical Field

[0001] This application relates to the field of gas distribution network optimization technology, and in particular to a hydrogen network optimization method, apparatus, computer equipment, and storage medium. Background Technology

[0002] With the booming development of my country's economy, oil consumption is continuously rising, while refining capacity is also expanding rapidly. However, during this capacity expansion, the degree of heavy and inferior quality of crude oil is constantly increasing. To improve crude oil quality, refineries need more hydrogenation processes to handle crude oil, thus increasing the demand for hydrogen. Therefore, hydrogen consumption in the refining process has become a major challenge for refineries. Thus, optimizing the design of the refinery's hydrogen network to utilize hydrogen resources rationally and efficiently is of great significance for achieving energy conservation, emission reduction, and cost control in the refining process.

[0003] Currently, to improve hydrogen purity, pressure swing adsorption (PSA) units are commonly used to purify low-purity feedstocks, meeting the demand for high-purity hydrogen from hydrogen-consuming devices and further improving hydrogen utilization efficiency. This also helps alleviate the pressure of hydrogen demand shortfalls. In related technologies, surrogate models are typically used to simulate the PSA process during hydrogen network optimization for PSA units. However, in practical applications, the surrogate models in these technologies have significant errors, severely affecting hydrogen network matching and the determination of optimal PSA unit parameters, ultimately leading to low reliability of the hydrogen network and thus low hydrogen utilization.

[0004] There is currently no effective solution to the problem of low reliability of hydrogen networks and low utilization rate of hydrogen in existing technologies. Summary of the Invention

[0005] Therefore, it is necessary to provide a hydrogen network optimization method, apparatus, computer equipment, and storage medium to address the aforementioned technical problems.

[0006] In a first aspect, this application provides a hydrogen network optimization method, the method comprising:

[0007] Obtain the parameter information of the pressure swing adsorption unit, and construct a pressure swing adsorption mechanism model based on the parameter information;

[0008] Based on the pressure swing adsorption mechanism model, multiple sets of input and output parameters are generated; each set of input and output parameters includes the input parameters of the pressure swing adsorption mechanism model and the corresponding output parameters.

[0009] Each set of input and output parameters is used as a constraint condition, and the hydrogen network is optimized according to the hydrogen network optimization model to obtain the hydrogen network and economic indicators corresponding to each set of input and output parameters.

[0010] A sample set is constructed based on the hydrogen network and economic indicators corresponding to each set of input and output parameters.

[0011] The sample set is iteratively optimized using the Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter set.

[0012] In one embodiment, generating multiple sets of input-output parameters based on the pressure swing adsorption mechanism model includes:

[0013] Based on the input parameter range of the pressure swing adsorption mechanism model, multiple sets of input parameters are selected; the input parameters include equipment parameters and operating parameters.

[0014] Multiple sets of input parameters are input into the pressure swing adsorption mechanism model to obtain corresponding output parameters; wherein, the output parameters include gas flow rate and gas concentration;

[0015] Multiple sets of input and output parameters are generated based on the multiple sets of input parameters and the corresponding output parameters.

[0016] In one embodiment, virtual hydrogen streams and / or new hydrogen streams are added to the hydrogen source corresponding to the hydrogen network optimized by the hydrogen network optimization model.

[0017] In one embodiment, the hydrogen network optimization model includes: hydrogen source / hydrogen trap constraints, pressure swing adsorption unit constraints, gas system constraints, and objective function constraints.

[0018] In one embodiment, the step of iteratively optimizing the sample set according to the Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter set includes:

[0019] The sample set is iteratively optimized using the Bayesian optimization algorithm, and the results of each iteration are added to the sample set; the iteration results include: the iterative input and output parameter set, as well as the corresponding hydrogen network and economic indicators;

[0020] If the number of iterations reaches the preset number, the iteration stops, and the updated sample set is obtained;

[0021] From the updated sample set, the input-output parameter set and hydrogen network corresponding to the minimum economic indicator are selected as the target input-output parameter set and target hydrogen network.

[0022] In one embodiment, when the Bayesian optimization algorithm iteratively optimizes the sample set, it uses a surrogate function to fit the relationship between the input-output parameter set and economic indicators.

[0023] In one embodiment, the method further includes:

[0024] Determine whether a virtual hydrogen stream and / or a new hydrogen stream have been added to the hydrogen source corresponding to the target hydrogen network;

[0025] If added, the target input / output parameter set and the target hydrogen network will be optimized.

[0026] Secondly, this application also provides a hydrogen network optimization device, the device comprising:

[0027] The acquisition module is used to acquire parameter information of the pressure swing adsorption unit and construct a pressure swing adsorption mechanism model based on the parameter information;

[0028] The generation module is used to generate multiple sets of input and output parameter groups based on the pressure swing adsorption mechanism model; each set of input and output parameter groups includes the input parameters of the pressure swing adsorption mechanism model and the corresponding output parameters.

[0029] The hydrogen network optimization module is used to optimize the hydrogen network according to the hydrogen network optimization model by taking each set of input and output parameters as constraints, and to obtain the hydrogen network and economic indicators corresponding to each set of input and output parameters.

[0030] The construction module is used to construct a sample set based on the hydrogen network and economic indicators corresponding to each set of input and output parameters.

[0031] The iterative optimization module performs iterative optimization on the sample set according to the Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter set.

[0032] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described in any of the embodiments of the first aspect above.

[0033] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the embodiments of the first aspect above.

[0034] The aforementioned hydrogen network optimization method, apparatus, computer equipment, and storage medium first acquire parameter information of the pressure swing adsorption (PSA) unit and construct a PSA mechanism model based on this information. Then, based on the PSA mechanism model, multiple sets of input-output parameter groups are generated. Each set includes the input parameters and corresponding output parameters of the PSA mechanism model. Each set of input-output parameter groups is used as a constraint condition, and the hydrogen network is optimized according to the hydrogen network optimization model to obtain the hydrogen network and economic indicators corresponding to each set of input-output parameter groups. A sample set is constructed based on the hydrogen network and economic indicators corresponding to each set of input-output parameter groups. The sample set is iteratively optimized using a Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter group. Based on the PSA mechanism model and the hydrogen network optimization model, the hydrogen network is optimized, improving its reliability, thereby increasing hydrogen utilization and reducing production costs. Attached Figure Description

[0035] The accompanying drawings, which are included to provide a further understanding of this application, form part of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation thereof. In the drawings:

[0036] Figure 1 This is a flowchart illustrating a hydrogen network optimization method in one embodiment;

[0037] Figure 2 This is a flowchart illustrating the hydrogen network optimization method in another embodiment;

[0038] Figure 3 This is a structural block diagram of a hydrogen network optimization device in one embodiment;

[0039] Figure 4 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0041] Unless otherwise defined, the technical or scientific terms used in this application shall have the general meaning as understood by one of ordinary skill in the art to which this application pertains. Words such as “a,” “an,” “an,” “the,” “the,” and “these,” used in this application, do not indicate quantitative limitation and may be singular or plural. The terms “comprising,” “including,” “having,” and any variations thereof used in this application are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or device that comprises a series of steps or modules (units) is not limited to the listed steps or modules (units) but may include steps or modules (units) not listed, or may include other steps or modules (units) inherent to such processes, methods, products, or devices. The terms “connected,” “linked,” and “coupled,” used in this application, are not limited to physical or mechanical connections but may include electrical connections, whether direct or indirect. The term “multiple” used in this application refers to two or more. The "and / or" operator describes the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: A alone, A and B simultaneously, and B alone. Typically, the character " / " indicates that the objects before and after it are in an "or" relationship. The terms "first," "second," and "third," etc., used in this application are merely for distinguishing similar objects and do not represent a specific ordering of the objects.

[0042] In one embodiment, such as Figure 1 As shown, Figure 1 This is a flowchart illustrating a hydrogen network optimization method in one embodiment. The hydrogen network optimization method includes the following steps:

[0043] Step S101: Obtain the parameter information of the pressure swing adsorption unit and construct a pressure swing adsorption mechanism model based on the parameter information.

[0044] The parameter information includes the input parameters of the pressure swing adsorption (PSA) mechanism model and gas characteristic information. The input parameters of the PSA mechanism model are the decision variables of the PSA mechanism model. The input parameters of the PSA mechanism model, i.e., the decision variables, determine the processing capacity, product concentration, and cost of the PSA unit. The gas characteristic information includes hydrogen adsorption characteristics and the adsorption characteristics of other impurities.

[0045] The pressure swing adsorption (PSA) mechanism model refers to the dynamic model of the two-tower, six-step PSA, serving as a proxy model for the PSA unit to simulate the PSA process. The two-tower, six-step model refers to the PSA unit comprising two adsorption towers, adsorption tower 1 and adsorption tower 2. Each cycle unit has six steps: adsorption, pressure equalization and reduction, pressure reduction, purging, pressure equalization and increase, and pressure increase. In the adsorption step, the feed gas is pressurized to the adsorption pressure p1 and then delivered to adsorption tower 1. The outlet is a high-purity hydrogen stream, part of which is sent to adsorption tower 2 as purging gas, and the remaining hydrogen stream flows out as product. At the end of the adsorption step, the adsorption capacity of adsorption tower 1 is close to saturation, and adsorption towers 1 and 2 are at their highest and lowest pressure states, respectively. To effectively utilize the pressure difference between the two towers and reduce the pressure in adsorption tower 1 for adsorbent regeneration, the gas in adsorption tower 1 is passed to adsorption tower 2, causing the pressure in adsorption tower 1 to drop to p2 and the pressure in adsorption tower 2 to rise to p3. This process is the pressure equalization and reduction step. In the subsequent depressurization step, the pressure in adsorption tower 1 is further reduced to p4, at which point the pressure inside adsorption tower 1 is close to atmospheric pressure. In the purging step, a high-purity hydrogen stream from adsorption tower 2 is introduced into adsorption tower 1, further desorbing impurities. During this process, the pressure in adsorption tower 1 is maintained at the minimum pressure p4. Then, a pressure equalization step is performed, which is the reverse of the depressurization step. Gas from adsorption tower 2 flows into adsorption tower 1, raising the pressure in adsorption tower 1 to p3. In the pressure increase step, adsorption tower 1 is further pressurized from the feed gas to the adsorption pressure, thus completing the entire cycle. It should be noted that the pressure swing adsorption mechanism model in this embodiment can be built using, but is not limited to, Aspen Adsorption software or languages ​​such as Matlab, Python, C++, C, and VB.

[0046] Specifically, based on the flow rate and concentration range of the purification stream in the hydrogen network, the input parameter range of the pressure swing adsorption (PSA) unit, i.e., the upper and lower limits of the decision variables, is determined. Based on this, a PSA mechanism model is established. It should be noted that the input parameter range, i.e., the upper and lower limits of the decision variables, needs to be set based on empirical values; this implementation does not impose specific limitations.

[0047] Step S102: Based on the pressure swing adsorption mechanism model, generate multiple sets of input and output parameter groups; each set of input and output parameter groups includes the input parameters of the pressure swing adsorption mechanism model and the corresponding output parameters.

[0048] The input and output parameter set includes the input parameters and corresponding output parameters of the pressure swing adsorption mechanism model.

[0049] Specifically, the input parameters of multiple pressure swing adsorption (PSA) mechanism models are input into the PSA mechanism models to obtain the corresponding output parameters of the multiple PSA mechanism models. The input parameters and corresponding output parameters of each PSA mechanism model are defined as input-output parameter sets, thereby generating multiple input-output parameter sets.

[0050] Step S103: Each set of input and output parameters is used as a constraint condition, and the hydrogen network is optimized according to the hydrogen network optimization model to obtain the hydrogen network and economic indicators corresponding to each set of input and output parameters.

[0051] The hydrogen network refers to the hydrogen quality exchange network formed by the hydrogen supply and consumption streams in a refinery, reflecting the layout of hydrogen streams. The hydrogen network includes the matching relationships between different hydrogen sources and hydrogen traps, i.e., the flow rate from hydrogen source to hydrogen trap. A hydrogen source refers to the unused, recyclable hydrogen stream contained in the outlet stream of a hydrogen production unit or the outlet stream of a hydrogen consumption unit; a hydrogen trap refers to the inlet stream of a hydrogen consumption unit. It should be noted that hydrogen network system integration methods include graphical methods (i.e., pinch methods) and mathematical programming methods. Graphical methods can quickly and clearly show the pinch locations and determine the minimum utility hydrogen consumption, but they can handle fewer variables, have a single objective function, and cannot solve complex problems with multiple variables. Mathematical programming methods can consider complex constraints and optimize the hydrogen network with total cost as the objective, determining the optimal hydrogen network distribution.

[0052] Among them, the economic indicators include at least the total annual cost, which refers to all the costs required to build and operate the hydrogen network within a single year.

[0053] Specifically, before establishing the hydrogen network optimization model, it is necessary to first determine the flow rate and stream component concentration of the hydrogen source participating in the optimization, as well as the required flow rate and component concentration of the hydrogen traps. Further, using the superstructure optimization method, a hydrogen network optimization model is established. Each set of input and output parameters is used as a constraint condition for the hydrogen network optimization model, and the model is then optimized to obtain the hydrogen network corresponding to each set of input and output parameters, as well as the corresponding economic indicator, i.e., the total annual cost. It should be noted that the hydrogen network optimization model in this embodiment can be implemented using, but is not limited to, GAMS software, Python, Matlab, C++, C, and VB languages.

[0054] Step S104: Construct a sample set based on the hydrogen network and economic indicators corresponding to each set of input and output parameters.

[0055] Specifically, in the process of optimizing the hydrogen network using the hydrogen network optimization model, the hydrogen network, economic indicators, and corresponding input parameters are added to the sample set according to the hydrogen network and economic indicators (i.e., the total annual cost) for each set of input and output parameters, and the sample set is updated and expanded.

[0056] Step S105: Iteratively optimize the sample set using the Bayesian optimization algorithm to obtain the target hydrogen network and the target input / output parameter set.

[0057] Among them, Bayesian optimization algorithm refers to optimization method based on Bayesian statistics and Gaussian process, which is used to solve the optimization problem of black box function; Bayesian optimization algorithm gradually approaches the optimal solution by continuously selecting the next sample point for evaluation.

[0058] Among them, the target hydrogen network refers to the optimal hydrogen network; the target input-output parameter set refers to the input-output parameter set corresponding to the minimum economic indicator, that is, the input parameters of the pressure swing adsorption mechanism model corresponding to the minimum annual total cost, and the output parameters corresponding to the input parameters.

[0059] Specifically, based on the Bayesian optimization algorithm, the sample set is iteratively optimized to obtain the target hydrogen network, i.e. the optimal hydrogen network, and the target input-output parameter set is selected, i.e. the input-output parameter set corresponding to the minimum economic index, i.e. the input parameters of the pressure swing adsorption mechanism model corresponding to the minimum annual total cost, and the output parameters corresponding to the input parameters.

[0060] In this embodiment, based on the pressure swing adsorption (PSA) mechanism model, the PSA process is simulated to generate multiple sets of input-output parameters. Then, based on the hydrogen network optimization model, the economic indicators corresponding to each set of input-output parameters are obtained to update and expand the sample set, laying a data foundation for improving the reliability of the hydrogen network. Furthermore, iterative optimization is performed based on the Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter set, thereby optimizing the hydrogen network, improving its reliability, increasing hydrogen utilization, and reducing production costs.

[0061] In one embodiment, generating multiple sets of input-output parameters based on the pressure swing adsorption mechanism model includes the following steps:

[0062] Step 1: Select multiple sets of input parameters based on the input parameter range of the pressure swing adsorption mechanism model; the input parameters include equipment parameters and operating parameters.

[0063] Step 2: Input multiple sets of input parameters into the pressure swing adsorption mechanism model to obtain the corresponding output parameters; among which, the output parameters include gas flow rate and gas concentration.

[0064] Step 3: Generate multiple sets of input and output parameter groups based on multiple sets of input parameters and their corresponding output parameters.

[0065] The input parameters include equipment parameters and operating parameters. Equipment parameters refer to those related to the pressure swing adsorption (PSA) unit, including at least the number of adsorption beds, bed dimensions, and adsorbent loading. The number of adsorption beds ranges from 1 to 5. Bed dimensions include bed length and diameter, with bed length ranging from 1 to 10 m and bed diameter from 0.2 to 4 m. The adsorbent loading ranges from 0.1 to 125 m³. 3 Operating parameters refer to the parameters related to the adsorption step, pressure equalization step, pressure reduction step, purging step, pressure equalization step, and pressure increase step, including at least the molar flow rate, pressure, temperature, and operating time for each step; wherein, the molar flow rate ranges from 1 to 55 mol / s; the pressure ranges from 0 to 60 bar; the temperature ranges from 20 to 40 °C; and the operating time ranges from 10 to 120 s. It should be noted that molar flow rate and operating time are negatively correlated with purification efficiency, while pressure is positively correlated with purification efficiency.

[0066] The output parameters include gas flow rate and gas concentration; the gas flow rate includes at least the average flow rate of the hydrogen stream at the outlet of the pressure swing adsorption unit; the gas concentration includes at least the concentrations of hydrogen, methane, carbon dioxide, and carbon monoxide.

[0067] Specifically, based on the input parameter range of the pressure swing adsorption (PSA) mechanism model, i.e., the upper and lower limits of the decision variables, multiple sets of input parameters are selected, namely equipment parameters and / or operating parameters. These multiple sets of input parameters are input into the PSA mechanism model to obtain the output parameters corresponding to the input parameters, namely the average flow rate of the hydrogen stream at the outlet of the PSA unit and the concentrations of hydrogen, methane, carbon dioxide, and carbon monoxide. Each input parameter and the output parameter corresponding to the current input parameter are regarded as a set of input-output parameters, and multiple sets of input-output parameter sets are generated accordingly. It should be noted that since the PSA mechanism model is a two-tower, six-step dynamic model, each set of input parameters needs to be run for 10 cycles to allow the PSA mechanism model to reach a stable state. Finally, the output parameter of the 10th running cycle is taken as the output parameter corresponding to that set of input parameters. Each running cycle includes six steps: adsorption step, equalization and depressurization step, depressurization step, purging step, equalization and depressurization step, and depressurization step.

[0068] In this embodiment, multiple sets of input parameters are input into the pressure swing adsorption mechanism model to obtain multiple sets of corresponding output parameters, thereby generating multiple sets of input and output parameter sets. This lays the foundation for the optimization of the hydrogen network, and further lays the foundation for improving the reliability of the hydrogen network and the utilization rate of hydrogen.

[0069] In one embodiment, virtual hydrogen streams and / or new hydrogen streams are added to the hydrogen source corresponding to the hydrogen network optimized by the hydrogen network optimization model.

[0070] Among them, the virtual hydrogen stream refers to the hydrogen stream with a hydrogen concentration of 0%, denoted as SR0; the new hydrogen stream refers to the hydrogen stream that originally exists in the hydrogen network, with a concentration range of 80%-100%, which is used to work with the virtual hydrogen stream SR0 to formulate the hydrogen source corresponding to the concentration and flow rate required by the pressure swing adsorption unit.

[0071] Specifically, by adding virtual hydrogen streams SR0 and / or new hydrogen streams to the hydrogen source corresponding to the hydrogen network involved in the optimization, the complex constraints between the inlet hydrogen stream of the pressure swing adsorption (PSA) unit and the hydrogen network can be eliminated, and infeasible points in the input parameter space of the PSA mechanism model can be converted into feasible points. Here, infeasible and feasible points are relative concepts. A set of input parameters is input into the PSA mechanism model to obtain corresponding output parameters, thereby determining the inlet and outlet hydrogen stream data of the PSA unit. When coupling the inlet and outlet hydrogen stream data with the hydrogen network, if the hydrogen source in the hydrogen network cannot meet the concentration and flow rate requirements of the PSA unit inlet for the hydrogen stream, the input parameters are considered infeasible points, requiring the addition of virtual hydrogen streams SR0 and / or new hydrogen streams to convert them into feasible points. If the hydrogen source in the hydrogen network meets the concentration and flow rate requirements of the PSA unit inlet for the hydrogen stream, the input parameters are considered feasible points, and the conversion using virtual hydrogen streams SR0 and / or new hydrogen streams is not necessary.

[0072] In this embodiment, by adding virtual hydrogen streams SRO and / or new hydrogen streams to the hydrogen sources corresponding to the hydrogen network participating in the optimization, the infeasible solutions generated during the hydrogen network optimization process are avoided from affecting the hydrogen network optimization results, thereby improving the reliability of the hydrogen network optimization model and the efficiency of hydrogen network optimization.

[0073] In one embodiment, the hydrogen network optimization model includes: hydrogen source / hydrogen trap constraints, pressure swing adsorption unit constraints, gas system constraints, and objective function constraints.

[0074] The hydrogen source / hydrogen trap constraints are shown in equations (1)-(3), where equations (1) and (2) represent the lower limits of the hydrogen trap's flow rate and the amount of pure hydrogen; equation (3) represents the upper and lower limits of the hydrogen source's usage. In equations (1)-(3), i represents the hydrogen source; j represents the hydrogen trap; and F represents the flow rate in mol·s⁻¹. -1 ;c represents concentration;F i,j This represents the flow rate from i to j, i.e., the flow rate from the hydrogen source to the hydrogen trap, expressed in mol·s⁻¹. -1 min and max represent the upper and lower limits of the hydrogen source flow rate, respectively.

[0075] ∑ i Fi,j ≥F j (1)

[0076] ∑ i F i,j ·c i ≥F j ·c j (2)

[0077] Fmin i ≤F i ≤Fmax i (3)

[0078] The constraints of the pressure swing adsorption (PSA) unit are shown in equations (4) and (5), where equations (4) and (5) indicate that the amount and flow rate of pure hydrogen supplied to the PSA unit must match the requirements. In equations (4) and (5), PSA represents the pressure swing adsorption unit; F i,PSA This represents the flow rate from hydrogen source i to the pressure swing adsorption unit, expressed in mol·s⁻¹. -1 ;in indicates the inlet of the pressure swing adsorption unit.

[0079] ∑ i F i,PSA =F in (4)

[0080] ∑ i F i,PsA ·c i =F in ·c in (5)

[0081] The gas system constraints are shown in equations (6) and (7), where equations (6) and (7) indicate that unused hydrogen source and pressure swing adsorption unit exhaust gas is introduced into the gas system. In equations (6) and (7), FGS represents the gas system; F i,FGS This represents the flow rate from hydrogen source i to the gas combustion system, expressed in mol·s⁻¹. -1 ;ex represents the exhaust gas discharged from the pressure swing adsorption unit.

[0082] ∑ i F i,FGS +F ex =F FGS (6)

[0083] F i,FGS =F i -∑ j F i,j (7)

[0084] The objective function constraints are shown in equations (8)-(14), where equation (8) represents the total annual cost including hydrogen cost and pressure swing adsorption (PSA) unit cost; equation (9) represents the hydrogen cost calculation equation; and equation (10) represents the PSA unit cost including investment cost and operating cost, which can be calculated from equations (11)-(14). In equations (8)-(14), TAC represents the total annual cost in M$; AC HP This indicates the cost of new hydrogen, in M$; AC PSA The following figures represent the costs of the pressure swing adsorption (PSA) unit, in M$; ACC represents the investment cost, in M$; AOC represents the operating cost, in M$; TPC represents the total equipment cost of the PSA unit, in M$; BMC represents the purchase cost, in M$; B, cool, com, and motor represent the adsorption tower, compressor, engine, and cooler, respectively; CA represents the adsorbent cost, in M$; UC represents the unit price of the adsorbent, in M$; and ρ represents the adsorbent density, in kJ / m³. -3 ε represents the adsorbent porosity; N is the number of adsorption towers; V is the volume in cubic meters. 3 .

[0085] TAC = AC HP +AC PSA (8)

[0086] AC HP = 8760 × 3600 × 2.37 × 10 -3 ×F HP (9)

[0087] AC PSA =ACC+AOC (10)

[0088] ACC = 0.0815 × TPC (11)

[0089] TPC = 1.716 × (BMC) B +BMC cool +BMC com +BMC motor (12)

[0090]

[0091] AOC = CA + 613.2HP com +0.055×TPC (14)

[0092] Furthermore, the hydrogen network optimization model also includes decision variables and objective function; among them, the decision variable of the hydrogen network optimization model is the hydrogen source-hydrogen trap matching flow rate, that is, the hydrogen flow stream exchange flow rate between equipment; the objective function of the hydrogen network optimization model is the economic indicator, that is, the total annual cost.

[0093] In this embodiment, the constraints of hydrogen source / hydrogen trap, pressure swing adsorption unit, gas system, and objective function are used as constraints of the hydrogen network optimization model. The matching flow rate of hydrogen source and hydrogen trap is used as the decision variable of the hydrogen network optimization model, and the economic index is used as the objective function of the hydrogen network optimization model. This lays the foundation for establishing an accurate, feasible, and effective hydrogen network optimization model.

[0094] In one embodiment, iteratively optimizing the sample set using a Bayesian optimization algorithm to obtain the target hydrogen network and the target input / output parameter set includes the following steps:

[0095] Step 1: Iteratively optimize the sample set using the Bayesian optimization algorithm, and add the results of each iteration to the sample set; the iteration results include: the iterative input and output parameter sets, as well as the corresponding hydrogen network and economic indicators.

[0096] Step 2: If the number of iterations reaches the preset number, stop the iteration and obtain the updated sample set.

[0097] Step 3: From the updated sample set, select the input-output parameter set and hydrogen network corresponding to the minimum economic indicator as the target input-output parameter set and target hydrogen network.

[0098] Preferably, when the Bayesian optimization algorithm iteratively optimizes the sample set, a surrogate function is used to fit the relationship between the input and output parameter sets and economic indicators.

[0099] Specifically, based on the Bayesian optimization algorithm, a surrogate function is used to fit the relationship between the economic indicator (i.e., the total annual cost) and the input parameters of the pressure swing adsorption (PSA) mechanism model, obtaining the corresponding fitting result. Based on the fitting result, the acquisition function is used to determine the next iteration point, i.e., the next set of input parameters. The surrogate function can be, but is not limited to, a Gaussian process function or a neural network model; preferably, this embodiment uses a Gaussian process function as the surrogate function. The acquisition function can be, but is not limited to, the UCB (Upper Confidence Bound) function, the EI (Expected Improvement) function, or the PI (Probability of Improvement) function. The UCB function can balance the relationship between exploring unknown regions and optimizing local regions, but its effect is sensitive to the parameter k; the EI function has fewer parameters and can balance the relationship between exploring unknown regions and optimizing local regions, but when using gradient-based methods, derivative information needs to be derived; the PI function is simple and easy to derive, but it only reflects the probability of improvement, not the amount of improvement. Preferably, in this embodiment, the UCB function is used as the acquisition function, and the maximum value of the UCB function in the entire parameter space is taken as the next iteration point; the UCB function is shown in equation (15), where μ is the mean of the surrogate function, σ is the standard deviation of the surrogate function, and k is the UCB adjustment parameter; where the surrogate function refers to the Gaussian process function.

[0100] UCB(x)=μ(x)+κσ(x) (15)

[0101] Specifically, based on the Bayesian optimization algorithm, a surrogate function, namely a Gaussian process function, is constructed. A UCB (Unified Input-Output) function is used to select the next iteration point (the next set of input parameters) from the existing sample set for iterative optimization, yielding the corresponding iterative results: the iterative input-output parameter set, the corresponding hydrogen network, and economic indicators. Further, it is determined whether the preset number of iterations has been reached. If the preset number of iterations has been reached, the iteration stops, and an updated sample set is obtained. From the updated sample set, the input-output parameter set and hydrogen network corresponding to the minimum economic indicator are selected as the target input-output parameter set and target hydrogen network. If the preset number of iterations has not been reached, the results of this iteration (the iterative input-output parameter set, the corresponding hydrogen network, and economic indicators) are added to the existing sample set to obtain an updated sample set. The process of selecting the next iteration point based on the Bayesian optimization algorithm is repeated, and this cycle continues until the preset number of iterations is reached. For example, in an existing sample set, a set of input parameters for a pressure swing adsorption (PSA) mechanism model is selected. The corresponding output parameters are obtained through the PSA mechanism model, and the corresponding input-output parameter set is determined. This set of input-output parameter sets is then substituted into a hydrogen network optimization model to obtain the corresponding hydrogen network and economic indicators. This process is repeated continuously, and the sample set is updated until the preset number of iterations is reached. From the updated sample set, the input-output parameter set and hydrogen network corresponding to the minimum economic indicator are selected as the target input-output parameter set and target hydrogen network. Here, the target hydrogen network refers to the optimal hydrogen network; the target input-output parameter set refers to the input parameters of the PSA mechanism model corresponding to the minimum economic indicator (i.e., the minimum annual total cost), and the corresponding output parameters.

[0102] In this embodiment, a surrogate function is constructed to fit the relationship between economic indicators and input parameters to obtain the corresponding fitting result. Then, based on the fitting result, an acquisition function is used to determine the next iteration point, and multiple iterations are performed to optimize and continuously update the sample set. Finally, the input and output parameter set corresponding to the minimum economic indicator and the hydrogen network can be selected from the updated sample set. Based on this, the utilization rate of hydrogen is improved and the production cost is reduced.

[0103] In one embodiment, it is determined whether a virtual hydrogen stream and / or a new hydrogen stream have been added to the hydrogen source corresponding to the target hydrogen network;

[0104] If added, the target input / output parameter set and the target hydrogen network will be optimized.

[0105] Specifically, to avoid generating infeasible solutions, virtual hydrogen source streams SR0 and / or new hydrogen streams are added to the hydrogen sources corresponding to the hydrogen networks participating in the optimization. This can transform infeasible points in the input parameter space into feasible points. It is understood that the addition of virtual hydrogen source streams SR0 and / or new hydrogen streams only affects the optimization results corresponding to infeasible points in the input parameter space, and does not affect the optimization results corresponding to originally feasible points. That is, virtual hydrogen source streams SR0 and / or new hydrogen streams exist in the hydrogen networks corresponding to originally infeasible points in the input parameter space, while virtual hydrogen source streams SR0 and / or new hydrogen streams do not exist in the hydrogen networks corresponding to originally feasible points. It should be noted that in practical applications, virtual hydrogen source streams SR0 cannot be added to the pressure swing adsorption (PSA) unit. If a virtual hydrogen source stream SR0 exists in the hydrogen source corresponding to the target hydrogen network, it needs to be removed, and the target input / output parameter set and the target hydrogen network need to be optimized so that the optimized target input / output parameter set and the target hydrogen network can be used in practical applications.

[0106] Specifically, based on the target input / output parameter set and the target hydrogen network, the input streams, the sum of input stream flow rates, and the concentration of the input streams for the corresponding pressure swing adsorption (PSA) units are determined. Based on this, it is determined whether a virtual hydrogen stream SR0 and / or a new hydrogen stream have been added to the hydrogen source corresponding to the target hydrogen network, i.e., to the input streams of the PSA units corresponding to the target hydrogen network. If no virtual hydrogen stream SR0 and / or a new hydrogen stream have been added, the target input / output parameter set and the target hydrogen network represent the final hydrogen network optimization result. If a virtual hydrogen stream SR0 and / or a new hydrogen stream have been added, then the virtual hydrogen stream SR0 and / or the new hydrogen stream need to be removed to obtain the sum of input stream flow rates and the corresponding concentrations after removing the virtual hydrogen stream SR0 and / or the new hydrogen stream. Based on this, the target input / output parameter set and the target hydrogen network are optimized and adjusted. It should be noted that, in order to ensure that the total input flow rate after removing the virtual hydrogen stream SR0 and / or the new hydrogen stream meets the requirements of different steps in the pressure swing adsorption unit (PSA), namely the adsorption step, pressure equalization step, pressure reduction step, purging step, pressure equalization step, and pressure increase step, it is necessary to adjust the inlet hydrogen stream flow rate and inlet hydrogen stream concentration in the adsorption step of the PSA unit. This will ensure that the total flow rate consumed by each step of the PSA unit is equal to the total input hydrogen stream flow rate after removing the virtual hydrogen stream SR0 and / or the new hydrogen stream, thereby achieving optimized adjustment of the target input and output parameter set and the target hydrogen network.

[0107] In this embodiment, by determining whether a virtual hydrogen stream and / or a new hydrogen stream have been added to the hydrogen source corresponding to the target hydrogen network, it can be ensured that the final target hydrogen network can be put into practical application. This not only improves the practicality of the hydrogen network, but also improves its reliability, thereby reducing production costs.

[0108] In one specific embodiment, during the initial design phase of the hydrogen network, the input parameters for the pressure swing adsorption (PSA) mechanism model are equipment parameters and operating parameters. The equipment parameters include at least the number of adsorption beds, bed size, and adsorbent loading. The operating parameters include at least molar flow rate, pressure, temperature, and operating time. During the hydrogen network optimization phase, the input parameters for the PSA mechanism model are operating parameters, which include at least molar flow rate, pressure, temperature, and operating time. At this stage, the equipment parameters are fixed and cannot be used as input parameters. Fixed parameters refer to equipment parameters other than the number of adsorption beds, bed size, and adsorbent loading, such as the type of adsorbent, adsorbent size, and compressor selection. It is understood that these fixed parameters are immutable during the hydrogen network optimization phase.

[0109] In one specific embodiment, a pressure swing adsorption (PSA) mechanism model, namely a two-tower, six-step PSA dynamic model, is established using Aspen Adsorption software. The adsorbents are activated carbon and 5A zeolite, with a volume ratio of 5:1. Input parameters include equipment parameters such as the number of adsorption beds, bed size, and adsorbent loading, as well as operating parameters such as molar flow rate, pressure, temperature, and operating time. Output parameters include gas flow rate, i.e., the average flow rate of the hydrogen stream at the outlet of the PSA unit, and gas concentrations such as hydrogen concentration, methane concentration, carbon dioxide concentration, and carbon monoxide concentration.

[0110] In one specific embodiment, such as Figure 2 As shown, Figure 2 Here is a flowchart illustrating the hydrogen network optimization method in another embodiment, including the following steps:

[0111] Step S201: Determine the hydrogen sources participating in the hydrogen network optimization, as well as the corresponding flow rate and concentration of the hydrogen sources, and determine the hydrogen traps participating in the hydrogen network optimization, as well as the corresponding flow rate and concentration requirements of the hydrogen traps.

[0112] Step S202: Randomly generate multiple sets of initial input parameters within the range of input parameters to form an initial sample set.

[0113] Step S203: Based on the Bayesian optimization algorithm, select a set of input parameters and input them into the pressure swing adsorption mechanism model to obtain the output parameters corresponding to the current input parameters.

[0114] Step S204: Generate the corresponding input and output parameter set based on the current input parameters and the corresponding output parameters.

[0115] Step S205: Add the current input and output parameter set to the hydrogen network optimization model to optimize the hydrogen network. During the hydrogen network optimization process, it is allowed to add virtual hydrogen source streams SRO and / or new hydrogen streams to the hydrogen network to keep the optimization results feasible.

[0116] Step S206: Obtain the hydrogen network and economic indicators corresponding to the current input and output parameter set.

[0117] Step S207: Determine whether the number of iterations has reached the preset number.

[0118] Specifically, if yes, then step S208 is executed to obtain the updated sample set, and from the updated sample set, the input parameters and hydrogen network corresponding to the minimum economic indicator are selected to complete the hydrogen network optimization process. If no, then step S209 is executed to add the current input parameters and the economic indicator and hydrogen network corresponding to the current input-output parameter group to the initial sample set, and update the initial sample set.

[0119] Step S208: Obtain the updated sample set, and from the updated sample set, select the input parameters and hydrogen network corresponding to the minimum economic indicator to complete the hydrogen network optimization process.

[0120] Step S209: Add the current input parameters, the economic indicators corresponding to the current input-output parameter group, and the hydrogen network to the initial sample set, and update the initial sample set.

[0121] Specifically, after step S209 is completed, proceed to step S203 and repeat the above process until the number of iterations reaches the preset number, and obtain the final minimum economic index and the corresponding input parameters and hydrogen network.

[0122] For example, assume that the hydrogen sources, their corresponding flow rates and concentrations, and the hydrogen traps involved in the hydrogen network optimization, along with their corresponding flow rate and concentration requirements, are shown in Table 1. Table 1 contains information on hydrogen sources and hydrogen traps involved in the hydrogen network optimization, where HCU is a hydrocracking unit, GOHT is an oil and gas hydrotreating unit, RHT is a residue hydrotreating unit, DHT is a diesel hydrotreating unit, NHT is a naphtha hydrotreating unit, HP is a new hydrogen stream, and CRU is a catalytic resetting unit. Assume that 50 initial input parameters are randomly generated within the input parameter range to form an initial sample set; and assume that the preset number of iterations is 1000. Based on the Bayesian optimization algorithm, steps S203-S206 are repeated until the number of iterations reaches 1000, obtaining an updated sample set. Based on the updated sample set, the minimum economic indicator, i.e., the minimum annual total cost, can be determined as 12.202 M$ / year, along with the corresponding input parameters: adsorption bed length of 6.38 m, adsorption bed diameter of 3.93 m, adsorption pressure of 20.00 bar, adsorption time of 78.3 s, and inlet hydrogen stream flow rate of 1.868 mol·s. -1 ·bar -1 The concentration of the imported hydrogen stream is 0.7282; among which, the adsorption bed length and adsorption bed diameter are the parameters corresponding to the adsorption bed size in the equipment parameters; the adsorption pressure, adsorption time, imported hydrogen stream flow rate and imported hydrogen stream concentration are the parameters corresponding to the adsorption steps in the operating parameters.

[0123] Table 1. Information on hydrogen sources and hydrogen traps involved in hydrogen network optimization.

[0124] HCU 0.8000 1156.49 HCU 0.8670 753.46 GOHT 0.7500 1024.49 GOHT 0.8360 685.19 RHT 0.7500 485.43 RHT 0.8260 320.65 DHT 0.7000 154.59 DHT 0.7490 99.21 NHT 0.6500 70.97 NHT 0.7270 47.60 HP 0.9500 1106.89 CRU 0.8000 214.46

[0125] In this embodiment, to avoid generating infeasible solutions, a virtual hydrogen source stream SRO and a new hydrogen stream with a hydrogen concentration of 100% were added to the original hydrogen source network. However, in practical applications, the virtual hydrogen source stream SRO cannot be added to the pressure swing adsorption (PSA) unit. Therefore, it is necessary to remove the virtual hydrogen source stream SRO and optimize, i.e., fine-tune, the inlet hydrogen stream flow rate and inlet hydrogen stream concentration of the adsorption step in the PSA unit to ensure that the optimized minimum economic index, the corresponding input parameters, and the hydrogen network can be put into practical application. The optimized inlet hydrogen stream flow rate for the adsorption step is 1.866 mol·s⁻¹. -1 ·bar -1The hydrogen concentration of the imported hydrogen stream is 0.7292, and the corresponding minimum economic indicator is reduced to 12.185 M$ / year. The optimal hydrogen network matching relationship obtained in this embodiment is shown in Table 2. Table 2 is a table of matching relationships between different hydrogen sources and hydrogen traps in the optimal hydrogen network. In the table, the hydrogen source represents the unused, recyclable hydrogen stream contained in the outlet stream of the hydrogen production unit or the outlet stream of the hydrogen consumption unit; the hydrogen trap represents the inlet stream of the hydrogen consumption unit. In Table 2, HCU is the hydrocracking unit, GOHT is the oil and gas hydrotreating unit, RHT is the residue hydrotreating unit, DHT is the diesel hydrotreating unit, NHT is the naphtha hydrotreating unit, HP is the new hydrogen stream, CRU is the catalytic resetting unit, and PSA is the pressure swing adsorption unit.

[0126] Table 2. Matching relationships between different hydrogen sources and hydrogen traps in the optimal hydrogen network.

[0127]

[0128] The aforementioned hydrogen network optimization method has three aspects: First, based on the pressure swing adsorption (PSA) mechanism model, it simulates the PSA process and generates multiple sets of input-output parameters, laying the foundation for hydrogen network optimization. Second, based on the hydrogen network optimization model, it obtains the economic indicators corresponding to each set of input-output parameters to update and expand the sample set, providing a data foundation for improving the reliability of the hydrogen network. Third, it uses a Bayesian optimization algorithm for iterative optimization to obtain the target hydrogen network and the target input-output parameter set, thus achieving hydrogen network optimization, improving the reliability of the hydrogen network, thereby increasing the utilization rate of hydrogen and reducing production costs.

[0129] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0130] Based on the same inventive concept, this application also provides a hydrogen network optimization device for implementing the hydrogen network optimization method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more hydrogen network optimization device embodiments provided below can be found in the limitations of the hydrogen network optimization method described above, and will not be repeated here.

[0131] In one embodiment, such as Figure 3 As shown, Figure 3 This is a structural block diagram of a hydrogen network optimization device in one embodiment. The hydrogen network optimization device includes: an acquisition module 301, a generation module 302, a hydrogen network optimization module 303, a construction module 304, and an iterative optimization module 305, wherein:

[0132] The acquisition module 301 is used to acquire parameter information of the pressure swing adsorption unit and construct a pressure swing adsorption mechanism model based on the parameter information;

[0133] The generation module 302 is used to generate multiple sets of input and output parameter sets based on the pressure swing adsorption mechanism model; each set of input and output parameter sets includes the input parameters of the pressure swing adsorption mechanism model and the corresponding output parameters.

[0134] The hydrogen network optimization module 303 is used to optimize the hydrogen network according to the hydrogen network optimization model by taking each set of input and output parameters as constraints, and obtain the hydrogen network and economic indicators corresponding to each set of input and output parameters.

[0135] Module 304 is used to construct a sample set based on the hydrogen network and economic indicators corresponding to each set of input and output parameters.

[0136] The iterative optimization module 305 performs iterative optimization on the sample set according to the Bayesian optimization algorithm to obtain the target hydrogen network and the target input and output parameter set.

[0137] In one embodiment, the generation module 302 is further used for

[0138] Based on the input parameter range of the pressure swing adsorption mechanism model, multiple sets of input parameters are selected; the input parameters include equipment parameters and operating parameters.

[0139] Multiple sets of input parameters are input into the pressure swing adsorption mechanism model to obtain the corresponding output parameters; among them, the output parameters include gas flow rate and gas concentration;

[0140] Generate multiple sets of input and output parameters based on multiple sets of input parameters and their corresponding output parameters.

[0141] In one embodiment, the hydrogen network optimization module 303 is further used for

[0142] In the hydrogen source corresponding to the hydrogen network optimized by the hydrogen network optimization model, add virtual hydrogen streams and / or new hydrogen streams.

[0143] In one embodiment, the hydrogen network optimization module 303 is further used for

[0144] The hydrogen network optimization model includes: hydrogen source / hydrogen trap constraints, pressure swing adsorption unit constraints, gas system constraints, and objective function constraints.

[0145] In one embodiment, the iterative optimization module 305 is further used for

[0146] The sample set is iteratively optimized using the Bayesian optimization algorithm, and the results of each iteration are added to the sample set. The iteration results include: the iterative input and output parameter sets, as well as the corresponding hydrogen network and economic indicators.

[0147] If the number of iterations reaches the preset number, the iteration stops, and the updated sample set is obtained;

[0148] From the updated sample set, the input-output parameter set and hydrogen network corresponding to the minimum economic indicator are selected as the target input-output parameter set and target hydrogen network.

[0149] In one embodiment, the iterative optimization module 305 is further used for

[0150] When Bayesian optimization algorithm iteratively optimizes a sample set, it uses a surrogate function to fit the relationship between the input and output parameter sets and economic indicators.

[0151] In one embodiment, the hydrogen network optimization module 303 is further used for

[0152] Determine whether virtual hydrogen streams and / or new hydrogen streams have been added to the hydrogen source corresponding to the target hydrogen network;

[0153] If added, the target input / output parameter set and the target hydrogen network will be optimized.

[0154] Each module in the aforementioned hydrogen network optimization device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.

[0155] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 4 As shown, Figure 4This is an internal structural diagram of a computer device in one embodiment. The computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores hydrogen network optimization-related data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements a hydrogen network optimization method.

[0156] Those skilled in the art will understand that Figure 4 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0157] In one embodiment, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above method embodiments.

[0158] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0159] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0160] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0161] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0162] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for optimizing hydrogen networks, characterized in that, The method includes: Obtain the parameter information of the pressure swing adsorption unit, and construct a pressure swing adsorption mechanism model based on the parameter information; Based on the pressure swing adsorption mechanism model, multiple sets of input and output parameters are generated; each set of input and output parameters includes the input parameters of the pressure swing adsorption mechanism model and the corresponding output parameters. Each set of input and output parameters is used as a constraint condition, and the hydrogen network is optimized according to the hydrogen network optimization model to obtain the hydrogen network and economic indicators corresponding to each set of input and output parameters. The hydrogen network optimization model includes: hydrogen source / hydrogen trap constraints, pressure swing adsorption unit constraints, gas system constraints, and objective function constraints. Virtual hydrogen streams and / or new hydrogen streams are added to the hydrogen source corresponding to the hydrogen network optimized by the hydrogen network optimization model. A sample set is constructed based on the hydrogen network and economic indicators corresponding to each set of input and output parameters. The sample set is iteratively optimized using the Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter set.

2. The method according to claim 1, characterized in that, The generation of multiple sets of input and output parameters based on the pressure swing adsorption mechanism model includes: Based on the input parameter range of the pressure swing adsorption mechanism model, multiple sets of input parameters are selected; the input parameters include equipment parameters and operating parameters. Multiple sets of input parameters are input into the pressure swing adsorption mechanism model to obtain corresponding output parameters; wherein, the output parameters include gas flow rate and gas concentration; Multiple sets of input and output parameters are generated based on the multiple sets of input parameters and the corresponding output parameters.

3. The method according to claim 1, characterized in that, The step of iteratively optimizing the sample set using the Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter set includes: The sample set is iteratively optimized using the Bayesian optimization algorithm, and the results of each iteration are added to the sample set; the iteration results include: the iterative input and output parameter set, as well as the corresponding hydrogen network and economic indicators; If the number of iterations reaches the preset number, the iteration stops, and the updated sample set is obtained; From the updated sample set, the input-output parameter set and hydrogen network corresponding to the minimum economic indicator are selected as the target input-output parameter set and target hydrogen network.

4. The method according to claim 3, characterized in that, When the Bayesian optimization algorithm iteratively optimizes the sample set, it uses a surrogate function to fit the relationship between the input and output parameter sets and economic indicators.

5. The method according to claim 3, characterized in that, The method further includes: Determine whether a virtual hydrogen stream and / or a new hydrogen stream have been added to the hydrogen source corresponding to the target hydrogen network; If added, the target input / output parameter set and the target hydrogen network will be optimized.

6. A hydrogen network optimization device, characterized in that, The device includes: The acquisition module is used to acquire parameter information of the pressure swing adsorption unit and construct a pressure swing adsorption mechanism model based on the parameter information; The generation module is used to generate multiple sets of input and output parameter groups based on the pressure swing adsorption mechanism model; each set of input and output parameter groups includes the input parameters of the pressure swing adsorption mechanism model and the corresponding output parameters. The hydrogen network optimization module is used to optimize the hydrogen network according to the hydrogen network optimization model, taking each set of input and output parameters as constraints, to obtain the hydrogen network and economic indicators corresponding to each set of input and output parameters. The hydrogen network optimization model includes: hydrogen source / hydrogen trap constraints, pressure swing adsorption unit constraints, gas system constraints, and objective function constraints. Virtual hydrogen streams and / or new hydrogen streams are added to the hydrogen sources corresponding to the hydrogen networks optimized by the hydrogen network optimization model. The construction module is used to construct a sample set based on the hydrogen network and economic indicators corresponding to each set of input and output parameters. The iterative optimization module performs iterative optimization on the sample set according to the Bayesian optimization algorithm to obtain the target hydrogen network and the target input-output parameter set.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images