A wear assessment method and system for aluminum alloy monofilament and an optimization method and system

Abstract

The present invention discloses a wear assessment method and system for aluminum alloy monofilaments and an optimization method and system. The wear assessment method comprises: obtaining training data based on a pre-constructed finite element model, and training a pre-constructed BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model; calculating multiple aluminum alloy monofilament wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model; using the Monte Carlo method to generate a set of random numbers of corresponding distribution, substituting them into the response surface function of the BP neural network model for simulation calculation to obtain the aluminum alloy monofilament wear limit state function value; performing probability analysis on all current aluminum alloy monofilament wear limit state function values ​​to obtain the wear reliability that meets the allowable error condition. The present invention shortens the calculation cycle for evaluating the wear of aluminum alloy monofilaments and is suitable for engineering applications.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid disaster prevention, and in particular to a wear assessment method and system for aluminum alloy monofilaments and an optimization method and system. Background Art

[0002] When the transmission line of a large-span project is in a continuous and stable wind zone, the aluminum alloy monofilaments in the ultra-high-strength stranded wire are seriously affected by the excitation of the wind. Since the conductor mainly dissipates wind energy by friction between the strands, under the action of continuous and stable wind, the wear between the strands of the ultra-high-strength aluminum alloy stranded wire gradually increases, thereby increasing the probability of the ultra-high-strength aluminum alloy stranded wire breaking due to strand wear. The large-span ultra-high-strength aluminum alloy stranded wire is the carrier of power transmission. Under the coupling effect of the force field generated by breeze vibration and the temperature field generated by current, the wear of the aluminum alloy monofilaments in the large-span ultra-high-strength aluminum alloy stranded wire causes cracks in the stranded wire at the wear point, which in turn causes fatigue damage to the stranded wire. The degree of fatigue damage is more serious than that of the large-span ultra-high-strength aluminum alloy stranded wire under the condition of only considering breeze vibration. At present, the evaluation methods for micro-motion fatigue wear of large-span conductors are mainly fatigue wear tests and fatigue finite element analysis. Fatigue wear test can directly evaluate the wear condition of aluminum alloy single wire, but the test cycle of this method is very long and the test cost is also very high; the wear condition of aluminum alloy single wire can also be evaluated by finite element calculation method, but the calculation cycle of this method is very long, so these two methods are not conducive to engineering applications. Summary of the invention

[0003] In order to solve the above-mentioned deficiencies in the prior art, the present invention provides a wear assessment method for aluminum alloy monofilament, comprising:

[0004] S1. Acquire training data based on a pre-constructed finite element model, and train a pre-constructed BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model;

[0005] S2. Calculating a plurality of aluminum alloy single wire wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model;

[0006] S3. Generate a set of random numbers of corresponding distribution by using Monte Carlo method, substitute them into the response surface function of BP neural network model for simulation calculation to obtain the wear limit state function value of aluminum alloy single wire; make probability analysis on all the current wear limit state function values ​​of aluminum alloy single wire, and calculate the wear reliability of aluminum alloy single wire;

[0007] S4. Determine whether the wear reliability of the aluminum alloy monofilament meets the set allowable error conditions. If the conditions are met, end the calculation and obtain the wear reliability of the current aluminum alloy monofilament; otherwise, update the number of aluminum alloy monofilament wear limit state function values ​​and execute S3.

[0008] Preferably, the acquiring of training data based on a pre-built finite element model comprises:

[0009] The total depth of sliding wear of aluminum alloy single wire was obtained based on the pre-built finite element model;

[0010] The finite element model is used multiple times to obtain the total depth of sliding wear of multiple aluminum alloy single wires;

[0011] The total depth of sliding wear of multiple aluminum alloy single wires was used as training data.

[0012] Preferably, the total depth of sliding wear of the aluminum alloy monofilament is obtained based on a pre-constructed finite element model, comprising:

[0013] S301, bringing the values ​​of the density, elastic modulus of the steel core, the density, elastic modulus, Poisson's ratio of the aluminum strands, and the friction coefficient between the strands into a pre-built finite element model;

[0014] S302, simulating the sliding of the aluminum alloy single wire in the stranded wire based on the finite element model, selecting multiple target nodes from the aluminum alloy single wire model, and extracting the curve belonging to the wear stage from the time history curve of each target node; dividing the extracted wear stage curve based on the set time interval, and extracting the stress value of each target node at the current wear number in each time interval;

[0015] S303, calculating the wear depth of a single sliding of the aluminum alloy single wire in each time interval based on the stress value of each target node at the current wear times and the measured hardness of the aluminum alloy single wire, the wear factor of the aluminum alloy single wire, the relative sliding speed, and the sliding time;

[0016] S304, calculating the wear depth of a single sliding of the aluminum alloy single wire in all time intervals based on the wear depth of a single sliding of the aluminum alloy single wire in each time interval;

[0017] S305, determining whether the number of wear times meets the set total number of wear times;

[0018] S306. When the set total number of wear times is not met, the wear depth of a single sliding of the aluminum alloy single wire in all time intervals corresponding to the current number of wear times is recorded, the number of wear times is updated, and the wear depth of a single sliding of the aluminum alloy single wire in all time intervals in the current simulation result is used as the initial condition of the aluminum alloy single wire model in the next simulation, and S302 is executed; otherwise, the iteration is terminated, and the total depth of sliding wear of the aluminum alloy single wire is obtained based on the wear depth of a single sliding of the aluminum alloy single wire in all time intervals corresponding to all the wear times.

[0019] Preferably, the construction of the finite element model includes:

[0020] To simulate the twisted wire at the half-wavelength antinode position, the calculation unit of the twisted wire strands is solid unit, the contact area between the twisted wire strands is contact unit, and the mapping method is used for meshing;

[0021] Constrain the degrees of freedom in the X, Y, and Z directions of the Z=0 face in the length direction.

[0022] Preferably, the wear depth of a single sliding of the aluminum alloy monofilament in each time interval is calculated according to the following formula:

[0023]

[0024] Where: Δh j,n,i is the wear depth of the aluminum alloy wire in a single sliding at the nth target node under the jth wear in the ith time interval; H is the hardness of the aluminum alloy wire; K is the wear factor of the aluminum alloy wire; v is the relative sliding velocity; Δt i is the i-th time interval; σ j,n,i is the stress value of the nth target node under the jth wear in the ith time interval.

[0025] Preferably, the wear depth of a single sliding of the aluminum alloy monofilament in all time intervals is calculated as follows:

[0026]

[0027] Where: h j is the wear depth of a single aluminum alloy wire sliding under the jth wear in all time intervals; Δh j,n,i is the wear depth of the aluminum alloy single wire at the nth target node under the jth wear in the i-th time interval; N is the number of target nodes; m is the number of time intervals.

[0028] Preferably, the total depth of sliding wear of the aluminum alloy monofilament is calculated as follows:

[0029]

[0030] Where: h is the total depth of sliding wear of aluminum alloy single wire; h j is the wear depth of a single aluminum alloy wire sliding under the jth wear in all time intervals; k is the total number of wear times.

[0031] Preferably, the construction of the BP neural network model includes:

[0032] The total area of ​​aluminum alloy monofilaments, the total area of ​​stranded steel wires, the number of wear times, displacement, tension, and temperature are used as the input layer, and the number of neurons in the input layer is set to 6;

[0033] Set a hidden layer and set the number of units in the hidden layer to 13;

[0034] The aluminum alloy single wire wear state function is used as the output layer, and the number of neurons in the output layer is set to 1;

[0035] Wherein, the BP neural network model includes the input layer, the hidden layer and the output layer.

[0036] Preferably, the response surface function mapping relationship of the BP neural network model is as shown in the following formula:

[0037]

[0038] Where: δ is the logarithmic sigmoid function; P is the number of units in the input layer; Q is the number of units in the hidden layer; U is the number of units in the output layer; w pq is the connection weight between the input layer and the hidden layer; v uq is the connection weight between the hidden layer and the output layer; b 2u is the threshold of the output layer; b 1q Hidden layer threshold; x p is the input layer vector.

[0039] Preferably, the wear limit state function value of the aluminum alloy single wire is calculated according to the following formula:

[0040] Z=g(X)=g(X1,X2,X3,X4,X5,X6)=h R -h(X1,X2,X3,X4,X5,X6)

[0041] Where: Z is the wear limit state function value of aluminum alloy single wire; g(X) is the wear state function of aluminum alloy single wire of stranded wire; X is the design variable; h R is the wear depth limit of aluminum alloy single wire in stranded wire; h(X1,X2,X3,X4,X5,X6) is the wear depth of aluminum alloy single wire in stranded wire, which is a random variable; X1 is the total area of ​​aluminum alloy single wire; X2 is the total area of ​​steel wire in stranded wire; X3 is the number of wear times; X4 is displacement; X5 is tension; X6 is temperature.

[0042] Based on the same inventive concept, the present invention also provides a wear assessment system for aluminum alloy monofilament, comprising:

[0043] A training module, used to obtain training data based on a pre-built finite element model, and train a pre-built BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model;

[0044] A first calculation module is used to calculate a plurality of aluminum alloy single wire wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model;

[0045] The second calculation module is used to generate a set of random numbers of corresponding distribution using the Monte Carlo method, substitute them into the response surface function of the BP neural network model for simulation calculation to obtain the wear limit state function value of the aluminum alloy single wire; perform probability analysis on all the current wear limit state function values ​​of the aluminum alloy single wire to obtain the wear reliability of the aluminum alloy single wire;

[0046] The judgment module is used to judge whether the wear reliability of the aluminum alloy monofilament meets the set allowable error conditions. When the conditions are met, the calculation is terminated to obtain the wear reliability of the current aluminum alloy monofilament; otherwise, the number of aluminum alloy monofilament wear limit state function values ​​is updated and the second calculation module is executed.

[0047] Preferably, the training module includes:

[0048] An acquisition unit, used for obtaining the total depth of sliding wear of the aluminum alloy single wire based on a pre-built finite element model;

[0049] A circulation unit, used for obtaining the total depth of sliding wear of multiple aluminum alloy single wires by using the finite element model multiple times;

[0050] The result unit is used to take the total depth of sliding wear of multiple aluminum alloy single wires as training data.

[0051] Based on the same inventive concept, the present invention also provides an optimization method based on the wear assessment method, comprising:

[0052] The wear state function of the aluminum alloy single wire of the stranded wire is calculated by using the response surface function mapping relationship of the BP neural network model.

[0053] The wear reliability of the aluminum alloy single wire of the stranded wire under wind vibration conditions is calculated using the Monte Carlo method based on the wear state function of the aluminum alloy single wire of the stranded wire;

[0054] Based on the pre-built optimization model, the design size of the aluminum alloy single wire of the stranded wire is continuously adjusted to make the wear reliability of the aluminum alloy single wire of the stranded wire close to the allowable wear reliability of the stranded wire, and on this basis, the cross-sectional area of ​​the aluminum wire and the cross-sectional area of ​​the steel wire of the stranded wire are given;

[0055] The optimized weight of the stranded wire is obtained by comparing the cross-sectional area of ​​the aluminum wire and the cross-sectional area of ​​the steel wire of the stranded wire.

[0056] Preferably, the construction of the optimization model includes:

[0057] The evaluation conditions are constructed based on the relationship between the reliability of the strands under the current stranding parameters and the set reliability;

[0058] The objective optimization function is constructed based on the minimum weight of the strand.

[0059] Preferably, the optimization model is shown as follows:

[0060]

[0061] Where: W(X) is the weight of the stranded wire; X1 is the total area of ​​the aluminum alloy monofilament; X2 is the total area of ​​the steel wire in the stranded wire; ρ1 is the density of aluminum; ρ2 is the density of steel; P{g(X)} is the reliability of the aluminum alloy monofilament in the stranded wire; R s Allows wear reliability for stranded wire.

[0062] Based on the same inventive concept, the present invention also provides an optimization system based on the wear assessment method, comprising:

[0063] The wear state function calculation module is used to calculate the wear state function of the aluminum alloy single wire of the stranded wire by using the response surface function mapping relationship of the BP neural network model;

[0064] A wear reliability calculation module is used to calculate the wear reliability of the aluminum alloy single wire of the stranded wire under wind vibration conditions using a Monte Carlo method based on the wear state function of the aluminum alloy single wire of the stranded wire;

[0065] The adjustment module is used to continuously adjust the design size of the aluminum alloy single wire of the stranded wire based on the pre-built optimization model, so that the wear reliability of the aluminum alloy single wire of the stranded wire is close to the allowable wear reliability of the stranded wire, and on this basis, the cross-sectional area of ​​the aluminum wire and the cross-sectional area of ​​the steel wire of the stranded wire are given;

[0066] The optimization module is used to calculate the cross-sectional area of ​​the aluminum wire of the optimized stranded wire and the cross-sectional area of ​​the steel wire to obtain the optimized weight of the stranded wire.

[0067] Preferably, the optimized stranding system further comprises a construction module for constructing an optimization model;

[0068] The building blocks include:

[0069] An evaluation condition unit, used for constructing an evaluation condition based on the relationship between the reliability of the stranded wire under the current stranded wire parameters and the set reliability;

[0070] The target optimization function unit is used to construct the target optimization function based on the minimum weight of the strand.

[0071] Compared with the prior art, the present invention has the following beneficial effects:

[0072] 1) The technical solution provided by the present invention includes: obtaining training data based on a pre-constructed finite element model, and training a pre-constructed BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model; calculating multiple aluminum alloy monofilament wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model; using the Monte Carlo method to generate a set of random numbers of corresponding distribution, substituting them into the response surface function of the BP neural network model for simulation calculation to obtain the aluminum alloy monofilament wear limit state function value; performing probability analysis on all current aluminum alloy monofilament wear limit state function values ​​to calculate the wear reliability of the aluminum alloy monofilament; judging whether the wear reliability of the aluminum alloy monofilament meets the set allowable error conditions, and when the conditions are met, terminating the calculation to obtain the wear reliability of the current aluminum alloy monofilament; otherwise updating the number of aluminum alloy monofilament wear limit state function values ​​and continuing the calculation. This method shortens the calculation cycle for evaluating the wear condition of the aluminum alloy monofilament and is suitable for engineering applications.

[0073] 2) The technical solution provided by the present invention provides a method for calculating the wear of aluminum alloy monofilaments in ultra-high strength stranded wires. This method can calculate the wear depth of aluminum alloy monofilaments under different tension, temperature and amplitude conditions, thereby providing sufficient sample data for establishing a mathematical model for optimizing the wear of aluminum alloy monofilaments based on reliability.

[0074] 3) The technical solution provided by the present invention establishes a calculation method for the wear reliability of aluminum alloy single wire based on the BP neural network response surface. This method can quickly predict the wear reliability of ultra-high strength aluminum alloy stranded wire under different tension, temperature and amplitude conditions, thereby saving a lot of calculation time, improving calculation efficiency, saving test and calculation costs, and facilitating engineering promotion and application.

[0075] 4) The technical solution provided by the present invention establishes an optimization model for aluminum alloy single wire wear based on reliability. After applying the optimization model to optimize the design of ultra-high strength aluminum alloy stranded wire, its wear can meet the allowable reliability control requirements and effectively reduce the weight of the connector terminal.

[0076] 5) The technical solution provided by the present invention establishes a wear assessment model for aluminum alloy monofilaments in stranded wires and evaluates the wear of the stranded wires under continuous and stable wind conditions, which is very necessary to avoid strand breakage due to strand wear and improve the reliability of power grid operation safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A flow chart of a wear assessment method for aluminum alloy monofilaments in a stranded wire provided by the present invention;

[0078] Figure 2 A flow chart for calculating the wear depth of aluminum alloy monofilaments in the large-span ultra-high-strength aluminum alloy stranded wire provided by the present invention;

[0079] Figure 3 A neural network model for calculating the response surface function of an aluminum alloy single wire according to the present invention;

[0080] Figure 4 A flow chart for calculating the wear reliability of aluminum alloy single wires in the large-span ultra-high-strength aluminum alloy stranded wire provided by the present invention. DETAILED DESCRIPTION

[0081] In order to better understand the present invention, the content of the present invention is further described below in conjunction with the accompanying drawings and examples.

[0082] like Figure 1 As shown, the present invention provides a wear assessment method for an aluminum alloy monofilament, comprising:

[0083] S1. Acquire training data based on a pre-constructed finite element model, and train a pre-constructed BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model;

[0084] S2. Calculating a plurality of aluminum alloy single wire wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model;

[0085] S3. Generate a set of random numbers of corresponding distribution by using Monte Carlo method, substitute them into the response surface function of BP neural network model for simulation calculation to obtain the wear limit state function value of aluminum alloy single wire; make probability analysis on all the current wear limit state function values ​​of aluminum alloy single wire, and calculate the wear reliability of aluminum alloy single wire;

[0086] S4. Determine whether the wear reliability of the aluminum alloy monofilament meets the set allowable error conditions. If the conditions are met, end the calculation and obtain the wear reliability of the current aluminum alloy monofilament; otherwise, update the number of aluminum alloy monofilament wear limit state function values ​​and execute S3.

[0087] S1. Acquiring training data based on a pre-constructed finite element model, and training a pre-constructed BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model, including:

[0088] 1. Sample data acquisition of wear depth of aluminum alloy single wire of stranded wire

[0089] The present invention adopts a discretization method to calculate the wear depth of the aluminum alloy single wire of the ultra-high strength stranded wire, including:

[0090] A certain time period Δt is selected as the time interval. The contact stress σ of the aluminum alloy single wire in this time interval Δt can be regarded as a constant value. The wear depth Δh in this time interval is calculated. j,n,i(where j represents the jth wear and n represents the nth target node), and then the wear depth is obtained by summing.

[0091] For the time interval Δt i The wear depth of a single sliding of an inner aluminum alloy single wire can be calculated by formula (1):

[0092]

[0093] Where: Δh j,n,i is the wear depth of the aluminum alloy wire in a single sliding at the nth target node under the jth wear in the ith time interval; H is the hardness of the aluminum alloy wire; K is the wear factor of the aluminum alloy wire; v is the relative sliding velocity; Δt i is the i-th time interval; σ j,n,i is the stress value of the nth target node under the jth wear in the ith time interval.

[0094] The wear depth of aluminum alloy wire in a single sliding operation in a time interval of m is h. j It can be expressed as:

[0095]

[0096] Where: h j is the wear depth of a single aluminum alloy wire sliding under the jth wear in all time intervals; Δh j,n,i is the wear depth of the aluminum alloy single wire at the nth target node under the jth wear in the i-th time interval; N is the number of target nodes; m is the number of time intervals.

[0097] The total depth of sliding wear of aluminum alloy single wire under all wear times can be expressed as:

[0098]

[0099] Where: h is the total depth of sliding wear of aluminum alloy single wire; h j is the wear depth of a single aluminum alloy wire sliding under the jth wear in all time intervals; k is the total number of wear times.

[0100] The specific process is divided into two steps: first, extract the curve belonging to the wear stage from the time history curve of the target node; second, divide the extracted wear stage curve and solve it using formula (1).

[0101] By using the finite element calculation method, the stress value on the target node of the aluminum alloy monofilament can be obtained more accurately. In order to calculate the stress value on the target node of the aluminum alloy monofilament, the present invention establishes a finite element model, including: taking a section of ultra-high-strength aluminum alloy stranded wire at the antinode position of half wavelength (λ / 2), establishing an ultra-high-strength aluminum alloy stranded wire finite element model, the stranded wire strand calculation unit of the model adopts solid unit, the contact area between the stranded wire strands adopts contact unit, and the mapping method is used for meshing.

[0102] The finite element model parameters of ultra-high strength aluminum alloy stranded wire are set as follows: the density of the steel core is 7780kg / m 3 , elastic modulus is 200GPa, and the density of aluminum wire strand is 2703kg / m 3 , elastic modulus is 62GPa, Poisson's ratio is 0.3, and the friction coefficient between the strands is 0.5. In the setting of boundary conditions, the degrees of freedom of the Z=0 plane in the X, Y, and Z directions are constrained in the length direction. Considering that the stranded wire is a symmetrical axial tension structure, a rigid domain point is established outside the Z=L plane, and then coupled with the surface into a rigid domain, forcing all nodes on the Z=L plane to have the same displacement in the axial direction of the wire; then tension F and temperature load T are applied to the rigid domain nodes of the model, and displacement load A is applied to it in the middle of the stranded wire. The stress values ​​on the target nodes of the aluminum alloy monofilament are calculated under different temperatures T, different tensions F, and different amplitudes.

[0103] In summary, if Figure 2 The finite element simulation method used to calculate the wear depth of aluminum alloy monofilament is mainly divided into two steps:

[0104] (1) Simulate the sliding of the aluminum alloy single wire, select the target node n from the simulated aluminum alloy single wire model, and extract the jth stress value σ required to calculate the wear depth j,n,i , use formula (1) to calculate the wear depth under each load step, and then combine formula (2) to calculate the wear depth of a single aluminum alloy wire sliding.

[0105] (2) The wear depth of the aluminum alloy single wire of the stranded wire in the previous simulation result is used as the initial condition of the model for the next simulation, that is, the previous wear value is substituted into the newly established model, and then the initial condition of the modeling is imported for simulation. This process is repeated and the total wear depth is obtained by formula (3).

[0106] 2 Establishment of reliability-based aluminum alloy single wire wear assessment model

[0107] In order to quickly evaluate the wear state of the aluminum alloy monofilament, the present invention establishes the following wear evaluation model for the aluminum alloy monofilament:

[0108]

[0109] In formula 4, X1 is the total area of ​​aluminum alloy single wires in the stranded wire, X2 is the total area of ​​steel wires in the stranded wire, X3 is the number of wear times, X4 is the displacement, X5 is the tension, X6 is the temperature, and f(X) is the calculation expression for the wear depth of aluminum alloy single wires in the stranded wire, which can be expressed as:

[0110] f(X)=h(X1,X2,X3,X4,X5,X6) (5)

[0111] In formula 4, R is the wear reliability of the stranded wire, R s is the allowable wear reliability of the stranded wire, R≥R s is the evaluation condition of strand wear, which can be further expressed as:

[0112]

[0113] g(X) is the wear state function, which can be expressed as:

[0114] Z=g(X)=g(X1,X2,X3,X4,X5,X6)=h R -h(X1,X2,X3,X4,X5,X6) (7)

[0115] when

[0116]

[0117] According to formula 8, when the wear depth of the aluminum alloy monofilament is close to the wear depth limit, it is in failure state. According to formula 7, the wear limit state equation of the aluminum alloy monofilament can be expressed as:

[0118] Z=g(X)=g(X1,X2,X3,X4,X5,X6)=h R -h(X1,X2,X3,X4,X5,X6)=0 (9)

[0119] Where: h R is the wear depth limit of aluminum alloy monofilament; h(X1,X2,X3,X4,X5,X6) is the wear depth of aluminum alloy monofilament, which is a random variable.

[0120] The key to solving the reliability of aluminum alloy single wire wear using formula 9 is to construct the mapping relationship between the aluminum alloy single wire wear state function g(X) and the design variable X. To construct this mapping relationship, it is necessary to know the joint probability density function of the state function Z of the aluminum alloy single wire wear. This function has an implicit function relationship with the influencing factors of the terminal temperature, which is difficult to calculate using traditional methods. To this end, the present invention uses the BP neural network response surface method to construct the relationship between the aluminum alloy single wire wear and its influencing factors.

[0121] This method first uses the sample data obtained by finite element calculation, and then uses the nonlinear mapping function of the neural network to simulate the explicit function expression (such as Figure 3 BP neural network is a multi-layer feedforward network, which includes input layer, hidden layer and output layer, and the units in the same layer are not connected. BP network can be regarded as a highly nonlinear mapping from input to output, that is: F: Rn→Rm, f(X)=Y. For the sample set: input x i (∈R n ) and output y i (∈R m ), it is believed that there is a mapping g, so that g(xi)=yi(i=1,2,…,N). When establishing the BP neural network response surface model of aluminum alloy single wire wear, the random input variables wear times X1, displacement X3, tension X4, temperature X5 are first used as 6 nodes of the network input layer, the aluminum alloy single wire wear function value Z is used as a node of the network output layer, and the number of nodes in the hidden layer is set to 13 (as shown in the attached figure). Figure 3 As shown). The transfer functions in the hidden layer and the output layer are all logarithmic sigmoid functions, so the mapping relationship between Z and random variables X = [X1, X2, X3, X4, X5, X6] can be expressed as:

[0122]

[0123] Where: δ is the logarithmic sigmoid function; P is the number of units in the input layer; Q is the number of units in the hidden layer; U is the number of units in the output layer; w pq is the connection weight between the input layer and the hidden layer; v uq is the connection weight between the hidden layer and the output layer; b 2u is the threshold of the output layer; b 1q Hidden layer threshold; x p is the input layer vector, that is, the vector composed of X1, X2, X3, X4, X5, and X6.

[0124] When training the BP neural network response surface model of aluminum alloy monofilament, the design variables X(X1, X2, X3, X4, X5, X6) are first input into the parametric finite element model of aluminum alloy monofilament to calculate the wear depth h(X1, X2, X3, X4, X5, X6) of the aluminum alloy monofilament, and then the functional function value Z is obtained. i . X i (X 1i ,X 2i ,X 3i ,X 4i ,X 5i ,X6i ) and Z i The sample points were input into the aluminum alloy single wire BP neural network response surface model to train the network. The learning rate, momentum factor and training error were selected as 0.6, 0.9 and 0.001 respectively.

[0125] S2. Calculating a plurality of aluminum alloy single wire wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model;

[0126] S3. Generate a set of random numbers of corresponding distribution by using Monte Carlo method, substitute them into the response surface function of BP neural network model for simulation calculation to obtain the wear limit state function value of aluminum alloy single wire; make probability analysis on all the current wear limit state function values ​​of aluminum alloy single wire, and calculate the wear reliability of aluminum alloy single wire;

[0127] S4. Determine whether the wear reliability of the aluminum alloy monofilament meets the set allowable error conditions. If the conditions are met, end the calculation and obtain the wear reliability of the current aluminum alloy monofilament; otherwise, update the number of aluminum alloy monofilament wear limit state function values ​​and execute S3.

[0128] Figure 4 The calculation process of the wear reliability of aluminum alloy single wire is given. The main calculation process is:

[0129] a) Calculate a series of aluminum alloy single wire limit state functions Z by using the response surface function mapping relationship of BP neural network i value, and find its mean μ zi and variance σ zi As the data for the next step of analysis;

[0130] b) Generate the i-th group of random numbers X of corresponding distribution using Monte Carlo method i1 , X i2 , X i3 , X i4 , X i5 , X 6i Substitute the response surface function of the BP neural network for simulation calculation to obtain the corresponding aluminum alloy single wire wear limit state function value Z i , and then make a probability analysis on the i limit state function values, and then find the wear reliability R of the aluminum alloy single wire;

[0131] c) Determine whether the conditions are met, RR s ≤ε, where R s is the allowable wear reliability of the stranded wire, ε is the allowable error, and the calculation ends if the condition is met, otherwise let n=n+1 and go to (b).

[0132] Based on the same inventive concept, the present invention provides a wear assessment system for aluminum alloy monofilament, comprising:

[0133] A training module, used to obtain training data based on a pre-built finite element model, and train a pre-built BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model;

[0134] A first calculation module is used to calculate a plurality of aluminum alloy single wire wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model;

[0135] The second calculation module is used to generate a set of random numbers of corresponding distribution using the Monte Carlo method, substitute them into the response surface function of the BP neural network model for simulation calculation to obtain the wear limit state function value of the aluminum alloy single wire; perform probability analysis on all the current wear limit state function values ​​of the aluminum alloy single wire to obtain the wear reliability of the aluminum alloy single wire;

[0136] The judgment module is used to judge whether the wear reliability of the aluminum alloy monofilament meets the set allowable error conditions. When the conditions are met, the calculation is terminated to obtain the wear reliability of the current aluminum alloy monofilament; otherwise, the number of aluminum alloy monofilament wear limit state function values ​​is updated and the second calculation module is executed.

[0137] In this embodiment, the training module includes:

[0138] An acquisition unit, used for obtaining the total depth of sliding wear of the aluminum alloy single wire based on a pre-built finite element model;

[0139] A circulation unit, used for obtaining the total depth of sliding wear of multiple aluminum alloy single wires by using the finite element model multiple times;

[0140] The result unit is used to take the total depth of sliding wear of multiple aluminum alloy single wires as training data.

[0141] Based on the same inventive concept, the present invention also provides an optimization method based on the wear assessment method. The specifically provided reliability-based strand wear assessment model can be used to optimize the design of ultra-high-strength aluminum alloy strands.

[0142] The weight function expression of ultra-high strength aluminum alloy stranded wire per unit length is:

[0143] W(X)=X1·ρ1+X2·ρ2 (11)

[0144] Where: X1 is the total area of ​​aluminum alloy single wires in the stranded wire, X2 is the total area of ​​steel wires in the stranded wire, ρ1 is the density of aluminum, and ρ2 is the density of steel.

[0145] The optimization model of the ultra-high strength aluminum alloy stranded wire provided by the present invention is:

[0146]

[0147] Where P{g(X)≥0}≥R s is the evaluation condition of strand wear, R s is the allowable wear reliability of the stranded wire, and g(x) is the state function.

[0148] The optimization design of ultra-high strength aluminum alloy stranded wire using the stranded wire optimization model can be divided into the following steps:

[0149] (a) Using Equation 10, calculate the wear state function g(X) of the stranded wire;

[0150] (b) Based on the wear assessment model of aluminum alloy monofilament, the wear reliability R of the stranded wire under wind vibration conditions is calculated using the Monte Carlo method;

[0151] (c) Using the optimization model given by Equation 12, the design dimensions of the strands are continuously adjusted to make the wear calculation reliability R of the strands close to the allowable wear design reliability R of the strands. s , and on this basis, the cross-sectional area of ​​the aluminum wire X1 and the cross-sectional area of ​​the steel wire X2 of the stranded wire are given

[0152] (d) Substitute the cross-sectional area X1 of the aluminum wire and the cross-sectional area X2 of the optimized stranded wire into formula (11) to obtain the optimized weight of the stranded wire.

[0153] It can be seen that the stranded wire wear assessment model based on reliability provided by the present invention can be used to optimize the design of the stranded wire, so that the weight of the stranded wire is minimized while ensuring the reliability of the stranded wire wear, thereby achieving good economic benefits.

[0154] The present invention provides an optimization system based on the same inventive concept and the wear assessment method, including:

[0155] The wear state function calculation module is used to calculate the wear state function of the aluminum alloy single wire of the stranded wire by using the response surface function mapping relationship of the BP neural network model;

[0156] A wear reliability calculation module is used to calculate the wear reliability of the aluminum alloy single wire of the stranded wire under wind vibration conditions using a Monte Carlo method based on the wear state function of the aluminum alloy single wire of the stranded wire;

[0157] The adjustment module is used to continuously adjust the design size of the aluminum alloy single wire of the stranded wire based on the pre-built optimization model, so that the wear reliability of the aluminum alloy single wire of the stranded wire is close to the allowable wear reliability of the stranded wire, and on this basis, the cross-sectional area of ​​the aluminum wire and the cross-sectional area of ​​the steel wire of the stranded wire are given;

[0158] The optimization module is used to calculate the cross-sectional area of ​​the aluminum wire of the optimized stranded wire and the cross-sectional area of ​​the steel wire to obtain the optimized weight of the stranded wire.

[0159] In an embodiment, the optimized stranding system further comprises a construction module for constructing an optimization model;

[0160] The building blocks include:

[0161] An evaluation condition unit, used for constructing an evaluation condition based on the relationship between the reliability of the stranded wire under the current stranded wire parameters and the set reliability;

[0162] The target optimization function unit is used to construct the target optimization function based on the minimum weight of the strand.

[0163] like Figure 2 As shown in the figure, the calculation process of the wear depth of aluminum alloy single wire in large span ultra-high strength aluminum alloy stranded wire includes:

[0164] Firstly, the wear finite element model of aluminum alloy monofilament is established, and the hardness, wear coefficient, elastic modulus and other parameters of the stranded wire are input into the finite element model to calculate the stress value of the contact surface node of the aluminum alloy monofilament.

[0165] Then, the wear depth under each load step is calculated by using Formula 1, and the wear depth of a single aluminum alloy wire sliding is calculated by combining Formula 1.

[0166] The wear depth of the aluminum alloy single wire in the previous simulation result is used as the initial condition of the model for the next simulation, that is, the previous wear value is substituted into the newly established model, and then the initial modeling conditions are imported for simulation, and this process is repeated to obtain the total wear depth through the formula.

[0167] like Figure 3 As shown, the neural network model used in the present invention to calculate the response surface function of the aluminum alloy single wire includes:

[0168] The model takes the total area X1 of the stranded aluminum alloy single wire, the total area X2 of the steel wire in the stranded wire, the number of wear times X3, the displacement X4, the tension X5, and the temperature X6 as the network input layer, and the number of input neurons is 6; the response surface function Z of the aluminum alloy single wire is taken as the output of the network, that is, the number of output neurons is 1; a hidden layer is selected, the number of nodes in the hidden layer is 13, and its topological structure is 6-13-1.

[0169] like Figure 4 As shown, the calculation method of aluminum alloy single wire wear reliability based on BP neural network response surface includes:

[0170] This method uses the response surface function mapping relationship of the BP neural network to calculate the limit state function value of the aluminum alloy single wire in a series of ultra-high strength aluminum alloy strands, and calculates its mean and variance as the data for the next step of analysis;

[0171] Then, the probability constraints corresponding to each state function in the optimization design model are converted into deterministic constraints, and the optimization design is performed to obtain the initial optimization design point;

[0172] Finally, the Monte Carlo method and the response surface function of the BP neural network were used to perform simulation calculations to obtain the corresponding limit state function and then calculate the reliability of the aluminum alloy single wire.

[0173] The stranded wire mentioned in the present invention is a power transmission wire composed of multiple strands of aluminum alloy monofilaments and steel wires.

[0174] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0175] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0176] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0177] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1The steps for the functions specified in one or more boxes.

[0178] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are included in the scope of the claims of the present invention to be approved.

Claims

1. A wear assessment method for aluminum alloy monofilament, characterized in that: include: S1. Acquire training data based on a pre-constructed finite element model, and train a pre-constructed BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model; S2. Calculating a plurality of aluminum alloy single wire wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model; S3. Generate a set of random numbers of corresponding distribution by using Monte Carlo method, substitute them into the response surface function of BP neural network model for simulation calculation to obtain the wear limit state function value of aluminum alloy single wire; make probability analysis on all the current wear limit state function values ​​of aluminum alloy single wire, and calculate the wear reliability of aluminum alloy single wire; S4, judging whether the wear reliability of the aluminum alloy monofilament meets the set allowable error condition, and when the condition is met, ending the calculation to obtain the wear reliability of the current aluminum alloy monofilament; Otherwise, update the number of aluminum alloy single wire wear limit state function values ​​and execute S3; The acquiring of training data based on the pre-built finite element model comprises: The total depth of sliding wear of aluminum alloy single wire was obtained based on the pre-built finite element model; The finite element model is used multiple times to obtain the total depth of sliding wear of multiple aluminum alloy single wires; The total depth of sliding wear of multiple aluminum alloy single wires is used as training data; The total depth of sliding wear of the aluminum alloy single wire is obtained based on the pre-built finite element model, including: S301, bringing the values ​​of the density, elastic modulus of the steel core, the density, elastic modulus, Poisson's ratio of the aluminum strands, and the friction coefficient between the strands into a pre-built finite element model; S302, simulating the sliding of the aluminum alloy single wire in the stranded wire based on the finite element model, selecting multiple target nodes from the aluminum alloy single wire model, and extracting the curve belonging to the wear stage from the time history curve of each target node; dividing the extracted wear stage curve based on the set time interval, and extracting the stress value of each target node at the current wear number in each time interval; S303, calculating the wear depth of a single sliding of the aluminum alloy single wire in each time interval based on the stress value of each target node at the current wear times and the measured hardness of the aluminum alloy single wire, the wear factor of the aluminum alloy single wire, the relative sliding speed, and the sliding time; S304, calculating the wear depth of a single sliding of the aluminum alloy single wire in all time intervals based on the wear depth of a single sliding of the aluminum alloy single wire in each time interval; S305, determining whether the number of wear times meets the set total number of wear times; S306. When the set total number of wear times is not met, the wear depth of a single sliding of the aluminum alloy single wire in all time intervals corresponding to the current number of wear times is recorded, the number of wear times is updated, and the wear depth of a single sliding of the aluminum alloy single wire in all time intervals in the current simulation result is used as the initial condition of the aluminum alloy single wire model in the next simulation, and S302 is executed; otherwise, the iteration is terminated, and the total depth of sliding wear of the aluminum alloy single wire is obtained based on the wear depth of a single sliding of the aluminum alloy single wire in all time intervals corresponding to all the wear times.

2. The method according to claim 1, characterized in that The construction of the finite element model includes: To simulate the twisted wire at the half-wavelength antinode position, the calculation unit of the twisted wire strands is solid unit, the contact area between the twisted wire strands is contact unit, and the mapping method is used for meshing; Constrain the degrees of freedom in the X, Y, and Z directions of the Z=0 face in the length direction.

3. The method according to claim 1, characterized in that The wear depth of a single sliding of the aluminum alloy monofilament in each time interval is calculated as follows: Where: Δh j,n,i is the wear depth of the aluminum alloy wire in a single sliding at the nth target node under the jth wear in the ith time interval; H is the hardness of the aluminum alloy wire; K is the wear factor of the aluminum alloy wire; v is the relative sliding velocity; Δt i is the i-th time interval; σ j,n,i is the stress value of the nth target node under the jth wear in the ith time interval.

4. The method according to claim 1, characterized in that The wear depth of a single aluminum alloy wire sliding in all the time intervals is calculated as follows: Where: h j is the wear depth of a single aluminum alloy wire sliding under the jth wear in all time intervals; Δh j,n,i is the wear depth of the aluminum alloy single wire at the nth target node under the jth wear in the i-th time interval; N is the number of target nodes; m is the number of time intervals.

5. The method according to claim 1, characterized in that The total depth of sliding wear of the aluminum alloy monofilament is calculated as follows: Where: h is the total depth of sliding wear of aluminum alloy single wire; h j is the wear depth of a single aluminum alloy wire sliding under the jth wear in all time intervals; k is the total number of wear times.

6. The method according to claim 1, characterized in that The construction of the BP neural network model includes: The total area of ​​aluminum alloy monofilaments, the total area of ​​stranded steel wires, the number of wear times, displacement, tension, and temperature are used as the input layer, and the number of neurons in the input layer is set to 6; Set a hidden layer and set the number of units in the hidden layer to 13; The aluminum alloy single wire wear state function is used as the output layer, and the number of neurons in the output layer is set to 1; Wherein, the BP neural network model includes the input layer, the hidden layer and the output layer.

7. The method according to claim 1, characterized in that The response surface function mapping relationship of the BP neural network model is shown in the following formula: Where: δ is the logarithmic sigmoid function; P is the number of units in the input layer; Q is the number of units in the hidden layer; U is the number of units in the output layer; w pq is the connection weight between the input layer and the hidden layer; v uq is the connection weight between the hidden layer and the output layer; b 2u is the threshold of the output layer; b 1q Hidden layer threshold; x p is the input layer vector.

8. The method according to claim 1, characterized in that The wear limit state function value of the aluminum alloy single wire is calculated as follows: Z=g(X)=g(X1,X2,X3,X4,X5,X6)=h R -h(X1,X2,X3,X4,X5,X6) Where: Z is the wear limit state function value of aluminum alloy single wire; g(X) is the wear state function of aluminum alloy single wire of stranded wire; X is the design variable; h R is the wear depth limit of aluminum alloy single wire in stranded wire; h(X1,X2,X3,X4,X5,X6) is the wear depth of aluminum alloy single wire in stranded wire, which is a random variable; X1 is the total area of ​​aluminum alloy single wire; X2 is the total area of ​​steel wire in stranded wire; X3 is the number of wear times; X4 is displacement; X5 is tension; X6 is temperature.

9. A wear assessment system for aluminum alloy monofilament, characterized in that: include: A training module, used to obtain training data based on a pre-built finite element model, and train a pre-built BP neural network model based on the training data to obtain a response surface function mapping relationship of the BP neural network model; A first calculation module is used to calculate a plurality of aluminum alloy single wire wear limit state function values ​​based on the response surface function mapping relationship of the BP neural network model; The second calculation module is used to generate a set of random numbers of corresponding distribution using the Monte Carlo method, substitute them into the response surface function of the BP neural network model for simulation calculation to obtain the wear limit state function value of the aluminum alloy single wire; perform probability analysis on all the current wear limit state function values ​​of the aluminum alloy single wire to obtain the wear reliability of the aluminum alloy single wire; A judgment module, used to judge whether the wear reliability of the aluminum alloy monofilament meets the set allowable error condition, and when the condition is met, end the calculation to obtain the wear reliability of the current aluminum alloy monofilament; Otherwise, the number of aluminum alloy single wire wear limit state function values ​​is updated and the second calculation module is executed; The training module comprises: An acquisition unit, used for obtaining the total depth of sliding wear of the aluminum alloy single wire based on a pre-built finite element model; A circulation unit, used for obtaining the total depth of sliding wear of multiple aluminum alloy single wires by using the finite element model multiple times; A result unit, used for taking the total depth of sliding wear of a plurality of aluminum alloy single wires as training data; The acquisition unit is specifically used for: S301, bringing the values ​​of the density, elastic modulus of the steel core, the density, elastic modulus, Poisson's ratio of the aluminum strands, and the friction coefficient between the strands into a pre-built finite element model; S302, simulating the sliding of the aluminum alloy single wire in the stranded wire based on the finite element model, selecting multiple target nodes from the aluminum alloy single wire model, and extracting the curve belonging to the wear stage from the time history curve of each target node; dividing the extracted wear stage curve based on the set time interval, and extracting the stress value of each target node at the current wear number in each time interval; S303, calculating the wear depth of a single sliding of the aluminum alloy single wire in each time interval based on the stress value of each target node at the current wear times and the measured hardness of the aluminum alloy single wire, the wear factor of the aluminum alloy single wire, the relative sliding speed, and the sliding time; S304, calculating the wear depth of a single sliding of the aluminum alloy single wire in all time intervals based on the wear depth of a single sliding of the aluminum alloy single wire in each time interval; S305, determining whether the number of wear times meets the set total number of wear times; S306. When the set total number of wear times is not met, the wear depth of a single sliding of the aluminum alloy single wire in all time intervals corresponding to the current number of wear times is recorded, the number of wear times is updated, and the wear depth of a single sliding of the aluminum alloy single wire in all time intervals in the current simulation result is used as the initial condition of the aluminum alloy single wire model in the next simulation, and S302 is executed; otherwise, the iteration is terminated, and the total depth of sliding wear of the aluminum alloy single wire is obtained based on the wear depth of a single sliding of the aluminum alloy single wire in all time intervals corresponding to all the wear times.

10. An optimization method for the wear assessment method according to any one of claims 1 to 8, characterized in that: include: The wear state function of the aluminum alloy single wire of the stranded wire is calculated by using the response surface function mapping relationship of the BP neural network model. The wear reliability of the aluminum alloy single wire of the stranded wire under wind vibration conditions is calculated using the Monte Carlo method based on the wear state function of the aluminum alloy single wire of the stranded wire; Based on the pre-built optimization model, the design size of the aluminum alloy single wire of the stranded wire is continuously adjusted to make the wear reliability of the aluminum alloy single wire of the stranded wire close to the allowable wear reliability of the stranded wire, and on this basis, the cross-sectional area of ​​the aluminum wire and the cross-sectional area of ​​the steel wire of the stranded wire are given; The optimized weight of the stranded wire is obtained by comparing the cross-sectional area of ​​the aluminum wire and the cross-sectional area of ​​the steel wire of the stranded wire.

11. The method according to claim 10, characterized in that The construction of the optimization model includes: The evaluation conditions are constructed based on the relationship between the reliability of the strands under the current stranding parameters and the set reliability; The objective optimization function is constructed based on the minimum weight of the strand.

12. The method according to claim 11, characterized in that The optimization model is shown as follows: Where: W(X) is the weight of the stranded wire; X1 is the total area of ​​the aluminum alloy monofilament; X2 is the total area of ​​the steel wire in the stranded wire; ρ1 is the density of aluminum; ρ2 is the density of steel; P{g(X)} is the reliability of the aluminum alloy monofilament in the stranded wire; R s Allows wear reliability for stranded wire.

13. An optimization system for the wear assessment method according to any one of claims 1 to 8, characterized in that: include: The wear state function calculation module is used to calculate the wear state function of the aluminum alloy single wire of the stranded wire by using the response surface function mapping relationship of the BP neural network model; A wear reliability calculation module, which calculates the wear reliability of the aluminum alloy single wire of the stranded wire under wind vibration conditions using a Monte Carlo method based on the wear state function of the aluminum alloy single wire of the stranded wire; The adjustment module is used to continuously adjust the design size of the aluminum alloy single wire of the stranded wire based on the pre-built optimization model, so that the wear reliability of the aluminum alloy single wire of the stranded wire is close to the allowable wear reliability of the stranded wire, and on this basis, the cross-sectional area of ​​the aluminum wire and the cross-sectional area of ​​the steel wire of the stranded wire are given; The optimization module is used to calculate the cross-sectional area of ​​the aluminum wire of the optimized stranded wire and the cross-sectional area of ​​the steel wire to obtain the optimized weight of the stranded wire.

14. The system of claim 13, wherein: The system also includes a building module for building an optimization model; The building blocks include: An evaluation condition unit, used for constructing an evaluation condition based on the relationship between the reliability of the stranded wire under the current stranded wire parameters and the set reliability; The target optimization function unit is used to construct the target optimization function based on the minimum weight of the strand.

Images