Method for classifying equipment exposure to environmental factors

Abstract

The application relates to an equipment exposure environment factor classification method, which comprises the following steps: S1, determining N types of natural environment factors constituting an equipment environment spectrum; S2, performing causal analysis between all environment factors in the environment spectrum, to obtain the causal relationship between the environment factors; S3, constructing a directed graph G=(N, E) of the N types of environment factors; S4, determining the coupling degree CD of the N types of environment factors; S5, judging whether the N types of environment factors are tightly coupled, and if yes, the next step is entered; S6, establishing a relationship matrix A between the environment factors; S7, calculating an adjacency matrix A'; S8, calculating a reachable matrix R; S9, establishing a hierarchical structure diagram of the N types of environment factors according to the reachable matrix, and further obtaining the input of an environment prediction model. The application is based on the environmental effect of equipment, quantitatively evaluates the coupling degree of environment factors by using directed graph theory, and determines the hierarchical relationship of tightly coupled environment factors by using an attribution analysis method, thereby providing a basis for selecting a suitable environment prediction method.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of environmental engineering, and particularly relates to a method for classifying exposure environmental factors of equipment. BACKGROUND

[0002] Equipment will be subjected to adverse effects of external climate, chemical and other environmental factors during service. To study such effects, a corresponding environmental spectrum needs to be established. The environmental spectrum is an important input for environmental adaptability examination and verification, life determination and life extension of equipment, and is based on different types of environmental data, and gives the action value, action time, action frequency and proportion of environmental factors that have significant effects on equipment, and quantitatively describes the environmental action process of equipment in a mission cycle through statistical and conversion methods.

[0003] The action value in the environmental spectrum represents the case that the equipment is subjected to extreme environmental action, but the environmental data used for compiling the environmental spectrum usually does not include the extreme state, so it is necessary to extrapolate and calculate the action value corresponding to the extreme state by using environmental prediction methods. Due to the difference in the region where the equipment is located, the corresponding environmental factors constituting the environmental spectrum are also different, and the mutual influence between different environmental factors cannot be ignored. For example, sand dust and corrosion medium in the air are two environmental factors that do not affect each other, and can be predicted separately; however, for relative humidity, it is significantly affected by solar radiation and temperature, solar radiation can cause temperature rise, temperature rise can cause the increase of saturated vapor pressure of humid air, and then cause the decrease of relative humidity, so the prediction of relative humidity needs to consider the three environmental factors comprehensively. However, for multiple environmental factors with strong coupling, the key prerequisite for carrying out environmental prediction is to clarify the cause-and-effect relationship between each environmental factor, so as to accurately define the input and output parameters in the environmental prediction model, otherwise it will lead to difficulty in solving the prediction model or calculation divergence, and the error of the prediction result is large, which cannot support the compilation of the environmental spectrum. SUMMARY

[0004] In view of the deficiencies in the prior art, the present application provides a method for classifying exposure environmental factors of equipment, which can firstly evaluate the coupling degree of environmental factors in the environmental spectrum by using directed graph theory, and then establish the hierarchical relationship of each factor by using attribution analysis method, so as to provide a basis for subsequent selection of appropriate environmental prediction method and determination of prediction model input and output.

[0005] To achieve the above-mentioned purpose, the present application discloses the following technical scheme:

[0006] Specifically, the present application provides a method for classifying exposure environmental factors of equipment, which comprises the following steps:

[0007] S1, according to the environmental profile of the equipment life cycle, combined with the composition structure and functional performance requirements of the equipment, screen the environmental factors that affect and act on the equipment, determine the N type natural environmental factors that constitute the equipment environment spectrum;

[0008] S2, based on the physical mechanism of each environmental factor, the causal analysis between all environmental factors in the environmental spectrum is carried out, and the causal relationship between the environmental factors is obtained, and the analysis criteria are as follows:

[0009] When the change of environmental factor A causes the change of environmental factor B simultaneously, then environmental factor A is the cause and environmental factor B is the effect, and the single-way symbol is used; when the change of environmental factor A causes the change of environmental factor B and the change of environmental factor B causes the change of environmental factor A simultaneously, then the two environmental factors are mutually causal, and the bidirectional symbol is used;

[0010] S3, according to the causal relationship between each environmental factor determined in step S2, a directed graph G=(N, E) of N type environmental factors is constructed, wherein: E represents the sum of the causal relationship between each environmental factor, wherein the single-way symbol is counted as 1, and the bidirectional symbol is counted as 2;

[0011] S4, the coupling degree CD of N type environmental factors is represented by the density of the directed graph G representing the causal relationship between each environmental factor. When the two-way symbol is used to represent the coupling degree of N type environmental factors, the sum of the causal relationship is expressed by the permutation formula , therefore, the calculation formula of the coupling degree CD is defined as:

[0012] ;

[0013] S5, set the value of b as the critical threshold value representing the coupling tightness, when b<CD, the N type environmental factor set is tightly coupled, otherwise it is loosely coupled, if the judgment result is tightly coupled, then enter the next step;

[0014] S6, when the N type environmental factor set is tightly coupled, the relationship matrix A between each environmental factor is established, wherein the matrix element represents that the environmental factor has a direct impact on the environmental factor , represents that the environmental factor has no impact on the environmental factor ;

[0015] S7, calculate the adjacency matrix , adjacency matrix The calculation formula is as follows:

[0016] ;

[0017] Wherein, A is the relation matrix, I is the unit matrix;

[0018] S8, power of the adjacency matrix is calculated until convergence, and the reachable matrix R is obtained; according to the reachable matrix R, the N types of environmental factors are decomposed to different levels, and a hierarchical structure diagram of the N types of environmental factors is established. In the hierarchical structure diagram of the N types of environmental factors, the environmental factors at the next level are the causes of the environmental factors at the previous level, and the topmost level is the total target, that is, the environmental factors at the next level are independent variables, and the environmental factors at the previous level are dependent variables;

[0019] S9, according to the hierarchical structure diagram of the N types of environmental factors, when a certain environmental factor is predicted, all the environmental factors at all levels below the environmental factor and having a connection with the environmental factor are taken as inputs of the environmental prediction model, and the environmental prediction model parameters are solved by using the determined sample corresponding to the inputs of the environmental prediction model and output as the prediction result of the environmental factor.

[0020] Preferably, when the numerical value is close to 1, the coupling degree indicates that each environmental factor tends to be tightly coupled, and the coupling degree is high; when the numerical value is close to 0, each environmental factor tends to be weakly coupled, and the coupling degree is low.

[0021] Preferably, the equipment life cycle environmental profile refers to a profile composed of the environments and environmental combinations / integrations and their time sequences that the equipment is subjected to.

[0022] Preferably, the environmental effect refers to the phenomenon that the materials, components and structural parts of the equipment are caused to fatigue, wear, corrode, age, degrade or degrade in performance, and the performance of the equipment is caused to decline or even lose the function under various single or integrated / combinational environmental effects in the life cycle of the equipment.

[0023] Preferably, the unit matrix I refers to a matrix with the same size as the relation matrix and diagonal elements of 1 and other elements of 0.

[0024] Preferably, the environmental factors in step S1 are temperature, relative humidity, solar radiation, wind, sand dust and acidic medium in the atmosphere.

[0025] Compared with the prior art, the present application has the following beneficial effects:

[0026] (1) The method of the present application aims at the problem of lacking analysis on the coupling degree of each environmental factor when compiling the environmental spectrum, quantitatively evaluates the coupling degree of environmental factors based on the environmental effect of equipment by using directed graph theory, and provides a basis for subsequent selection of appropriate environmental prediction method. When the environmental factors are tightly coupled, the input parameters of the environmental prediction model are determined, the samples corresponding to the input environmental factors are determined as the input to solve the model output, the final environmental prediction model is determined, the accuracy of the output result is ensured, the method can be applied to environmental factor prediction of various equipment in specific exposure and use process, and the method can be applied to classification of environmental factors of equipment exposure in multiple fields.

[0027] (2) The method of the present application aims at the strong coupling of various environmental factors, uses the attribution analysis method to perform structural analysis on all environmental factors, establishes the hierarchical relationship of all environmental factors, determines the input-output relationship between all environmental factors, according to the hierarchical structure diagram of N types of environmental factors, when predicting an environmental factor, all environmental factors under the environmental factor and having a connection with the environmental factor are taken as the input of the environmental prediction model, the corresponding input environmental data samples are used to solve the prediction model parameters and output as the environmental prediction result, which can well support the analysis and calculation work of the environmental prediction model. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 is a specific flowchart of the present application;

[0029] Figure 2 is a directed graph in the embodiment of the present application;

[0030] Figure 3 is a hierarchical structure diagram of environmental factors in the embodiment of the present application.

[0031] Figure 4 is a relative humidity environmental prediction result diagram in the embodiment of the present application. DETAILED DESCRIPTION

[0032] Exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the drawings represent functionally the same or similar elements. Although various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless specifically indicated.

[0033] Specifically, the present application provides an equipment exposure environmental factor classification method, which comprises the following steps:

[0034] S1, according to the environmental profile of the equipment life cycle, combining the composition structure and functional performance requirements of the equipment, screening the environmental factors which have significant influence and effect on the equipment, and determining N types of natural environmental factors constituting the environmental spectrum of the equipment.

[0035] The equipment life cycle environmental profile refers to a profile composed of the environment and environmental combination / integration and its time sequence description experienced by the equipment. The environmental effect refers to a phenomenon that the equipment performance is degraded or lost due to fatigue, wear, corrosion, aging, performance degradation or degradation of materials, components and structural parts of the equipment under various single or integrated / combined environmental effects during the life cycle of the equipment.

[0036] In a specific application, the determination process is based on the environmental effect of the equipment, for example, the equipment is composed of electronic components and is therefore more sensitive to temperature and humidity; the equipment is composed of metal materials and is therefore sensitive to corrosive media; the equipment is composed of rubber, plastic and other polymer materials and is therefore sensitive to solar radiation.

[0037] The environmental factors are temperature, relative humidity, solar radiation, wind, sand and dust, and acidic media in the atmosphere.

[0038] S2, based on the physical mechanism of forming each environmental factor, a causal analysis is performed between all environmental factors in the environmental spectrum, and the causal relationship between the environmental factors is obtained, and the analysis criteria are as follows:

[0039] When the change of a certain environmental factor A can simultaneously cause the change of a certain environmental factor B and the change of the environmental factor B cannot cause the change of the environmental factor A, i.e. only the change of one of the environmental factors can cause the change of the other environmental factor, it is determined that the environmental factor A is the cause and the environmental factor B is the effect, and a one-way symbol is used to represent; when the change of the environmental factor A simultaneously causes the change of the environmental factor B and the change of the environmental factor B simultaneously causes the change of the environmental factor A, i.e. when the two environmental factors change, they can cause the change of the other, it is determined that the two environmental factors are cause and effect of each other, and a two-way symbol is used to represent.

[0040] S3, according to the causal relationship between the environmental factors determined in step S2, a directed graph G=(N, E) of N types of environmental factors is constructed, wherein: E represents the sum of the causal relationships between the environmental factors, wherein the one-way symbol is counted as 1, and the two-way symbol is counted as 2.

[0041] S4, the coupling degree CD of the N types of environmental factors is represented by the density of the count of the directed compliance representing the causal relationship of the environmental factors in the directed graph G, when the two-way symbol is used to represent the causal relationship between the environmental factors, at this time, the coupling degree of the N types of environmental factors is the highest, the sum of the causal relationships is expressed by the permutation formula , therefore, the calculation formula of the coupling degree CD is defined as:

[0042] ;

[0043] S5, set b value as a critical threshold value representing coupling tightness, when b < CD, the N-class environment factor set is tightly coupled, otherwise it is loosely coupled. If the judgment result is tight coupling, go to the next step. If it is judged as loose coupling, the input and output of the environment data are relatively clear, and the input of the environment factor prediction model can be directly predicted.

[0044] In specific applications, the coupling degree is distinguished by numerical values 0 and 1. When the numerical value is closer to 1, it indicates that each environment factor tends to be tightly coupled, and the coupling degree is high. When the numerical value is close to 0, it indicates that each environment factor tends to be weakly coupled, and the coupling degree is low. If it is judged that the environment factors are tightly coupled, the next step is performed.

[0045] S6, when the N-class environment factor set is tightly coupled, a relationship matrix A between the environment factors is established, wherein the matrix element represents that the environment factor has a direct impact on the environment factor . represents that the environment factor has no impact on the environment factor .

[0046] S7, calculate the adjacency matrix , the calculation formula of the adjacency matrix is:

[0047] ;

[0048] Wherein, A is the relationship matrix, I is the unit matrix; in specific applications, the unit matrix I refers to a matrix with the same size as the relationship matrix and diagonal elements of 1 and other elements of 0.

[0049] S8, calculate the power of the adjacency matrix until convergence, obtain the reachable matrix R; according to the reachable matrix R, decompose the N-class environment factors into different levels, and establish the hierarchical structure diagram of the N-class environment factors. In the hierarchical structure diagram of the N-class environment factors, the environment factors in the next layer are the reasons for the environment factors in the last layer, and the top layer is the total target, that is, the next layer environment factors are independent variables, and the last layer environment factors are dependent variables; in specific applications, the attribution analysis method is used to analyze the structure of all environment factors, establish the hierarchical relationship of all environment factors, and clarify the input-output relationship between all environment factors.

[0050] S9, according to the hierarchical structure diagram of the N-class environment factors, when predicting a certain environment factor, all environment factors under the environment factor and having a connection with the environment factor are taken as the input of the environment prediction model, and the environment data samples corresponding to the input are used to solve the prediction model parameters and output as the environment prediction result.

[0051] Another aspect of the present application also provides an equipment exposure environment factor classification system, which comprises a natural environment factor determination unit, a causality analysis unit, a directed graph construction unit, a coupling degree calculation unit, a coupling degree judgment unit, a relation matrix construction unit, an adjacency matrix calculation unit, a reachable matrix calculation unit and an environment prediction model input determination unit.

[0052] The natural environment factor determination unit is used to determine N types of natural environment factors constituting an equipment environment spectrum; the causality analysis unit performs causality analysis between all environment factors in the environment spectrum to obtain causality between the environment factors; the directed graph construction unit is used to construct a directed graph G=(N, E) of the N types of environment factors; the coupling degree calculation unit is used to determine the coupling degree CD of the N types of environment factors; the coupling degree judgment unit judges whether the N types of environment factors are tightly coupled, and if so, the next step is entered; in this unit, the value of b is set as a critical threshold value representing the coupling tightness, and when b<CD, the N types of environment factor sets are tightly coupled, otherwise, they are loosely coupled. If the result of the judgment is tight coupling, the next step is entered. If the result of the judgment is loose coupling, the input of the environment factor prediction model can be directly predicted because the input and output of the environment data are relatively clear. The relation matrix construction unit is used to establish a relation matrix A between the environment factors; the adjacency matrix calculation unit is used to calculate an adjacency matrix ; the reachable matrix calculation unit is used to calculate a reachable matrix R, and according to the reachable matrix R, the N types of environment factors are decomposed into different levels to establish a hierarchical structure diagram of the N types of environment factors. In the hierarchical structure diagram of the N types of environment factors, the environment factors in the next level are the causes of the environment factors in the previous level, and the topmost level is the total target, i.e., the environment factors in the next level are independent variables, and the environment factors in the previous level are dependent variables; the environment prediction model input determination unit is used to, according to the hierarchical structure diagram of the N types of environment factors, when predicting a certain environment factor, take all the environment factors in all levels below the environment factor and having a connection with the environment factor as the input of the environment prediction model, solve the prediction model parameters using the corresponding input environment data samples and output as the environment prediction result.

[0053] Meanwhile, the present application also provides a computer device, which is installed with the above-mentioned equipment exposure environment factor classification system. Specific embodiments

[0054] The embodiment of the present application provides a helicopter equipment exposure environment factor classification method, which specifically comprises the following steps:

[0055] S1, a certain type of helicopter is deployed in a certain area, because the service life of a certain type of helicopter is mainly parked, so the natural environment spectrum of a certain type of helicopter parking phase is compiled. The certain area is a coastal industrial area, and the atmosphere contains acidic medium formed by industrial activities to form acidic atmosphere, analyze the environmental effects of all natural environmental factors on a certain type of helicopter, and determine that the parking environment spectrum of a certain type of helicopter consists of 6 types of environmental factors including temperature, relative humidity, solar radiation, wind, sand dust and acidic medium in the atmosphere.

[0056] S2, define 6 types of environmental factors as F1~F6. Based on the physical mechanism of forming each environmental factor, the causal analysis between all N types of environmental factors in the environmental spectrum is carried out, as follows

[0057] a) solar radiation will cause temperature rise, use "F1 " to represent;

[0058] b) temperature, solar radiation, relative humidity and wind speed change will cause sand dust concentration change, use "F1 ", "F2 ", "F3 ", "F4 " to represent;

[0059] c) temperature and wind speed change will cause relative humidity change, use "F1 ", "F4 " to represent;

[0060] d) wind speed change will cause the concentration of acidic medium in the atmosphere to change, use "F4 " to represent.

[0061] S3, according to the dependence relationship of 6 types of environmental factors in the environmental spectrum, a directed graph G=(6, E) is constructed, where: E represents the dependence relationship between environmental factors, the number of edges is |E|, as shown in the accompanying Figure 1 , wherein |E|=8.

[0062] S4, calculate the coupling degree CD of 6 types of environmental factors, the formula is as follows:

[0063] ;

[0064] S5, define the critical threshold b=0.2 of the coupling tightness degree of environmental factors, since CD=0.27>0.2, 6 types of environmental factors are tightly coupled.

[0065] S6, establish the relationship matrix A between 6 types of environmental factors, as shown below:

[0066] ;

[0067] Add the unit matrix I to the relationship matrix A to obtain the adjacency matrix i.e.

[0068] ;

[0069] S7, calculate the power of the adjacency matrix until convergence, obtain the reachable matrix R, i.e.

[0070] ;

[0071] S8, according to the reachable matrix R, decompose the six types of environmental factors into different levels, establish a hierarchical structure diagram of the six types of environmental factors, a total of five levels, as shown in the accompanying Figure 2

[0072] S9, for the hierarchical structure diagram of the six types of environmental factors, when predicting a certain environmental factor, all the environmental factors under the level and connected to it are taken as the input of the environmental prediction model, and the environmental data samples corresponding to the input are used to solve the prediction model parameters and output as the environmental prediction result.

[0073] Predict the relative humidity inside the cockpit of a certain type of helicopter. According to the hierarchical structure diagram, the relative humidity is at the third level, and the environmental factors at the fourth and fifth levels are temperature and solar radiation, respectively, so temperature and solar radiation are the input parameters for relative humidity prediction.

[0074] An Elman neural network method is used to model and predict the cabin relative humidity environment. The external solar radiation intensity, external atmospheric temperature, external relative humidity and cabin temperature are selected as the input matrix X(t), and the cabin relative humidity is the output matrix Y(t). When establishing the network model, the network topology structure is set to 4-5-1 according to the number of input and output variables and engineering experience, i.e. the number of input layer nodes is 4, the number of hidden layer nodes is 5, and the number of output layer nodes is 1.

[0075] The measured data is used to train the network, and the number of iterations is set to 10000. With the increase of training times, the gradient of cost function decreases rapidly, and when the network training is completed, it has decreased to 0.1, at this time the learning rate is 0.01, which has reached a small value. The mean square error of training is 2.62.

[0076] Select environmental data that does not participate in modeling to verify the accuracy of the model. The external atmospheric temperature, external relative humidity and cabin air temperature are input, and the relative humidity in the cockpit is output. The relative humidity environmental prediction result of a certain day is shown in the accompanying Figure 4

[0077] ​​​The embodiment of the present application adopts the attribution analysis method to perform structural analysis on all the environmental factors, establishes the hierarchical relationship of all the environmental factors, and clearly defines the input-output relationship between all the environmental factors. According to the hierarchical structure diagram of the N types of environmental factors, when a certain environmental factor is predicted, all the environmental factors under the certain environmental factor and having a connection with the certain environmental factor are taken as the input of the environmental prediction model. The environmental data samples corresponding to the input are used to solve the prediction model parameters and output as the environmental prediction result, which can well support the analysis and calculation of the environmental prediction model.

[0078] The above-described embodiments are merely preferred embodiments of the present application, and are not intended to limit the scope of the present application. Various modifications and improvements to the technical solutions of the present application made by those of ordinary skill in the art without departing from the design spirit of the present application shall fall within the protection scope of the present application as defined by the claims.

Claims

1. An apparatus for classifying exposure environmental factors, the method comprising: It comprises the following steps: S1, according to the environmental profile of the equipment life cycle, combined with the composition structure and functional performance requirements of the equipment, screen the environmental factors that affect and act on the equipment, determine the N type natural environmental factors that constitute the equipment environment spectrum; S2, based on the physical mechanism of each environmental factor, the causal analysis between all environmental factors in the environmental spectrum is carried out, and the causal relationship between the environmental factors is obtained, and the analysis criteria are as follows: When the change of environmental factor A causes the change of environmental factor B, then the environmental factor A is determined as the cause and the environmental factor B as the effect, and a single-headed arrow is used to represent When the change of environmental factor A causes the change of environmental factor B and the change of environmental factor B causes the change of environmental factor A, then the two environmental factors are determined as cause and effect of each other, and a double-headed arrow is used to represent ​ S3. According to the causal relationship between each environmental factor determined in step S2, a directed graph G=(N, E) of N types of environmental factors is constructed, wherein: E represents the sum of the causal relationships between each environmental factor, wherein the single-headed symbol counts as 1, and the double-headed symbol counts as 2. S4. Characterizing the coupling degree CD of N environmental factors with the density of the directed consistent count representing the causal relationship of each environmental factor in the directed graph G, when the two-way sign between each pair of N environmental factors The symbol represents the sum of the causal relationships of N environmental factors, which is calculated by the permutation formula Expression, therefore, the calculation formula of the coupling degree CD is defined as: ； S5, set b value as critical threshold value representing coupling tightness, when b < CD, N type environmental factor set is tight coupling, otherwise is loose coupling, if the judgment result is tight coupling, then enter the next step; S6、When the N sets of environmental factors are tightly coupled, a relationship matrix A between the environmental factors is established, where the matrix elements represent that the environmental factor has a direct influence on the environmental factor . represent that the environmental factor has no influence on the environmental factor . S7, computing the adjacency matrix adjacency matrix The formula for computing the adjacency matrix is ； Wherein, A is the relationship matrix, I is the unit matrix; S8、calculating the adjacency matrix power of the adjacency matrix until convergence to obtain the reachable matrix R; according to the reachable matrix R, decomposing the N types of environmental factors into different levels to establish a hierarchical structure diagram of the N types of environmental factors. In the hierarchical structure diagram of the N types of environmental factors, the environmental factors at the next level are the causes of the environmental factors at the previous level, and the topmost level is the overall goal, i.e. the environmental factors at the next level are the independent variables and the environmental factors at the previous level are the dependent variables. S9, according to the hierarchical structure diagram of N type environmental factors, when predicting a certain environmental factor, all environmental factors under the environmental factor and having contact with the environmental factor are taken as the input of the environmental prediction model, and the sample corresponding to the determined input of the environmental prediction model is used to solve the environmental prediction model parameters and output as the prediction result of the environmental factor.

2. The method of claim 1, wherein: The coupling degree, when the numerical value is close to 1, indicates that each environmental factor tends to be tight coupling, and the coupling degree is high; When the numerical value is close to 0, it indicates that each environmental factor tends to be weak coupling, and the coupling degree is low.

3. The method of claim 1, wherein: The equipment life cycle environmental profile refers to the profile composed of the environment and the combination / composition of the environment and its time sequence description that the equipment is subjected to.

4. The method of claim 1, wherein: The unit matrix I refers to the matrix with the same size as the relationship matrix, and the diagonal elements are 1 and the other elements are 0.

5. The method of claim 1, wherein: The environmental factors in step S1 are temperature, relative humidity, solar radiation, wind, sand dust and acidic medium in the atmosphere.

Images