CN110135040B - 3K planetary reducer reliability evaluation method based on neural network - Google Patents

Abstract

The invention proposes a reliability evaluation method of a 3K planetary reducer based on a neural network. From the perspective of Monte Carlo simulation method, the method defines the failure properties of the components of the 3K planetary reducer system based on the neuron structure, and then establishes a simplified system reliability hierarchical network model without components based on the neural network structure. , and then consider the common cause failure to determine the system reliability assessment strategy, and finally establish a Monte Carlo simulation method that matches the reliability hierarchical network model. The invention alleviates the error caused by the simplification in the reliability modeling of the 3K planetary reducer by using the traditional reliability evaluation method, and avoids the difficulty of solving multiple integrals when calculating the reliability of a mechanical system with common cause failure by the full probability model method. It provides a basis for designers to evaluate the reliability of a system with a large number of components and a complex structure, and improves the accuracy of system reliability evaluation.

Description

Reliability evaluation method of 3K planetary reducer based on neural network

technical field

The invention relates to the field of reliability evaluation of an engineering transmission system, in particular to a reliability evaluation method of a 3K planetary reducer based on a neural network, which can be applied to occasions where the 3K planetary reducer is applied such as aerospace.

Background technique

In recent years, 3K planetary reducer (K: center wheel) has been more and more frequently used in various fields of the machinery industry due to its compact structure, light weight, large transmission ratio and high efficiency. Especially in the aerospace field, the 3K planetary reducer is widely used in the flaps and slats of the aircraft. It is responsible for opening the flaps and slats during take-off and landing to improve lift; retracting the flaps and slats during normal flight at high altitudes to reduce drag. However, in the 3K planetary reducer of aircraft flaps and slats, there are many parts, complex structure and many system failure factors, and its reliability is very high due to safety considerations. Therefore, the correct evaluation of the reliability of the 3K planetary reducer plays a key role in ensuring the safe and smooth take-off, cruise and landing of the aircraft.

At present, there are two main methods for evaluating the reliability of the transmission system with a large number of parts and a complex structure, such as 3K planetary reducer. It is a classic mechanical system reliability evaluation method based on stress intensity interference theory and parts series-parallel theory. The reliability of parts is calculated through stress intensity interference, and then the system reliability is calculated through series-parallel theory. The first method is difficult to achieve in terms of time and financial resources due to the complex production and processing of 3K planetary reducer, high cost, long test time, and large sample size. The second type requires a lot of simplification of the system during the implementation process, and only retains the main components that cause the system to fail, which will inevitably lead to overestimation of the reliability calculation value. For high-reliability products such as aerospace equipment or precision machinery, There is a certain risk, in addition, the series-parallel model method assumes that the failures of components are independent of each other, but in fact the common cause failure represented by the load is widespread and will cause errors in the reliability calculation value. Therefore, the reliability modeling of the system components of the 3K planetary reducer is carried out without simplification, and the reliability assessment is carried out under the condition of common cause failure, in order to better analyze and improve the reliability of the 3K planetary reducer, and ensure the precision mechanical transmission. The stable and reliable work of the system has important theoretical and practical significance.

SUMMARY OF THE INVENTION

Aiming at the defects of the existing reliability assessment methods, the present invention provides a reliability assessment method for 3K planetary reducer based on neural network. Based on the system layered reliability model, considering the common cause failure of the system, from the perspective of Monte Carlo simulation, a reliability simulation strategy of 3K planetary reducer is formulated to predict the reliability of the system, so as to facilitate the next arrangement of operation and maintenance. , improve the reliability of system operation.

The technical scheme of the present invention is:

Described a kind of 3K planetary reducer reliability assessment method based on neural network is characterized in that: comprises the following steps:

Step 1: Define the failure of each component of the 3K planetary reducer based on the neuron structure. The failure definition includes three attributes, namely input attribute, failure trigger attribute and output attribute; the failure trigger attribute includes external failure trigger conditions and internal failure triggering conditions;

Step 2: After defining the failure of each part, determine the internal failure mechanism, external failure mechanism and failure propagation mechanism of each part;

Step 3: Based on the primary and secondary structure of the neural network, according to the assembly relationship between the components and the importance in the power transmission process, the reliability of the 3K planetary reducer is modeled in layers, and the reliability network diagram and reliability network diagram are obtained. There are Z layers in China: the first layer is the core components of the 3K planetary reducer, the second layer is the secondary components that directly affect the core components, the third layer is the components that directly affect the secondary components, and so on. The stratification of each component in the 3K planetary reducer; establish a pointing sequence for the components of each stratum according to the direction of power transmission;

Step 4: Take the failure rate of the 3K planetary reducer as the evaluation index, initialize the number of simulations N=1, and initialize the cumulative safety number of the system n=0;

Step 5: According to the common cause failure Monte Carlo simulation strategy, the common cause of failure is uniformly sampled and then transferred to each component, and the non-failure common cause parameters are independently sampled according to the parameter distribution;

Step 6: According to the internal failure triggering and external failure triggering conditions of each component, update each layer sequentially in the order of layer Z, layer Z-1, layer Z-2, ... layer 2, layer 1 The component state T _i ; the component state is divided into two states: safety and failure;

Step 7: Determine whether the last component in the first layer of the reliability network diagram pointing to the sequence fails, if not, set the system state variable k=1, indicating that the system is safe; otherwise, k=0, indicating that the system has failed;

Step 8: Update the number of simulations N=N+1 and the cumulative number of safety systems n=n+k;

Step 9: Judging whether the number of simulations satisfies N<Num, Num represents the total number of simulations set, if so, return to step 5 to continue the iteration, otherwise terminate the program and count the cumulative safety times n of the system, and calculate the reliability R=n/Num, The failure rate Y=1-R, compare the failure rate with the established failure probability, and determine the operating state of the 3K planetary reducer according to the result.

A further preferred solution, the described a kind of 3K planetary reducer reliability evaluation method based on neural network, is characterized in that: in step 1, the input attribute includes the state input k, load F, vibration x, torque T of parts and components; output The attribute refers to the output state; the external failure triggering condition refers to the condition that another component connected to the component and at the upper level of the system power transmission fails and the component cannot receive the power of the system; the internal failure triggering condition refers to the component itself of each failure mode.

beneficial effect

1. For the mechanical transmission system with a large number of components and complex relationships between components, which is difficult to model by the series-parallel method, create a reliability modeling method for complex systems without component simplification based on neural network structure, so that the established reliability is reliable. Sexual models have high accuracy.

2. Starting from the Monte Carlo method and based on the full probability model of reliability, a Monte Carlo simulation strategy for common cause failure is established, which reduces the error caused by the simplification of the reliability modeling of the 3K planetary reducer by using the traditional reliability evaluation method, and avoids the The problem that multiple integrals are difficult to solve when calculating the reliability of mechanical systems with common cause failures by the full probability model method makes the system reliability calculation more efficient and accurate.

Additional aspects and advantages of the present invention will be set forth, in part, from the following description, and in part will be apparent from the following description, or may be learned by practice of the invention.

Description of drawings

The above and/or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of embodiments taken in conjunction with the accompanying drawings, wherein:

Fig. 1 is the flow chart of the method of the present invention.

Figure 2 is a diagram of component failure definition.

Figure 3 is a schematic diagram of the transmission structure of the 3K planetary reducer.

Figure 4 is the assembly drawing of the parts of the 3K planetary reducer.

Figure 5 is a reliability network diagram including core components.

Figure 6 is a reliability network diagram including minor components.

Figure 7 is a reliability network diagram that includes all components.

Detailed ways

The embodiments of the present invention are described in detail below. This embodiment takes a large aircraft slat 3K planetary reducer as an example to illustrate the specific implementation of the reliability evaluation method for a 3K planetary reducer. The described embodiments are intended to explain the present invention, and cannot It is construed as a limitation of the present invention.

Figure 1 shows the flow chart of the method of the present invention. The transmission structure diagram of the 3K planetary reducer of the large aircraft slat selected this time is shown in Figure 3. Input from N3, output from N4, the input speed of the main shaft n _N3 = 1500r/min , the output torque is T _e =94.26N·m, and the design life is 2000h. In this embodiment, the structure is analyzed to determine the rotational speed and bearing load of the main gears, shafts and other components. The basic parameters of the slat 3K reducer are shown in Table 1.

Table 1 Introduction of basic parameters of slat 3K planetary reducer

输入转速n_N3＝1500r/min，结合附图3可以计算输入转矩、各个齿轮转速和啮合力。(本实施例重在解释本发明方法，此处输入转矩、各个齿轮转速和啮合力的计算过程均省略。)其中n_N4＝17.86r/min，n_N5＝1155.77r/min。Known total gear ratio

The input rotational speed n _N3 =1500r/min, the input torque, the rotational speed of each gear and the meshing force can be calculated in conjunction with FIG. 3 . (This embodiment focuses on explaining the method of the present invention, and the calculation process of input torque, rotational speed of each gear and meshing force is omitted here.) Among them, n _N4 =17.86r/min, n _N5 =1155.77r/min.

Input torque:

but:

T _e =T _N4 =58.5T _a

Normal shear stress at the meshing of N3 and N2:

Normal shear stress at the meshing of N1 and N2:

Normal shear stress at the meshing of N4 and N5:

其中np为行星轮个数3。

Where np is the number of planetary wheels 3.

Step 1: Define the failure of each component of the 3K planetary reducer based on the neuron structure. The failure definition includes three attributes, namely input attribute, failure trigger attribute and output attribute; the failure trigger attribute includes external failure trigger Conditions and internal failure trigger conditions; the schematic diagram of component failure definition is shown in Figure 2.

The input attributes include the state input k, load F, vibration x, and torque T of the component; the output attribute refers to the output state; the external failure triggering condition indicates that another component connected to the component and at the upper level of the system power transmission fails. The conditions that cause the component to fail to receive system power; internal failure triggering conditions represent the component's own failure modes.

In this embodiment, the input attribute is selected as load, and the output attribute is two states of component safety and failure.

Step 2: After defining the failure of each part, determine the internal failure mechanism, external failure mechanism and failure propagation mechanism of each part.

In this embodiment, the internal failure mechanism, external failure mechanism and failure propagation mechanism of each part are analyzed according to the components of the 3K planetary reducer (Table 2) and FIG. 4 .

Table 2 Parts list of slat 3K planetary reducer

(1) No. 10 (shaft 2, material 40CrNiMoA): shaft 2 is the power input shaft, which is only affected by torsional shear stress after analysis.

External failure trigger mechanism: failure of parts 9, 18, 37 will cause them to fail;

Internal failure trigger mechanism: the torsional shear stress is greater than the torsional strength, resulting in failure;

Failure propagation mechanism: Failure of part 10 will cause failure of part 31.

The internal failure evaluation is based on the distribution sampling of the input load T _a and the value of the diameter D to calculate the maximum stress and torsional strength at the dangerous place, and then determine whether the shaft fails or not after comparing the magnitudes. in:

Maximum stress at hazard

The torsional strength σ _τ can be obtained by referring to the relevant chapters of the "Mechanical Design Manual" to obtain its distribution: σ _τ ～N (576.34, 86.45 ² ).

(2) No. 31 (gear N3, material 38CrMoAl), gear 31 is the center wheel in the planetary transmission.

External failure trigger mechanism: failure of parts 10 and 16 will cause them to fail;

Internal failure trigger mechanism: including tooth surface contact fatigue failure and tooth root bending fatigue failure;

Failure Propagation Mechanism: Failure of Part 31 will cause Part 33 to fail.

The internal failure evaluation calculates the contact stress and bending stress respectively, and then compares them with the contact strength and bending strength. When the strength of both is greater than the stress, the gear is judged to be safe, otherwise, as long as one of the stresses is greater than the strength, the gear is judged to fail. in:

Tooth contact stress

Tooth contact strength

σ _HG =σ _Hlim Z _NT Z _L Z _V Z _R Z _W Z _X

tooth root bending stress

Root bending strength

σ _FG =σ _Flim Y _ST Y _NT Y _δrelT Y _RrelT Y _X

The above tangential force F _t is known, the contact stress and strength calculation parameters of the tooth surface, and the calculation parameters of the contact stress and strength of the tooth surface can be obtained by referring to the relevant chapters of the "Mechanical Design Manual" to obtain their distribution.

(3) No. 33 (planetary shafts N2, N5), the No. 33 part is a planetary wheel shaft, which is integrated with the gears N2 and N5, so its failure includes the failure of the two gears. And because the 3K planetary actuator includes three planetary gears, the three planetary gears are treated as a 2/3 voting system from a practical point of view.

External failure trigger mechanism: the failure of parts 31 and 46 will cause the voting system to fail, and the failure of parts 29 and 30 will cause the planetary gear to fail and conduct 2/3 voting;

Internal failure trigger mechanism: including tooth surface contact fatigue failure and tooth root bending fatigue failure of gears N2 and N5;

Failure Propagation Mechanism: Failure of the voting system will cause Part 47 to fail.

The internal failure evaluation of planetary gears N2 and N5 calculates the contact stress and bending stress respectively, and then compares them with the contact strength and bending strength. When the strength of both is greater than the stress, the gear is judged to be safe, otherwise as long as one of the stresses is greater than the strength, the gear is judged to fail. . The calculation process of the contact stress, contact strength, bending stress and bending strength of the planetary gears N2 and N5 is similar to that of the No. 31 part (gear N3), and will not be described here.

(4) No. 47 (gear N1), the ring gear in the planetary transmission of gear N1.

External failure trigger mechanism: failure of the planetary axle 2/3 voting system will cause it to fail;

Internal failure trigger mechanism: including tooth surface contact fatigue failure and tooth root bending fatigue failure;

Failure Propagation Mechanism: Failure of Part 47 will cause Part 11 to fail.

The internal failure evaluation calculates the contact stress and bending stress respectively, and then compares them with the contact strength and bending strength. When the strength of both is greater than the stress, the gear is judged to be safe, otherwise, as long as one of the stresses is greater than the strength, the gear is judged to fail. The calculation process of the contact stress, contact strength, bending stress, and bending strength of the gear N1 is similar to that of the No. 31 part (gear N3), and will not be described here.

(5) No. 11 (gear N4).

External failure trigger mechanism: failure of parts 47 and 34 will cause them to fail;

Internal failure trigger mechanism: including tooth surface contact fatigue failure and tooth root bending fatigue failure.

Failure Propagation Mechanism: Failure of Part 11 will cause Part 4 to fail.

The internal failure evaluation calculates the contact stress and bending stress respectively, and then compares them with the contact strength and bending strength. When the strength of both is greater than the stress, the gear is judged to be safe, otherwise, as long as one of the stresses is greater than the strength, the gear is judged to fail. The calculation process of the contact stress, contact strength, bending stress and bending strength of the gear 11 is similar to that of the No. 31 part (gear N3 ), and will not be described here.

(6) No. 4 (No. 1 output shaft): Part 4 is the No. 1 output shaft, which is only affected by torsional shear stress after analysis.

External failure trigger mechanism: failure of parts 7 and 11 will cause them to fail;

Internal failure trigger mechanism: the torsional shear stress is greater than the torsional strength, resulting in failure;

Failure propagation mechanism: Part 4 failure will cause Part 1 failure.

The internal failure evaluation process of part 4 is similar to that of part 10 (axis 2), and will not be described here.

(7) No. 1 (No. 2 output shaft): Part 1 is the No. 2 output shaft, which is only affected by torsional shear stress after analysis.

External failure trigger mechanism: failure of parts 4, 5, 43, 45 will cause them to fail;

Internal failure trigger mechanism: the torsional shear stress is greater than the torsional strength, resulting in failure;

Failure Propagation Mechanism: Failure of Part 1 will cause the entire actuator to fail.

The internal failure evaluation process of part 1 is similar to that of part 10 (axis 2), and will not be described here.

(8) No. 9 and No. 37 (bearings): Bearings are standard parts.

External failure trigger mechanism: failure of parts 36, 38, 39 will cause them to fail;

Internal failure trigger mechanism: the bearing is a standard part and an assembly. The reliability of this calculation can be calculated by the relationship between the rated life reliability (known at the factory and is 0.9) and the design life of the actuator, so its internal failure Sampling can be done directly according to reliability;

Failure propagation mechanism: failure of part 9 and part 37 will lead to failure of part 10.

The internal failure assessment process of the bearing is as follows:

Reliability calculation formula:

It can be calculated according to the design life of the 3K planetary reducer and related parameters of the bearing (which can be obtained from the "Mechanical Design Manual").

(9) No. 18 (bearing): The bearing is a standard part.

External failure trigger mechanism: failure of parts 19, 20, 22, 24, 26 will cause them to fail;

Internal failure trigger mechanism: the bearing is a standard part and an assembly, and the failure causes are more complicated. Its reliability can be calculated by the relationship between the rated life reliability (known at the factory and is 0.9) and the design life of the actuator, so its Internal failures can be sampled directly according to reliability;

Failure Propagation Mechanism: Failure of Part 18 will cause Part 10 to fail.

The internal failure evaluation process of part 18 is similar to that of part 9 (bearing), and will not be described here.

(10) No. 16 (retaining ring): The elastic retaining ring is a standard part.

External failure trigger mechanism: none;

Internal failure trigger mechanism: the retaining ring is a standard part, which is mainly used to limit the axial movement of the gear. The reliability is very high when the gear is not subjected to axial force. The reliability can be set to 0.999, and then random sampling method to determine its state at any time.

Failure propagation mechanism: failure of part 16 will cause failure of part 31.

(11) No. 29 (bearing): The bearing is a standard part.

External failure trigger mechanism: none;

Internal failure trigger mechanism: the bearing is a standard part and an assembly, and the failure causes are more complicated. Its reliability can be calculated by the relationship between the rated life reliability (known at the factory and is 0.9) and the design life of the actuator, so its Internal failures can be sampled directly according to reliability;

Failure propagation mechanism: Failure of part 29 will cause failure of part 33.

The internal failure evaluation process of part 29 is similar to that of part 9 (bearing), and will not be described here.

(12) No. 30 (copper sleeve), No. 34 (slotted long cylindrical end set screw M5×6): can be regarded as standard parts.

External failure trigger mechanism: none;

Internal failure trigger mechanism: copper sleeves and screws are highly reliable, and the reliability can be set to 0.999, and then the state at any time is determined according to the method of random sampling;

Failure propagation mechanism: failure of part 30 will cause failure of part 33, failure of part 34 will cause failure of part 11.

(13) No. 46 (planet carrier)

External failure trigger mechanism: failure of parts 12, 13, 14, 15, 17, 32 will cause failure of part 46;

Internal failure trigger mechanism: the planet carrier itself is highly reliable, and the reliability can be set to 0.999; then determine its state at any time by random sampling;

Failure propagation mechanism: Failure of part 46 will cause failure of part 33.

(14) No. 7 (bearing): The bearing is a standard part.

External failure trigger mechanism: failure of parts 8 and 40 will cause them to fail;

Internal failure trigger mechanism: the bearing is a standard part and an assembly, and the failure causes are more complicated. Its reliability can be calculated by the relationship between the rated life reliability (known at the factory and is 0.9) and the design life of the actuator, so its Internal failures can be sampled directly according to reliability;

Failure propagation mechanism: Failure of part 7 will cause failure of part 4.

The internal failure evaluation process of part 7 is similar to that of part 9 (bearing), and will not be described here.

(15) No. 5 (spline sleeve), No. 45 (gasket)

External failure trigger mechanism: none;

Internal failure trigger mechanism: The reliability of the spline sleeve and the gasket itself is very high, and its reliability can be set to 0.999; then determine its state at any time according to the method of random sampling;

Failure propagation mechanism: failure of part 5 and part 45 will lead to failure of part 1.

(16) No. 43 (bearing): The bearing is a standard part.

External failure trigger mechanism: failure of parts 2, 3, 41, 42, 44 will cause them to fail;

Internal failure trigger mechanism: the bearing is a standard part and an assembly, and the failure causes are more complicated. Its reliability can be calculated by the relationship between the rated life reliability (known at the factory and is 0.9) and the design life of the actuator, so its Internal failures can be sampled directly according to reliability;

Failure Propagation Mechanism: Failure of Part 43 will cause Part 1 to fail.

The internal failure evaluation process of part 43 is similar to that of part 9 (bearing), and will not be described here.

The above only shows the key components that directly affect the power transmission, and the components that act on the key components to assist in the power transmission (here called secondary components). The failure triggering mechanism and failure propagation mechanism of the remaining parts are the same as those analyzed above similar.

Step 3: Based on the primary and secondary structure of the neural network, according to the assembly relationship between the components and the importance in the power transmission process, the reliability of the 3K planetary reducer is modeled in layers, and the reliability network diagram and reliability network diagram are obtained. There are Z layers in China: the first layer is the core components of the 3K planetary reducer, such as input shaft, output shaft, intermediate transmission shaft, sun gear, planetary gear, ring gear, etc., and the second layer is the second layer that directly affects the core components. The main parts, such as bearings, housings, retaining rings, sleeves, spline sleeves, etc., the third layer is the parts that directly affect the secondary parts, such as planet carriers, screws, nuts, washers, end caps, etc., in order By analogy, determine the layer to which each component in the 3K planetary reducer belongs. Establish a pointing sequence for the components of each layer according to the direction of power transmission; the arrow in the reliability neural network diagram of the core components of the first layer of the 3K planetary reducer represents the direction of power transmission and also the direction of failure propagation, and the arrow points to the order of the last component. The state represents the system state, that is, if the failure of the last component in the first layer is not triggered, it means that the power transmission of the system is normal and the system is safe. On the contrary, if the failure of the last component is triggered, it means that the power transmission of the system fails and the system fails; The arrows in layers 3, .

For this embodiment, from the perspective of power transmission, the main power transmission components of the slat 3K planetary reducer include transmission shafts and gears. The reliability network diagram (layer 1) of the core components of the slat 3K planetary reducer is as follows As shown in Figure 5, 10 is the transmission shaft, 31 is the gear N3, 33 is the planetary gear, that is, includes the gear N2 and the gear N5, 47 is the gear N1, 11 is the gear N4, 4 is the No. 1 output shaft, and 1 is No. 2. Output shaft.

The secondary components that directly act on the key components to assist in power transmission belong to the second layer of the reliability network diagram of the slat 3K planetary reducer, and the reliability network diagram including the secondary components is shown in Figure 6. 7, 9, 18, 29, 37, 43 are bearings, 16 are retaining rings, 30 are copper sleeves, 46 are planet carriers, 34 are screws, 5 are spline sleeves, 45 are gaskets, and 35 are housings.

According to the above-mentioned reliable network segmentation modeling method, a reliability network diagram including all parts of the slat 3K planetary reducer is established as shown in Figure 7, and the names and quantities of the parts represented by the specific serial numbers are shown in Table 2.

Step 4: Take the failure rate Y=0.0600 of the slat 3K planetary reducer as the evaluation index, set the total number of simulation times Num to 105 times, initialize the number of simulation times N=1, and the cumulative safety times of the system n=0.

Step 5: According to the common cause failure Monte Carlo simulation strategy, the common cause of failure is uniformly sampled and then transferred to each component, and the non-failure common cause parameters are independently sampled according to the parameter distribution.

The formulation of the Monte Carlo simulation strategy for common cause failure utilizes the full probability model of the reliability of the common cause failure system;

For a series system with common cause failure, the reliability full probability model can be expressed as:

For a parallel system with common cause failure, the reliability full probability model can be expressed as:

The full reliability full probability model is obtained. When dealing with the common cause failure problem, the stress that produces the common cause of failure is only integrated once, R represents the system reliability, f(σ) represents the generalized stress distribution density function, and f(S) represents the generalized strength Distribution density function, p represents the number of parts in series, q represents the number of parts in parallel. The Monte Carlo simulation strategy of common cause failure utilizes the characteristics of the full reliability full probability model in dealing with the common cause failure problem. During the simulation, the stress that produces the common cause of failure is only sampled once.

In this embodiment, according to the Monte Carlo simulation strategy for common cause failure, the input load T _a is taken as the common cause of failure, and Monte Carlo random sampling is performed on its distribution. The failure evaluation method in the failure analysis performs random sampling of stress and strength in the program, and the non-failure common cause parameters are distributed and independently sampled in the program according to the failure evaluation method in the failure analysis above.

Step 6: According to the internal failure triggering and external failure triggering conditions of each component, update each layer sequentially in the order of layer Z, layer Z-1, layer Z-2, ... layer 2, layer 1 The component state T _i ; the component state is divided into two states: safety and failure.

In this embodiment, according to the internal failure triggering and external failure triggering conditions of each component, the state of each component in FIG. 7 is sequentially updated according to the fourth layer, the third layer, the second layer, and the first layer in FIG. 7 . Ti, Ti=1 means the ith component is safe, and Ti =0 means the _ith component fails.

Step 7: Determine whether the last component in the sequence indicated by the arrow in the first layer of Fig. 7 fails, that is, whether the state Ti of part 1 is 1; if the part 1 does not fail, set the system state variable _k =1, indicating that the system is safe Otherwise, k=0, indicating that the system fails.

Step 8: Update the number of simulations N=N+1 and the number of times of safety accumulated by the system n=n+k.

Step 9: Judging whether the number of simulations satisfies N<Num, Num represents the total number of simulations set, if so, return to step 5 to continue the iteration, otherwise terminate the program and count the cumulative safety times n of the system, and calculate the reliability R=n/Num, The failure rate Y=1-R, compare the failure rate with the established failure probability, determine the operating state of the 3K planetary reducer according to the results, and reasonably arrange the operation and maintenance.

The system reliability of the slat 3K planetary reducer calculated in this embodiment is R=0.9526, and the failure rate Y=0.0474. The failure rate of the obtained slat 3K planetary reducer is less than the required failure rate, indicating that the slat 3K planetary reducer is in a safe state at this stage.

Although the embodiments of the present invention have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and those of ordinary skill in the art will not depart from the principles and spirit of the present invention Variations, modifications, substitutions, and alterations to the above-described embodiments are possible within the scope of the present invention without departing from the scope of the present invention.

Claims (1)

1. A3K planetary reducer reliability assessment method based on a neural network is characterized in that: the method comprises the following steps:

step 1: carrying out failure definition on each part of the 3K planetary reducer based on a neuron structure, wherein the failure definition comprises three attributes which are an input attribute, a failure trigger attribute and an output attribute respectively; the input attributes comprise state variables k, load F, vibration x and torque T of the parts; the output attribute refers to an output state; the failure triggering attributes comprise an external failure triggering condition and an internal failure triggering condition; the external failure triggering condition represents a condition that another component connected to the component and at a level higher than the power transmission level of the system fails to receive the power of the system; the internal failure triggering conditions represent each failure mode of the part;

step 2: after failure definition is carried out on each part, an internal failure mechanism, an external failure mechanism and a failure propagation mechanism of each part are determined;

and 3, step 3: based on the primary and secondary structure of the neural network, performing reliability layered modeling on the 3K planetary reducer according to the assembly relation among parts and the importance degree in the power transmission process to obtain a reliability network diagram, wherein the reliability network diagram has a Z layer: the method comprises the following steps that 1, the layer 1 is a core component of the 3K planetary reducer, 2, the layer 2 is a secondary component directly influencing the core component, 3, the layer 3 is a component directly influencing the secondary component, and the rest is repeated to determine the layer of each component in the 3K planetary reducer; establishing a pointing sequence for the parts of each layer according to the power transmission direction;

and 4, step 4: taking the fault rate of the 3K planetary reducer as an evaluation index, initializing the simulation times N to be 1, and initializing the accumulated safety times N to be 0;

and 5: uniformly sampling failure common causes according to a common cause failure Monte Carlo simulation strategy, transmitting the failure common causes to each part, and independently sampling non-failure common cause parameters according to parameter distribution;

step 6: according to the internal failure trigger and the external failure trigger conditions of each part, the part state T of each layer is updated in sequence according to the sequence of the Z layer, the Z-1 layer, the Z-2 layer, the … 2 layer and the 1 layer _i (ii) a The state of the part is divided into a safe state and a failure state, T _i 1 denotes the safety of the ith component, T _i 0 indicates that the ith part fails;

and 7: judging whether the last part in the direction sequence in the layer 1 of the reliability network diagram fails, if not, setting a system state variable k to be 1 to represent system safety, otherwise, setting k to be 0 to represent system failure;

and 8: updating simulation times N which is N +1 and system accumulated safety times N which is N + k;

and step 9: and judging whether the simulation times meet N < Num, wherein Num represents the set total simulation times, if so, returning to the step 5 to continue iteration, otherwise, terminating the program and counting the system accumulated safety times N, calculating the reliability R equal to N/Num, and the failure rate Y equal to 1-R, comparing the failure rate with the formulated fault probability, and judging the running state of the 3K planetary reducer according to the result.

Images