A safety prediction method for bridge cables based on gradient lifting machine

Abstract

The present invention proposes a bridge cable safety prediction method based on a gradient hoist, and the specific steps are as follows: (1) Conduct a steel wire corrosion simulation test to obtain a steel wire corrosion time-tensile strength curve; (2) Conduct a high temperature simulation test after steel wire corrosion to obtain a temperature-tensile strength curve under different corrosion conditions; (3) Calculate and evaluate the safety of a single steel wire and the entire cable. Based on the current status of bridge cables, the present invention adopts a corrosion and high temperature coupling simulation test, obtains the high temperature corrosion coupling change law of bridge cables according to the gradient hoist, and evaluates the cable status.

Description

Technical Field

[0001] The invention belongs to the field of bridge engineering, and in particular is a method for predicting the safety of bridge cables based on a gradient lifting machine. Background Art

[0002] With the development of modern bridge engineering, bridges, as important transportation hubs, bear huge transportation pressure. However, the main material of bridge cable structure is steel, which has poor corrosion resistance and good thermal conductivity, making it threatened by high temperature of fire. Oxygen and moisture in the atmosphere can easily cause corrosion of steel. The popularization of energy transportation and trams has increased the risk of bridge fires. Since fireproof materials are always in the process of corrosion and will be affected by industrial products, after a fire occurs, protective materials are more susceptible to high temperatures and cause steel damage, affecting structural safety. Therefore, it is necessary to carry out research on cable safety after high temperature and corrosion coupling.

[0003] At present, some studies have tested and simulated the mechanical properties of cables under corrosion and high temperature, but the research on bridge cables under the coupled influence of high temperature and corrosion is still immature. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes a bridge cable safety prediction method based on a gradient hoist. As the service time of the cable increases, the mechanical properties of its material will be affected and damaged due to natural factors, human factors, etc., and even break. Through experiments, the degradation of structural performance after the coupling of high temperature and corrosion is obtained, and finally the tensile strength under the coupling of steel wire corrosion and high temperature is calculated.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] A bridge cable safety prediction method based on a gradient lifting machine, the specific steps are as follows, and the characteristics are as follows:

[0007] Step 1: Carry out corrosion tests on a single steel wire under different corrosion conditions and different corrosion times;

[0008] Step 2: After the corroded steel wire is subjected to a high temperature heating test, a tensile test is performed on it to obtain the tensile strength f of the corroded steel wire after heating at different high temperatures TE. C,T (i,TE), combine the measured data into a training data set and establish a gradient boosting machine model;

[0009] Step 3: According to the actual bridge fire and cable corrosion conditions, the corrosion condition C(i), corrosion time T(i), and the maximum temperature of the outermost steel wire of the cable T max , using the gradient lifting machine model to calculate the tensile strength f of a single steel wire under corrosion and high temperature C,T (i,TEmax );

[0010] Step 4: Based on the cable force f measured by the bridge monitoring system and the cross-section A0 of the cable, calculate the stress S(i) borne by a single steel wire i;

[0011] Step 5: According to the tensile strength f of a single steel wire under corrosion and high temperature C,T (i,TE max ), the stress S(i), when f C,T (i,TE max )>S(i), the single steel wire is safe, otherwise the single steel wire cannot continue to bear the load until it breaks;

[0012] Step 6: When the number of steel wires i broken is n, recalculate the effective cross-section A of the cable structure n , calculate the stress S(i) borne by a single steel wire i;

[0013] Step 7: Repeat steps 3 to 6 to calculate the wire breakage rate. When the wire breakage rate is greater than 2%

[0014] The cable structure is considered unsafe when the load is below 0.05, otherwise the cable structure is in a safe state.

[0015] Furthermore, the process of establishing the gradient boosting machine model in step 2 can be expressed as follows:

[0016] Step 2.1, initialize the model, initialize a constant model F0(x), this constant is the average value of the labels of all training data, x represents the corrosion condition, corrosion time, and steel wire temperature in the data sample;

[0017] Step 2.2, iterative training

[0018] Calculate the residual r mi :

[0019]

[0020] In the formula, r mi represents the residual of the ith sample at the mth iteration, where m = 1, 2, ..., M, m is the number of model iterations, and also represents the index of the weak learner currently being trained, and M is the total number of iterations or the number of weak learners; L(y i ,F(x i ) is the loss function, y i is the true label of the i-th sample, F(x i ) is the predicted value of the current model for the i-th sample, is to find the gradient;

[0021] Use the residual rmi As labels, we train a new weak learner h m (x); h m (x) represents the weak learner trained in the mth iteration, which is a decision tree; a constant γ is found by line search. m , so that F m (x) = F m-1 (x)+γ m h m (x) can minimize the loss function, and its loss function can be expressed as:

[0022]

[0023] Update the model as follows:

[0024] F m (x) = F m-1 (x)+vγ m h m (x)

[0025] Where v is the learning rate, which controls the m (x) Contribution to the final model, F m (x) is the cumulative model after the mth iteration, F m-1 (x) is the cumulative model after the m-1th iteration;

[0026] Step 2.3, model fusion, weighted and processed all weak learners to obtain the final gradient boosting machine F(x):

[0027]

[0028] Furthermore, in step 3, the gradient boosting model is used to calculate the tensile strength f of a single steel wire under corrosion and high temperature. C,T (i,TE max ) process can be expressed as follows:

[0029] f C,T (i,TE max )=F(x')

[0030] Where F() is the gradient lift, x' is the corrosion condition C(i), corrosion time T(i), and the maximum temperature of the outermost steel wire of the cable T according to the actual bridge fire and cable corrosion conditions. max .

[0031] The present invention provides a bridge cable safety prediction method based on a gradient hoist, with beneficial effects: The technical effects of the present invention are:

[0032] 1. The present invention fully considers the fact that the bridge cable structure is in a corrosive state in actual use and is subsequently affected by fire damage, and introduces corrosion damage, high temperature damage, and steel wire breakage, so that the calculation results are more in line with the actual situation and the calculated results are more accurate;

[0033] 2. The present invention has a clear idea, is easy to use in a process-based manner, and more accurately reflects the stress condition of the cable after corrosion and high temperature coupling, making the resistance borne by each steel wire clearer, and ultimately being able to more scientifically evaluate the structural safety of the cable after corrosion coupling. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 is a flow chart of the present invention;

[0035] Figure 2 The variation law of corrosion time and ultimate strength of the steel wire of the present invention;

[0036] Figure 3 It is a schematic diagram of a high temperature time-stress curve of a corroded steel wire of the present invention;

[0037] Figure 4 This is the temperature field distribution diagram of a single corroded steel wire of the present invention. DETAILED DESCRIPTION

[0038] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments:

[0039] This paper proposes a bridge cable safety prediction method based on gradient lifting machine, please refer to Figures 1 to 4 , the steps of the present invention are described in detail below.

[0040] Step 1: Carry out corrosion tests on a single steel wire under different corrosion conditions and different corrosion times;

[0041] Step 2: After the corroded steel wire is subjected to a high temperature heating test, a tensile test is performed on it to obtain the tensile strength f of the corroded steel wire after heating at different high temperatures TE. C,T (i,TE), combine the measured data into a training data set and establish a gradient boosting machine model;

[0042] Step 2.1, initialize the model, initialize a constant model F0(x), this constant is the average value of the labels of all training data, x represents the corrosion condition, corrosion time, and steel wire temperature in the data sample;

[0043] Step 2.2, iterative training

[0044] Calculate the residual r mi :

[0045]

[0046] In the formula, r mi represents the residual of the ith sample at the mth iteration, where m = 1, 2, ..., M, m is the number of model iterations, and also represents the index of the weak learner currently being trained, and M is the total number of iterations or the number of weak learners; L(y i ,F(x i ) is the loss function, y i is the true label of the i-th sample, F(x i ) is the predicted value of the current model for the i-th sample, is to find the gradient;

[0047] Use the residual r mi As labels, we train a new weak learner h m (x); h m (x) represents the weak learner trained in the mth iteration, which is a decision tree; a constant γ is found by line search. m , so that F m (x) = F m-1 (x)+γ m h m (x) can minimize the loss function, and its loss function can be expressed as:

[0048]

[0049] Update the model as follows:

[0050] F m (x) = F m-1 (x)+vγ m h m (x)

[0051] Where v is the learning rate, which controls the m (x) Contribution to the final model, F m (x) is the cumulative model after the mth iteration, F m-1 (x) is the cumulative model after the m-1th iteration;

[0052] Step 2.3, model fusion, weighted and processed all weak learners to obtain the final gradient boosting machine F(x):

[0053]

[0054] Step 3: According to the actual bridge fire and cable corrosion conditions, the corrosion condition C(i), corrosion time T(i), and the maximum temperature of the outermost steel wire of the cable T max , using the gradient lifting machine model to calculate the tensile strength f of a single steel wire under corrosion and high temperature C,T (i,TEmax );

[0055] f C,T (i,TE max )=F(x')

[0056] Where F() is the gradient lift, x' is the corrosion condition C(i), corrosion time T(i), and the maximum temperature of the outermost steel wire of the cable T according to the actual bridge fire and cable corrosion conditions. max .

[0057] Step 4: Based on the cable force f measured by the bridge monitoring system and the cross-section A0 of the cable, calculate the stress S(i) borne by a single steel wire i;

[0058] Step 5: According to the tensile strength f of a single steel wire under corrosion and high temperature C,T (i,TE max ), the stress S(i), when f C,T (i,TE max )>S(i), the single steel wire is safe, otherwise the single steel wire cannot continue to bear the load until it breaks;

[0059] Step 6: When the number of steel wires i broken is n, recalculate the effective cross-section A of the cable structure n , calculate the stress S(i) borne by a single steel wire i;

[0060] Step 7: Repeat steps 3 to 6 to calculate the wire breakage rate. When the wire breakage rate is greater than 2%

[0061] The cable structure is considered unsafe when the load is below 0.05, otherwise the cable structure is in a safe state.

[0062] The above description is only a preferred embodiment of the present invention and does not constitute any other form of limitation to the present invention. Any modification or equivalent change made based on the technical essence of the present invention still falls within the scope of protection required by the present invention.

Claims

1. A bridge cable safety prediction method based on a gradient lifting machine, the specific steps are as follows, characterized in that: Step 1: Carry out corrosion tests on a single steel wire under different corrosion conditions and different corrosion times; Step 2: After the corroded steel wire is subjected to a high temperature heating test, a tensile test is performed on it to obtain the tensile strength f of the corroded steel wire after heating at different high temperatures TE. C,T (i,TE), combine the measured data into a training data set and establish a gradient boosting machine model; The process of establishing the gradient boosting machine model in step 2 is expressed as follows: Step 2.1, initialize the model, initialize a constant model F0(x), this constant is the average value of the labels of all training data, x represents the corrosion condition, corrosion time, and steel wire temperature in the data sample; Step 2.2, iterative training Calculate the residual r mi : In the formula, r mi represents the residual of the ith sample at the mth iteration, where m = 1, 2, ..., M, m is the number of model iterations, and also represents the index of the weak learner currently being trained, and M is the total number of iterations or the number of weak learners; L(y i ,F(x i ) is the loss function, y i is the true label of the i-th sample, F(x i ) is the predicted value of the current model for the i-th sample, is to find the gradient; Use the residual r mi As labels, we train a new weak learner h m (x); h m (x) represents the weak learner trained in the mth iteration, which is a decision tree; a constant γ is found by line search. m , so that F m (x) = F m-1 (x)+γ m h m (x) can minimize the loss function, and its loss function can be expressed as: Update the model as follows: F m (x)=F m-1 (x)+vγ m h m (x) Where v is the learning rate, which controls the m (x) Contribution to the final model, F m (x) is the cumulative model after the mth iteration, F m-1 (x) is the cumulative model after the m-1th iteration; Step 2.3, model fusion, weighted and processed all weak learners to obtain the final gradient boosting machine F(x): Step 3: According to the actual bridge fire and cable corrosion conditions, the corrosion condition C(i), corrosion time T(i), and the maximum temperature of the outermost steel wire of the cable T max , using the gradient lifting machine model to calculate the tensile strength f of a single steel wire under corrosion and high temperature C,T (i,TE max ); In step 3, the gradient lifting model is used to calculate the tensile strength f of a single steel wire under corrosion and high temperature. C,T (i,TE max ) process is as follows: f C,T (in,THE max )=F(x') Where F() is the gradient lift, x' is the corrosion condition C(i), corrosion time T(i), and the maximum temperature of the outermost steel wire of the cable T measured based on the actual bridge fire and cable corrosion conditions. max； Step 4: Based on the cable force f measured by the bridge monitoring system and the cross-section A0 of the cable, calculate the stress S(i) borne by a single steel wire i; Step 5: According to the tensile strength f of a single steel wire under corrosion and high temperature C,T (i,TE max ), the stress S(i), when f C,T (i,TE max )>S(i), the single steel wire is safe, otherwise the single steel wire cannot continue to bear the load until it breaks; Step 6: When the number of steel wires i broken is n, recalculate the effective cross-section A of the cable structure n , calculate the stress S(i) borne by a single steel wire i; Step 7: Repeat steps 3 to 6 to calculate the wire breakage rate. When the wire breakage rate is greater than 2% The cable structure is considered unsafe when the load is below 0.05, otherwise the cable structure is in a safe state.

Images