Hybrid intelligent simplification method for low-carbon alternative fuel mechanism of engine

Abstract

The invention discloses a hybrid intelligent simplification method for an engine low-carbon alternative fuel mechanism, and the method specifically comprises the following steps: initialization: generating a population represented by binary codes through a random algorithm, and the code of each individual in the population is composed of 0 and 1; for each binary individual generated in the previous step, calculating the fitness value of each binary individual according to a preset objective function; and updating the population according to a fitness calculation result. In the step, the optimal particle and the global optimal particle in the current population are determined, and the speed (V) of the particles is calculated at the same time; performing selection, crossover and variation updating operation to obtain a new population with an optimal adaptive value; outputting an optimal solution, namely the simplest mechanism; the invention belongs to the field of combustion reaction kinetics numerical simulation, and according to the ideas of the particle swarm optimization and the genetic algorithm, the size of the combustion mechanism of the low-carbon alternative fuel of the engine can be remarkably reduced, the complexity of the combustion mechanism is reduced, and calculation resources are saved for combustion numerical simulation work.

Description

Technical Field

[0001] The present invention belongs to the field of numerical simulation of combustion reaction dynamics, and in particular relates to a hybrid intelligent simplification method for low-carbon alternative fuel mechanism of an engine. Background Art

[0002] As the world pays more and more attention to environmental protection and addressing climate change, exploring low-carbon or even zero-carbon alternative fuels to apply to existing engine systems to achieve a significant reduction in carbon emissions while meeting power requirements has become a hot research topic in the current energy and transportation industries. Ammonia can not only be used as a substitute for engine fuel, but is also considered a potential clean energy source due to its low carbon emission characteristics. However, complex combustion reaction mechanisms often involve a large number of components and reactions, resulting in extremely high computational complexity. Especially when high temperatures, complex reaction paths, and reactions of multiple components are involved, the time and computational cost of numerical simulation are huge. Therefore, simplification of the combustion reaction mechanism is the key to optimizing the ammonia combustion process.

[0003] At present, the research methods for simplifying the combustion mechanism of low-carbon alternative fuels in engines include some simplified methods based on sensitivity analysis that can identify and eliminate components and reactions that have little impact on certain key indicators of the combustion process, and simplified methods based on error propagation analysis that take into account the correlation between reactions. However, these methods still face the problem of balancing accuracy and computational efficiency. Therefore, how to simplify the combustion mechanism while retaining its accuracy and computational efficiency has become a challenge for the application of ammonia as an alternative fuel in engines.

[0004] Aiming at the limitations of some current mechanism simplification methods, this paper proposes a hybrid intelligent simplification method for low-carbon alternative fuel mechanism of engine, which provides an efficient and accurate combustion mechanism simplification strategy based on the ideas of particle swarm algorithm and genetic algorithm. Summary of the invention

[0005] In view of the above problems, the present invention provides a hybrid intelligent simplification method for the low-carbon alternative fuel mechanism of engines by means of the ideas of particle swarm and genetic algorithm.

[0006] The method of the present invention can effectively simplify the detailed mechanism of ammonia, a low-carbon alternative fuel for engines, improve the calculation efficiency in numerical simulation of combustion, and provide strong technical support for the further application and development of ammonia fuel in engines.

[0007] The present invention is realized by the following technical scheme: a hybrid intelligent simplified method for low-carbon alternative fuel mechanism of an engine, comprising the following steps:

[0008] Step 1: Initialization: Generate a population represented by binary code through a random algorithm. The code of each individual in the population consists of 0 and 1.

[0009] Step 2: For each binary individual generated in the previous step, calculate their fitness value according to the preset objective function;

[0010] Step 3: Update the population according to the results of fitness calculation. This step will determine the best particle (Pbest) and the global best particle (Gbest) in the current population, and calculate the particle velocity (V);

[0011] Step 4: Perform selection, crossover and mutation update operations to obtain a new population with the optimal fitness value;

[0012] Step 5: Output the optimal solution, which is the simplest mechanism.

[0013] Preferably, the step 1 is specifically:

[0014] When initializing the population, a random algorithm is used to generate multiple binary chromosome samples, denoted as Q i (i=1,2,...,S), where 0 means that the component is eliminated in the new mechanism, 1 means that the component is retained in the new mechanism, S represents the total number of samples, Q i Represents the total number of components in the mechanism.

[0015] Preferably, the step 2 is specifically:

[0016] The fitness of each random sample is calculated, and the fitness function is shown in formula (1):

[0017]

[0018] Among them, the simplified mechanism ignition delay time error of the i-th sample is δ i It means that under the kth operating condition, the ignition delay time of the detailed mechanism is The ignition delay time of the simplified mechanism is X is the set of initial conditions, t0 is the cutoff time, t com is the time spent on the simplified mechanism calculation process. The total number of components in the simplified mechanism is represented by S species It indicates that the cutoff error is set to 20, and t0 is set to twice the time consumed by the detailed mechanism to calculate the ignition delay in X space.

[0019] Preferably, the steps three and four are specifically as follows:

[0020] The next generation of particles is determined by four coefficients: W (own inertia factor), C1 (learning rate for Pbest), C2 (learning rate for Gbest) and V (speed). In the update iteration, some random sample arrays are selected for reorganization and mutation, and then continuous iterations are performed to find the best adaptive particles. When generating a new generation of particles, C1=C2=0.3, W=0.3, V=0.1 are set, the crossover method is single-point crossover, the crossover probability is 50%, and the mutation probability is set to 1%.

[0021] Preferably, the invention uses a simplified initial particle swarm algorithm to quickly locate local areas with greater potential, and then uses a genetic algorithm to further search in these areas and their surroundings, jump out of the local optimum, and explore a better global simplification solution, avoiding the problems of premature convergence or inefficient search that may occur in a single algorithm, making the entire simplification process more efficient, and ultimately obtaining a streamlined engine low-carbon alternative fuel mechanism.

[0022] The present invention provides a hybrid intelligent simplified method for low-carbon alternative fuel mechanism of engines. Compared with the existing technology, it has the following beneficial effects:

[0023] The present invention focuses on the simplified technology research of the ammonia mechanism of alternative fuels for engines, and adopts a simplified method combining particle swarm algorithm and genetic algorithm, giving full play to the advantages of strong global search ability of particle swarm algorithm and fine local search of genetic algorithm, and good at handling discrete problems, so as to better handle the common nonlinear problems in the field of combustion and efficiently simplify the complex ammonia combustion reaction mechanism. The present invention can significantly reduce the size of the ammonia combustion mechanism of alternative fuels for engines, reduce the complexity of the combustion mechanism, and save computing resources for the numerical simulation of combustion. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a simplified schematic diagram of the mechanism based on the hybrid intelligent algorithm.

[0025] Figure 2 This is a graph showing the relationship between the number of elementary reactions in the simplified mechanism and the number of iterations when the simplification method of the hybrid intelligent algorithm is used to simplify the NH3 / O2 / AR mechanism.

[0026] Figure 3 , Figure 4 It is a comparison diagram of the ignition delay time of the simplified mechanism and the detailed mechanism of NH3 / O2 / AR under different working conditions obtained by the present invention.

[0027] Figure 5 This is a graph showing the relationship between the number of elementary reactions in the simplified mechanism and the number of iterations when the simplification method of the hybrid intelligent algorithm is used to simplify the NH3 / O2 / AR mechanism.

[0028] Figure 6 , Figure 7 , Figure 8 , Fig. 9 It is a comparison diagram of the ignition delay time of the simplified mechanism and the detailed mechanism of NH3 / O2 / AR under different working conditions obtained by the present invention. DETAILED DESCRIPTION

[0029] In the embodiments of the present invention, the technical solutions are described in detail with reference to the corresponding drawings, but what is described here is only a part of the many embodiments of the present invention. Based on the embodiments included in the present invention, all other embodiments obtained by ordinary researchers and technicians in the field without creative work are within the scope of protection of the present invention.

[0030] Please see attached Figure 1-5 The present invention provides two implementation columns: a hybrid intelligent simplified method for low-carbon alternative fuel mechanism of an engine, the implementation examples are as follows:

[0031] Specific example 1:

[0032] This hybrid intelligent method was applied to simplify the ammonia oxidation mechanism developed by Shrestha et al., which includes 1090 elementary reactions and 125 components. The paper is titled: Detailed Kinetic Mechanism for the Oxidation of Ammonia Including the Formation and Reduction of Nitrogen Oxides.

[0033] according to Figure 1 As shown, the present invention provides a hybrid intelligent simplified method for the combustion mechanism of low-carbon alternative fuels in an engine, comprising the following steps:

[0034] Step 1: Simplify in a wide range and set the initial conditions: pressure p range is 1.4~30atm, equivalence ratio range is 0.5~1.0, temperature T range is 1550~2500K; the initial population number is 100, and the number of iterations is 200;

[0035] Step 2: By increasing the number of iterations, the number of reactions in the combustion mechanism is continuously reduced. After more than 100 iterations, the number of reactions has been greatly reduced and gradually tends to be constant, and finally a simplified mechanism of 17 reactions and 16 components is obtained, which is 1073 reactions less than the original detailed mechanism, greatly reducing the size of the mechanism. The elementary reactions obtained by simplifying the NH3 / O2 / AR combustion mechanism are shown in Table 1 below, and the component names are shown in Table 2 below. The iteration diagram of the simplified NH3 / O2 / AR detailed mechanism using this invention is shown in Table 2 below. Figure 2 shown.

[0036] Table 1 Elementary reactions obtained by simplifying the NH3 / O2 / AR mechanism

[0037]

[0038] Table 2 Component names obtained by simplifying the NH3 / O2 / AR mechanism

[0039]

[0040]

[0041] Step 3: Calculate and compare the ignition delay time for the simplified mechanism with 17 reactions, according to Figure 3 As shown, the simplified NH3 / O2 / AR mechanism of reaction 17 can accurately predict the ignition delay time, and the maximum error of the ignition delay time is less than 10%. This result demonstrates the effectiveness of the simplification of this invention.

[0042] Specific example 2:

[0043] In order to further test the simplification capability of the invention, the hybrid intelligent method was applied to simplify the Zhou mechanism, which contains 1268 elementary reactions and 169 components. The document is titled: An experimental and kinetic modeling study on NH3 / air, NH3 / H2 / air, NH3 / CO / air, and NH3 / CH4 / air premixed laminarflames at elevated temperature.

[0044] Step 1: In this embodiment, the initial conditions for calculation are: the pressure p ranges from 1.4 to 30 atm, the equivalence ratio is 0.5 to 2.0, the temperature T ranges from 1550 to 2500 K, the number of initialized populations is 100, and the number of iterations is 200.

[0045] Step 2: In the early stage of iteration, due to a large number of non-critical reactions, the number of reactions in the mechanism dropped rapidly. When the iteration was halfway through, the number of reactions in the mechanism had been reduced to 20, and the simplification rate was about 98%. In the subsequent iterations, the number of reactions slowly decreased. After 109 iterations, the number of reactions remained unchanged. Finally, a simplified mechanism of 16 reactions and 14 components was obtained. The elementary reactions obtained by simplifying the NH3 / O2 / AR mechanism are shown in Table 3 below, and the component names are shown in Table 4 below. The iteration diagram of the simplified NH3 / O2 / AR detailed mechanism using this invention is shown in Table 4 below. Figure 4 shown.

[0046] Table 3 Elementary reactions obtained by simplifying the NH3 / O2 / AR mechanism

[0047]

[0048]

[0049] Table 4 Component names obtained by simplifying the NH3 / O2 / AR mechanism

[0050]

[0051] Step 3: Analyze the ignition prediction capability of the NH3 / O2 / AR combustion mechanism. Figure 5 The ignition prediction capability of the simplified mechanism of 16 elementary reactions of NH3 / O2 / AR combustion over a wide range is demonstrated. Through the corresponding simplification, the simplified mechanism of 16 elementary reactions is obtained, and the simplified mechanism can be used to accurately predict the ignition delay time of NH3 / O2 / AR over a wide range. Compared with the original mechanism, this simplified mechanism reduces the number of reactions by 1252, which only accounts for 1.26% of the original mechanism. It can be seen that the simplified method based on the hybrid algorithm has an excellent performance in accurately predicting the ignition delay time of NH3 / O2 / AR.

[0052] In the process of in-depth exploration of the combustion mechanism, it can be seen from the simplified size results that this invention can obtain a more compact simplified mechanism. In terms of the ignition delay time prediction results, the invention shows excellent prediction ability, which can accurately predict the ignition delay time of the combustion mechanism, providing reliable data support for related research. Therefore, this invention has shown significant superiority in the important field of combustion mechanism simplification. It not only helps to optimize the structure of the engine's low-carbon fuel combustion mechanism, making it more concise and efficient, but also performs outstandingly in accurately predicting the ignition delay time, providing a highly efficient and intelligent strategy for the simplification of the mechanism of highly complex chemical reaction systems.

[0053] The above is a preferred implementation scheme of a hybrid intelligent simplified method for a low-carbon alternative fuel mechanism for an engine. The above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made by any technician familiar with the technical field within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A hybrid intelligent simplified method for low-carbon alternative fuel mechanism of an engine, characterized by: The following steps are involved: Step 1: Initialization: Generate a population represented by binary code through a random algorithm. The code of each individual in the population consists of 0 and 1. Step 2: For each binary individual generated in the previous step, calculate their fitness value according to the preset objective function; Step 3: Update the population according to the results of fitness calculation. This step will determine the best particle (Pbest) and the global best particle (Gbest) in the current population, and calculate the particle velocity (V); Step 4: Perform selection, crossover and mutation update operations to obtain a new population with the optimal fitness value; Step 5: Output the optimal solution, which is the simplest mechanism.

2. A hybrid intelligent simplified method for low-carbon alternative fuel mechanism of an engine according to claim 1, characterized in that: In the step 1, when the population is initialized, a random algorithm is used to generate multiple binary chromosome samples, denoted as Qi (i=1, 2, ..., S), where 0 represents the elimination of the component in the new mechanism, 1 represents the retention of the component in the new mechanism, S represents the total number of samples, and Qi represents the total number of components in the mechanism.

3. A hybrid intelligent simplified method for low-carbon alternative fuel mechanism of an engine according to claim 1, characterized in that: The step 2 is specifically as follows: The fitness of each random population is calculated, and the fitness function is shown in formula (1): Among them, the simplified mechanism ignition delay time error of the i-th sample is δ i It means that under the kth operating condition, the ignition delay time of the detailed mechanism is The ignition delay time of the simplified mechanism is X is the set of initial conditions, t0 is the cutoff time, t com is the time spent on the simplified mechanism calculation process. The total number of components in the simplified mechanism is represented by S species It indicates that the cutoff error is set to 20, and t0 is set to twice the time consumed by the detailed mechanism to calculate the ignition delay in X space.

4. A hybrid intelligent simplified method for low-carbon alternative fuel mechanism of an engine according to claim 1, characterized in that: The steps three and four are specifically as follows: The next generation of particles is determined by four coefficients: W (own inertia factor), C1 (learning rate for Pbest), C2 (learning rate for Gbest) and V (speed). In the update iteration, some random sample arrays are selected for reorganization and mutation, and then continuous iterations are performed to find the best adaptive particles. When generating a new generation of particles, C1=C2=0.3, W=0.3, V=0.1 are set, the crossover method is single-point crossover, the crossover probability is 50%, and the mutation probability is set to 1%.

5. The hybrid intelligent simplified method for low-carbon alternative fuel mechanism of an engine according to claim 1, characterized in that: The particle swarm algorithm is used in the initial stage of simplification to quickly locate local areas with greater potential, and then the genetic algorithm is used to further search in these areas and their surrounding areas to escape the local optimum and explore a better global simplification solution, avoiding the problems of premature convergence or inefficient search that may occur in a single algorithm, making the entire simplification process more efficient and ultimately obtaining a streamlined engine low-carbon alternative fuel mechanism.

Images