A method and system for detecting assembly process of a gas distribution mechanism

Abstract

The application discloses a kind of gas distribution mechanism assembly process detection method and system, comprising: based on the relationship between the tolerance value of each component historical assembly and the running state of gas distribution mechanism Bayes network is constructed;Bayes network is optimized based on the dragonfly algorithm of fusion artificial fish behavior, the dragonfly algorithm of fusion artificial fish behavior is: the foraging, aggregation and pursuit behavior in artificial fish school algorithm is introduced into the dragonfly behavior of dragonfly algorithm, to update the moving direction of dragonfly, the score function of Bayes network is regarded as the fitness function of dragonfly, the difference value of the fitness value of the tth iteration step and the t+1th iteration step and the fitness value of the tth iteration step are determined, and the ratio of the tth iteration step is combined with the field of vision of artificial fish school and dragonfly behavior to change the optimization range of dragonfly;The tolerance value of each component actually assembled in gas distribution mechanism is detected using the optimized Bayes network for gas distribution mechanism assembly process.Effectively control the process parameters of gas distribution mechanism, improve the thermal efficiency of engine.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of engine intelligent assembly, in particular to a valve train assembly process detection method and system. BACKGROUND

[0002] The statements in this section merely provide background information related to the present application and do not necessarily constitute the prior art.

[0003] The valve train is one of the important components of the engine, and its function is to open and close the intake valve and exhaust valve of each cylinder according to the working order and working cycle requirements of the engine, so that the mixture enters the cylinder and the exhaust gas is discharged from the cylinder. The power, economy and reliability of the engine are closely related to the valve train, and whether the design of the valve train is reasonable will affect the stability and charging efficiency of the engine.

[0004] The smooth progress of the breathing process requires the periodic closing of the exhaust valve, and the intake and exhaust valves of the valve train have an early opening angle and a late closing angle. Early or late opening of the intake and exhaust valves is not conducive to improving the thermal efficiency of the engine, and the assembly process parameters of the valve train also have a certain influence on the early opening angle and the late closing angle. If an unreasonable valve train assembly process is used, it will have an adverse effect on the stable operation of the engine.

[0005] Furthermore, although the relationship between the changes in the early opening angle and the late closing angle of the intake and exhaust valves and the assembly process parameters can be established through a Bayesian network, the learning problem of the structure of the Bayesian network is an NP difficult problem, so the existing methods cannot efficiently mine the optimal Bayesian network structure from the data. SUMMARY

[0006] In order to solve the above problems, the present application provides a valve train assembly process detection method and system, which effectively controls the process parameters of the valve train, designs a more reasonable valve train assembly process, and thus more accurately and effectively controls the early opening angle and late closing angle of the intake and exhaust valves, improves the thermal efficiency of the engine, and ensures the stability and charging efficiency of the engine.

[0007] In order to achieve the above purpose, the present application adopts the following technical solutions:

[0008] In a first aspect, the present application provides a valve train assembly process detection method, comprising:

[0009] Obtaining the tolerance value of the historical assembly of each component in the valve train, and constructing a Bayesian network based on the relationship between the tolerance value of the historical assembly of each component and the operating state of the valve train;

[0010] The artificial fish swarm algorithm is introduced into the dragonfly algorithm to update the moving direction of the dragonfly, the score function of the Bayesian network is taken as the fitness function of the dragonfly, the ratio of the difference between the fitness value of the tth iteration step and the fitness value of the t+1th iteration step and the fitness value of the tth iteration step is determined, and the artificial fish swarm field is combined with the dragonfly behavior to change the dragonfly optimization range.

[0011] The optimized Bayesian network is used to detect the assembly process of the valve train according to the tolerance value of the actual assembly of each component in the valve train.

[0012] As an optional implementation, the components in the valve train include a camshaft, a tappet, a push rod, a rocker arm, an intake valve and an exhaust valve, the operating state of the valve train includes the early opening angle and the late closing angle of the intake and exhaust valves, and the Bayesian network is constructed according to the relationship between the early opening angle and the late closing angle of the intake and exhaust valves and the tolerance value of the assembly of each component in the valve train.

[0013] As an optional implementation, the foraging behavior in the artificial fish swarm algorithm is introduced into the dragonfly algorithm.

[0014]

[0015] The aggregation behavior in the artificial fish swarm algorithm is introduced into the dragonfly algorithm.

[0016]

[0017] The pursuit behavior in the artificial fish swarm algorithm is introduced into the dragonfly algorithm.

[0018]

[0019] Wherein, F i is the model of the i th individual attracted by food, C i is the model of the aggregation behavior of the i th individual, E i is the model of the i th individual away from the enemy, X + is the position of the food, X - is the position of the enemy, X i is the position of the individual, X c is the center position of the aggregation of the artificial fish after the dragonfly is regarded as the artificial fish, ω1, ω2, ω3 are weights, rand is a random number between 0 and 1, and Step is a step length.

[0020] As an optional implementation, the moving direction of the dragonfly is:

[0021] ΔX t+1 = (fF i+cC i +eE i )+ωΔX t

[0022] Wherein, ΔX, ΔX+1 are the moving direction of the tth iteration step and the t+1th iteration step, f is the food weight, c is the weight corresponding to the group behavior, e is the enemy weight, ω is the inertia weight, and t is the iteration number.

[0023] As an optional implementation, the dragonfly optimization range with the artificial fish vision field is vis-df = vis-df * alpha + vis-df min ; wherein, vis-df min is the optimization range change step, and alpha is the change coefficient of the optimization range.

[0024] As an optional implementation, the change of the dragonfly optimization range includes:

[0025] When is less than a set threshold value, the change coefficient of the optimization range is alpha = 0.5 + 0.4cos(fF + cC);

[0026] When is not less than the set threshold value, the change coefficient of the optimization range is alpha = 3 + 2cos(eE);

[0027] Wherein, the fitness value of the ith individual at the tth iteration step is KT t , the fitness value of the ith individual at the t+1th iteration step is KT t+1 , F i is the model of the ith individual attracted by food, C i is the model of the group behavior of the ith individual, E i is the model of the ith individual away from the enemy, f is the food weight, c is the weight corresponding to the group behavior, and e is the enemy weight.

[0028] As an optional implementation, in the process of optimizing the Bayesian network based on the dragonfly algorithm fused with the artificial fish behavior, each Bayesian network is a dragonfly with the artificial fish optimization characteristics, and the best Bayesian network is found according to the score of the Bayesian network under the score function of the Bayesian network.

[0029] In a second aspect, the present application provides a gas distribution mechanism assembly process detection system, comprising:

[0030] A network construction module is configured to obtain the tolerance value of historical assembly of each component in the gas distribution mechanism, and construct a Bayesian network based on the relationship between the tolerance value of historical assembly of each component and the running state of the gas distribution mechanism.

[0031] The network optimization module is configured to optimize the Bayesian network based on a dragonfly algorithm fusing artificial fish behaviors, the dragonfly algorithm fusing artificial fish behaviors being: introducing foraging, aggregation and pursuit behaviors in an artificial fish school algorithm into dragonfly behaviors of the dragonfly algorithm to update a moving direction of the dragonfly, taking a score function of the Bayesian network as an adaptability function of the dragonfly, determining a ratio of a difference between adaptability values of a tth iteration step and a t+1th iteration step and the adaptability value of the tth iteration step, and combining a field of vision of the artificial fish school with the dragonfly behaviors to change a dragonfly optimization range;

[0032] The process detection module is configured to detect a gas distribution mechanism assembly process by using the optimized Bayesian network on a tolerance value actually assembled by each component in the gas distribution mechanism.

[0033] In a third aspect, the present application provides an electronic device, comprising a memory and a processor, and computer instructions stored in the memory and running on the processor, when the computer instructions are run by the processor, the method of the first aspect is completed.

[0034] In a fourth aspect, the present application provides a computer readable storage medium for storing computer instructions, when the computer instructions are executed by a processor, the method of the first aspect is completed.

[0035] Compared with the prior art, the present application has the following beneficial effects:

[0036] The present application provides a gas distribution mechanism assembly process detection method and system, the early opening angle and the late closing angle of the intake and exhaust are associated with the camshaft, the tappet, the push rod, the rocker arm, the intake valve and the exhaust valve by the Bayesian network, the foraging, aggregation and pursuit behaviors in the artificial fish school algorithm are introduced into the dragonfly behaviors of the dragonfly algorithm, the field of vision of the artificial fish school is combined with the dragonfly behaviors to improve the dragonfly algorithm, and the improved dragonfly algorithm fusing artificial fish behaviors is applied to the learning of the Bayesian network structure, forming a gas distribution structure assembly process detection method based on an intelligent Bayesian network, effectively controlling the process parameters of the gas distribution mechanism, designing a more reasonable gas distribution mechanism assembly process, thereby more accurately and effectively controlling the early opening angle and the late closing angle of the intake and exhaust valves, improving the thermal efficiency of the engine, and ensuring the stability and charging efficiency of the engine.

[0037] The advantages of the additional aspects of the present application will be partially given in the following description, partially become obvious from the following description, or be known by the practice of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0038] The drawings accompanying the specification of the present application form a part thereof, serve to provide further understanding of the present application, and together with the description of the exemplary embodiments of the present application and the explanation thereof serve to explain the present application, and do not constitute an improper limitation on the present application.

[0039] Figure 1 The flow chart of the detection method for the assembly process of the gas distribution mechanism provided in Embodiment 1 of the present application is shown in FIG. 1. DETAILED DESCRIPTION

[0040] The present application will be further described below with reference to the accompanying drawings and embodiments.

[0041] It should be noted that the following detailed description is exemplary in nature and is intended to provide further description of the present application. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0042] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments in accordance with the present application. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof.

[0043] The embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0044] Embodiment 1

[0045] The present embodiment provides a gas distribution mechanism assembly process detection method based on an intelligent Bayesian network, comprising:

[0046] Obtaining the tolerance values of the historical assembly of each component in the gas distribution mechanism, and constructing a Bayesian network based on the relationship between the tolerance values of the historical assembly of each component and the running state of the gas distribution mechanism;

[0047] Optimizing the Bayesian network based on a dragonfly algorithm fusing artificial fish behavior, wherein the dragonfly algorithm fusing artificial fish behavior is: introducing the foraging, aggregation and pursuit behaviors in the artificial fish school algorithm into the dragonfly behavior of the dragonfly algorithm to update the moving direction of the dragonfly, taking the score function of the Bayesian network as the fitness function of the dragonfly, determining the ratio of the difference between the fitness values of the tth iteration step and the t+1th iteration step and the fitness value of the tth iteration step, and combining the field of vision of the artificial fish with the dragonfly behavior to change the optimization range of the dragonfly;

[0048] Using the optimized Bayesian network to detect the assembly process of the gas distribution mechanism for the tolerance values of the actual assembly of each component in the gas distribution mechanism.

[0049] In the embodiment, the components of the valve train include camshaft, tappet, push rod, rocker arm, intake valve and exhaust valve; the operating states of the valve train include early opening angle and late closing angle of the intake and exhaust valves; the early opening angle and late closing angle of the intake and exhaust valves are influenced by camshaft tolerance, tappet tolerance, push rod tolerance, rocker arm tolerance, gap between rocker arm and valve, intake valve tolerance and exhaust valve tolerance, and the smaller camshaft tolerance will make the intake and exhaust valves open late and close early; the tappet tolerance, push rod tolerance and rocker arm tolerance will also affect the early opening angle and late closing angle of the intake and exhaust valves, and the too large gap between rocker arm and valve will make the valve open for a short time and the intake be insufficient, and the too small gap will make the valve close loosely and leak. Therefore, the embodiment establishes the relationship between the early opening angle and late closing angle of the intake and exhaust valves and the influencing factors thereof by the Bayesian network, and constructs the Bayesian network.

[0050] When solving practical application problems based on the Bayesian theory, a key task is to establish a Bayesian network topology structure capable of describing the relationship between different attributes according to the characteristics of the research target. Since learning the optimal BN structure is an NP-hard problem, generally, swarm intelligence optimization methods are used to solve the problem, but such methods still have problems such as low search efficiency and low optimization accuracy. In view of the above problems, the dragonfly algorithm fused with artificial fish behavior is applied to the learning of the Bayesian network structure to improve the search efficiency and optimization accuracy.

[0051] The dragonfly algorithm fused with artificial fish behavior is specifically as follows:

[0052] The foraging, aggregation and pursuit behaviors in the artificial fish swarm algorithm are introduced into the dragonfly behavior of the dragonfly algorithm to improve the performance of the dragonfly algorithm; the position of dragonfly i in a population consisting of m dragonflies is represented as X i =(x i1 ,x i2 ,...x in ), wherein i = 1, 2,... m; n is the dimension;

[0053] Then, the description of the dragonfly algorithm includes:

[0054] Food attraction behavior:

[0055] F i =X + -X i

[0056] Aggregation behavior:

[0057]

[0058] Enemy avoidance behavior:

[0059] E i =X - +X i

[0060] The foraging behavior in artificial fish swarm algorithm is introduced into the dragonfly algorithm:

[0061]

[0062] Where F i is the mathematical modeling of the i-th dragonfly individual attracted by food, X + is the position of food, X i is the position of the i-th dragonfly individual, ω1 is the weight, and rand is a random number between 0 and 1.

[0063] The aggregation behavior in artificial fish swarm algorithm is introduced into the dragonfly algorithm:

[0064]

[0065] Where C i is the mathematical model of the aggregation behavior of the i-th dragonfly individual, X i is the position of the i-th dragonfly individual, X c is the center position of the aggregation after the dragonfly is regarded as artificial fish, and ω2 is the weight.

[0066] The pursuit behavior in artificial fish swarm algorithm is introduced into the dragonfly algorithm:

[0067]

[0068] Where E i is the mathematical modeling of the i-th dragonfly individual moving away from the enemy, X - is the position of the enemy, X i is the position of the i-th dragonfly individual, and ω3 is the weight.

[0069] Under the joint influence of the three modes, the moving direction of the dragonfly is:

[0070] ΔX t+1 = (fF i + cC i + eE i ) + ωΔX t

[0071] Where ΔX, ΔX+1 are the step vectors of the t-th iteration step and the t+1-th iteration step, which represent the moving direction of the dragonfly, f is the weight of food, c is the weight corresponding to the aggregation behavior, e is the weight of the enemy, ω is the inertia weight, and t is the iteration number.

[0072] The artificial fish swarm visual field is combined with the foraging behavior, flocking behavior and enemy-avoiding behavior of dragonfly to obtain the optimization range of the dragonfly, so that the dragonfly algorithm has the optimization characteristics of the artificial fish. The current optimization range of the dragonfly with the optimization characteristics of the artificial fish is vis-df, vis-df min is the step length of the change of the optimization range, and is equal to 0.01-0.03vis-df. The fitness value of the tth iteration step is Kt t , and the fitness value of the t+1th iteration step is Kt+1 t+1 .

[0073] In order to improve the search ability of the dragonfly, when is less than a set threshold (1% in this embodiment), vis-df is changed to vis-df*α+vis-df min to increase the local search ability, and α=0.5+0.4cos(fF+cC) is taken as the change coefficient of the optimization range.

[0074] When is not less than the set threshold, the global search ability is strengthened, and α=3+2cos(eE) is taken as the change coefficient of the optimization range.

[0075] The dragonfly algorithm based on the fusion of the artificial fish behavior is used to optimize the Bayesian network. In the process of searching for the optimal Bayesian network, each Bayesian network is a candidate, and each reasonable Bayesian network topology is regarded as a dragonfly with the optimization characteristics of the artificial fish.

[0076] Suppose that the Bayesian network has n nodes, and the topology structure of the Bayesian network can be represented by an n*n adjacency matrix X:

[0077]

[0078] In the formula, x ij =1 indicates that there is a directed edge between the node i and the node j, and the arrow direction is i pointing to j. When x ij =0, it indicates that there is no edge between the node i and the node j.

[0079] The abstract change of the topology structure can be converted into the change of the elements in the matrix by using the adjacency matrix to represent the topology structure of the Bayesian network. The optimization algorithm is used to optimize the topology structure of the Bayesian network, and in fact, the optimization algorithm is used to optimize the adjacency matrix. In the structure learning of the Bayesian network, a randomization algorithm is used to randomly generate a plurality of legal directed acyclic graphs as the initial population, and the maximum number of father nodes of each node in the initial population is limited. The initial population obtained by using the random generation algorithm is used as the input of the improved dragonfly algorithm, and the topology structure of the Bayesian network can be obtained after the iteration optimization of the algorithm.

[0080] The initial number of dragonflies is set as NP, the initial position (solution) of each dragonfly is randomly generated, and the number of parent nodes of each variable is not more than 1, and the score function KT of the Bayesian network is used t As the fitness (target) function of the dragonfly, that is, the score function of the Bayesian network is as follows:

[0081]

[0082] Where D is a given training set, and G is a Bayesian network structure.

[0083] In the score function KT of the Bayesian network i Under the mechanism, if the score of a network is higher, the topology structure of the network is better, and finally the best Bayesian network topology structure is formed.

[0084] In this embodiment, the tolerance values of the actual assembly of each component in the valve train mechanism are collected by using an industrial Internet of Things data collection device, and then the optimized Bayesian network is used for detection of the valve train assembly process, so as to effectively control the process parameters of the valve train mechanism, design a more reasonable valve train assembly process, and more accurately and effectively control the early opening angle and late closing angle of the intake and exhaust valves, improve the thermal efficiency of the engine, and ensure the stability and charging efficiency of the engine.

[0085] Embodiment 2

[0086] The embodiment provides a valve train assembly process detection system, which comprises:

[0087] The network construction module is configured to obtain the tolerance values of the historical assembly of each component in the valve train mechanism, and construct a Bayesian network based on the relationship between the tolerance values of the historical assembly of each component and the running state of the valve train mechanism.

[0088] The network optimization module is configured to optimize the Bayesian network based on a dragonfly algorithm fused with artificial fish behavior, wherein the dragonfly algorithm fused with artificial fish behavior is as follows: the foraging, aggregation and pursuit behaviors in the artificial fish swarm algorithm are introduced into the dragonfly behavior of the dragonfly algorithm to update the moving direction of the dragonfly, the score function of the Bayesian network is taken as the fitness function of the dragonfly, the ratio of the difference between the fitness values of the tth iteration step and the (t+1)th iteration step and the fitness value of the tth iteration step is determined, and the field of view of the artificial fish swarm is combined with the dragonfly behavior to change the optimization range of the dragonfly.

[0089] The process detection module is configured to use the optimized Bayesian network to detect the valve train assembly process based on the tolerance values of the actual assembly of each component in the valve train mechanism.

[0090] It should be noted that the above modules correspond to the steps described in Embodiment 1, and the above modules have the same examples and application scenarios as the corresponding steps, but are not limited to the content disclosed in Embodiment 1. It should be noted that the above modules can be executed in a computer system such as a set of computer executable instructions as part of a system.

[0091] In more embodiments, there are also provided:

[0092] An electronic device includes a memory and a processor, and computer instructions stored in the memory and running on the processor, when the computer instructions are run by the processor, the method described in Embodiment 1 is completed. For brevity, it will not be described here.

[0093] It should be understood that in this embodiment, the processor can be a central processing unit CPU, and the processor can also be other general-purpose processors, digital signal processors DSP, application-specific integrated circuits ASIC, ready-to-program gate arrays FPGA or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0094] The memory can include read-only memory and random access memory, and provide instructions and data to the processor, and a part of the memory can also include non-volatile random access memory. For example, the memory can also store device type information.

[0095] A computer readable storage medium for storing computer instructions, when the computer instructions are executed by a processor, the method described in Embodiment 1 is completed.

[0096] The method in Embodiment 1 can be directly embodied as a hardware processor to complete, or a combination of hardware and software modules in the processor to complete. The software module can be located in a mature storage medium in the art such as random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, register, etc. The storage medium is located in the memory, and the processor reads the information in the memory to complete the steps of the above method in combination with its hardware. To avoid repetition, it will not be described in detail here.

[0097] Those of ordinary skill in the art can realize that the units of the examples described in combination with the present embodiment, i.e. the algorithm steps, can be realized in electronic hardware or in combination of computer software and electronic hardware. Whether the functions are executed in hardware or software mode depends on the specific application of the technical solution and the design constraints. The skilled person can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0098] The above describes the specific embodiments of the present application in combination with the drawings, but is not a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications or variations made by those skilled in the art on the basis of the technical solutions of the present application without creative labor are still within the protection scope of the present application.

Claims

1. A method for testing the assembly process of a valve train mechanism, characterized in that, include: Obtain the historical assembly tolerance values ​​of each component in the gas distribution mechanism, and construct a Bayesian network based on the relationship between the historical assembly tolerance values ​​of each component and the operating state of the gas distribution mechanism. The dragonfly algorithm based on the fusion of artificial fish behavior optimizes the Bayesian network. The dragonfly algorithm based on the fusion of artificial fish behavior is as follows: the foraging, swarming and tail-chasing behaviors in the artificial fish swarm algorithm are introduced into the dragonfly behavior of the dragonfly algorithm to update the dragonfly's movement direction. The score function of the Bayesian network is used as the fitness function of the dragonfly. The ratio of the difference between the fitness values ​​of the t-th iteration step and the (t+1)-th iteration step and the fitness value of the t-th iteration step is determined. The vision of the artificial fish swarm is combined with the dragonfly behavior to change the dragonfly's optimization range. An optimized Bayesian network was used to inspect the assembly process of the valve train by measuring the actual assembly tolerances of each component in the valve train.

2. The method for testing the assembly process of a valve train mechanism as described in claim 1, characterized in that, The components of the valve train include a camshaft, tappets, pushrods, rocker arms, intake valves, and exhaust valves. The operating states of the valve train include the early opening angle and late closing angle of the intake and exhaust valves. A Bayesian network is constructed based on the relationship between the early opening angle and late closing angle of the intake and exhaust valves and the assembly tolerance values ​​of the components in the valve train.

3. The method for testing the assembly process of a valve train mechanism as described in claim 1, characterized in that, Introducing foraging behavior from the artificial fish swarm algorithm into the dragonfly algorithm: Introducing the swarming behavior from the artificial fish swarm algorithm into the dragonfly algorithm: Introducing the tail-chasing behavior from the artificial fish swarm algorithm into the dragonfly algorithm: in, Let's consider a model where the i-th individual is attracted to food. Let's model the clustering behavior of the i-th individual. For the model where the i-th individual is farthest from the enemy, The location of the food. The enemy's location. For the location of the individual, To position the dragonfly as the central point of the artificial fish colony, , , As weight, It is a random number between [0,1], Step is the step size, and m is the total number of dragonflies.

4. The method for testing the assembly process of a valve train mechanism as described in claim 3, characterized in that, The dragonfly moves in the following direction: in, , Let f be the movement direction at iteration step t and iteration step (t+1), f be the food weight, c be the weight corresponding to the herding behavior, and e be the enemy weight. It is the inertia weight, and t is the number of iterations.

5. The method for testing the assembly process of a valve train mechanism as described in claim 1, characterized in that, Dragonfly optima search range with artificial fish swarm vision is ;in, To optimize the step size for range variation, The coefficient representing the variation range of the optimization range.

6. The method for testing the assembly process of a valve train mechanism as described in claim 5, characterized in that, The changes in the dragonfly's search range include: when When the value is less than the set threshold, the coefficient of variation of the optimization range is taken as follows: ; when When the value is not less than the set threshold, the coefficient of variation of the optimization range is taken as follows: ; Wherein, the fitness value of the i-th individual at the t-th iteration step is The fitness value of the i-th individual at iteration step t+1 is... ; Let's consider a model where the i-th individual is attracted to food. Let's model the clustering behavior of the i-th individual. Let f be the food weight, c be the weight corresponding to the herd behavior, and e be the enemy weight.

7. The method for testing the assembly process of a valve train mechanism as described in claim 1, characterized in that, In the process of optimizing Bayesian networks based on the dragonfly algorithm that incorporates artificial fish behavior, each Bayesian network is a dragonfly with the optimization characteristics of artificial fish. Under the scoring function of the Bayesian network, the best Bayesian network is found based on the score of the Bayesian network.

8. A valve train assembly process inspection system, characterized in that, include: The network construction module is configured to obtain the historical assembly tolerance values ​​of each component in the gas distribution mechanism, and construct a Bayesian network based on the relationship between the historical assembly tolerance values ​​of each component and the operating state of the gas distribution mechanism. The network optimization module is configured to optimize the Bayesian network based on the dragonfly algorithm that integrates artificial fish behavior. The dragonfly algorithm that integrates artificial fish behavior is as follows: the foraging, swarming, and tail-chasing behaviors in the artificial fish swarm algorithm are introduced into the dragonfly behavior of the dragonfly algorithm to update the dragonfly's movement direction. The score function of the Bayesian network is used as the fitness function of the dragonfly. The difference between the fitness values ​​of the t-th iteration step and the (t+1)-th iteration step is determined, and the fitness value of the t-th iteration step is compared with the fitness value of the t-th iteration step. The vision of the artificial fish swarm is combined with the dragonfly behavior to change the dragonfly's optimization range. The process inspection module is configured to inspect the assembly process of the valve train by using an optimized Bayesian network to measure the actual assembly tolerances of each component in the valve train.

9. An electronic device, characterized in that, It includes a memory and a processor, as well as computer instructions stored in the memory and running on the processor, which, when executed by the processor, perform the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, Used to store computer instructions, which, when executed by a processor, perform the method described in any one of claims 1-7.

Images