Multi-objective hot rolling quality detection method based on progressive layered extraction

Abstract

Disclosed in the present invention is a multi-objective hot rolling quality detection method based on progressive layered extraction. The present invention comprises the following steps: firstly, collecting hot rolling production process data, and determining a plurality of hot rolling quality detection parameters corresponding to the hot rolling production process data, so as to construct a hot rolling production data set; then performing data preprocessing on the hot rolling production data set to obtain a preprocessed data set; then training, on the basis of the preprocessed data set, a multi-objective network based on progressive layered extraction to obtain a multi-objective hot rolling quality detection model; and finally, inputting preprocessed hot rolling production data to be predicted into the multi-objective hot rolling quality detection model, outputting a multi-objective hot rolling quality detection result, and calculating predicted values of the plurality of hot rolling quality detection parameters on the basis of the multi-objective hot rolling quality detection result. According to the present invention, by performing weighted fusion on features of different quality detection parameters, regularization is achieved, thereby improving the accuracy and anti-interference performance of hot rolling quality detection.

Description

A multi-objective hot rolling quality detection method based on progressive layered extraction Technical Field

[0001] The present invention relates to a product quality detection parameter determination method for hot rolling production process fault prediction and diagnosis, and in particular to a multi-target hot rolling quality detection method based on progressive layered extraction. Background Art

[0002] Hot rolling is widely used in the manufacture of general structural parts such as buildings, bridges, ships, and vehicles due to its low energy consumption, good plastic processing, low deformation resistance, insignificant work hardening, and easy rolling. However, since the hot rolling production process is affected by variables such as rolling force, rolling speed, roll gap value, bending roll force, and temperature, it has the characteristics of multivariate, strong coupling, nonlinearity, and time-varying. At the same time, since the quality of hot-rolled products is determined by multiple parameters at the same time, how to accurately evaluate the various quality inspection parameters in the hot rolling production process has become an important guarantee and challenge to ensure product quality.

[0003] Currently, the evaluation and testing of hot rolling production processes can be based on two approaches: mechanism models and data models. Mechanism models require a significant amount of prior knowledge to establish the model. Furthermore, these methods simplify and idealize hot rolling production, leaving their robustness and practicality untested. Data-driven approaches, on the other hand, do not require extensive prior knowledge. Furthermore, with the advancement of hardware and the growth of data volumes in recent years, more and more researchers are turning their attention to data-driven approaches. These data-driven approaches have found widespread application in many scenarios and demonstrate high generalizability and robustness.

[0004] However, the application of traditional data-driven methods in the industrial field is greatly limited. First, the data of the hot rolling production process has a lot of noise; second, due to the different customer demands for products, such as production specifications and quantities, the data of hot rolling production is highly unbalanced, and the performance of traditional data-driven methods on unbalanced data is often unsatisfactory; third, it is understood that the data-driven methods currently used in the field of fault prediction and diagnosis of hot rolling production processes are all single-objective prediction models. Although they can achieve good prediction results on some parameters, their performance on other parameters often deteriorates. At the same time, the single-objective model also requires additional time for model training, and also requires additional storage space and computing power for real-time prediction.

[0005] Summary of the Invention

[0006] In view of this, in order to solve the problems existing in the background technology, the purpose of the present invention is to propose a multi-objective hot rolling quality detection method based on progressive hierarchical extraction, so that it can realize the simultaneous prediction of multiple quality detection parameters on an unbalanced hot rolling production data set under noise interference, and improve the accuracy of the prediction.

[0007] The technical solutions of the present invention are as follows:

[0008] Step 1: Collect hot rolling production process data, determine multiple hot rolling quality inspection parameters corresponding to the hot rolling production process data, and thus construct a hot rolling production data set;

[0009] Step 2: Preprocess the hot rolling production data set to obtain a preprocessed data set;

[0010] Step 3: Based on the preprocessed data set, a multi-objective network based on progressive hierarchical extraction is trained to obtain a multi-objective hot rolling quality detection model;

[0011] Step 4: Input the preprocessed hot rolling production data to be predicted into the multi-objective hot rolling quality detection model, output the multi-objective hot rolling quality detection results, and calculate the predicted values ​​of multiple hot rolling quality detection parameters based on the multi-objective hot rolling quality detection results.

[0012] In step 1, the hot rolling production process data includes rolling force, rolling speed, bending roll force, roll gap value, roll shifting amount, rolling thickness and inlet temperature.

[0013] In step 1, the hot rolling quality inspection parameters include convexity, centerline offset, outlet temperature, wedge shape, width and straightness.

[0014] The step 2 is specifically as follows:

[0015] Firstly, the outliers in the hot rolling production dataset are eliminated according to the 3-sigma principle to obtain the eliminated dataset; then, the eliminated dataset is normalized to obtain the normalized dataset; finally, the data in the normalized dataset is processed using the sliding window method to obtain the preprocessed dataset.

[0016] The multi-objective network based on progressive hierarchical extraction consists of a first-layer progressive hierarchical extraction network, a second-layer progressive hierarchical extraction network and a multi-objective output network;

[0017] The first-layer progressive hierarchical extraction network consists of k first private networks, 1 first shared network, and k+1 first gating networks; the second-layer progressive hierarchical extraction network consists of k second private networks, 1 second shared network, and k second gating networks; the multi-target output network consists of k output networks;

[0018] The inputs of the multi-target network based on progressive hierarchical extraction are all used as inputs of k first private networks and 1 first shared network, the k first gating networks are respectively connected to the corresponding first private networks, the k first gating networks are also respectively connected to the first shared network, and the remaining first gating network is connected to both the k first private networks and the first shared network. The inputs of the multi-target network based on progressive hierarchical extraction are all connected to k+1 first gating networks, and the outputs of the k+1 first gating networks are used as the outputs of the first layer of progressive hierarchical extraction network and are respectively connected to the k second private networks and 1 second shared network of the second layer of progressive hierarchical extraction network;

[0019] The k second private networks are respectively connected to the corresponding second gating networks, the second shared networks are all connected to the k second gating networks, the outputs of the k first gating networks are also respectively connected to the corresponding second gating networks, and the k second gating networks are respectively connected to the k output networks.

[0020] The k first private networks and the one first shared network have the same structure, and are all composed of a convolutional neural network, a gated recurrent network, and a long short-term memory network in parallel, wherein the convolutional neural network is composed of two one-dimensional convolutional layers, one one-dimensional maximum pooling layer, and one fully connected layer connected in sequence, the gated recurrent network is composed of one gated recurrent unit and one fully connected layer connected in sequence, and the long short-term memory network is composed of one long short-term memory unit and one fully connected layer connected in sequence;

[0021] The k second private networks and the one second shared network have the same structure, and are both composed of three fully connected layer networks in parallel.

[0022] The k+1 first gating networks have the same structure, and are both composed of one layer of gated recurrent units and one layer of fully connected layers connected in sequence; the k second gating networks have the same structure, and are both composed of two layers of fully connected layers connected in sequence.

[0023] The single-objective loss function of the multi-objective network based on progressive hierarchical extraction during training is the mean square error, the initial weight of each single-objective loss function is 1, and training is performed according to the dynamic weighted average rule.

[0024] In step 4, the pre-processed hot rolling production data to be predicted is obtained by performing the pre-processing in step 2 on the collected hot rolling production data to be predicted.

[0025] In step 4, after the multi-objective hot rolling quality detection results are denormalized, predicted values ​​of multiple hot rolling quality detection parameters are obtained.

[0026] The beneficial effects of the present invention are:

[0027] 1. The present invention proposes a multi-objective prediction method for hot-rolled product quality inspection parameters. Through the progressive layered extraction method, the features between different quality inspection parameters can be integrated to perform model regularization and improve the model prediction accuracy.

[0028] 2. Compared with the single-objective data-driven model, the present invention can obtain relatively satisfactory prediction results using less training data, and can overcome the situation where there may be insufficient sample data for a single type of hot rolling in the hot rolling production process.

[0029] 3. The present invention can predict multiple quality detection parameters simultaneously, taking up less space and using shorter prediction time to obtain more accurate prediction results. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] FIG1 is a flow chart of the method of the present invention.

[0031] FIG2 is a network structure diagram of the present invention.

[0032] FIG3 is a diagram showing the experimental results of the present invention. DETAILED DESCRIPTION

[0033] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, but the protection scope of the present invention is not limited thereto.

[0034] As shown in FIG1 , the embodiments of the present invention are as follows:

[0035] Step 1: Collect hot rolling production process data, determine multiple hot rolling quality detection parameters corresponding to the hot rolling production process data, and thus construct a hot rolling production data set; Data set It can be expressed as: Among them, x n represents the time series of the operating variables in the nth hot rolling production process, y k represents the time series of the kth quality detection parameter, where k is the number of quality detection parameters;

[0036] Hot rolling process data includes rolling force, rolling speed, bending force, roll gap, roll shifting, roll thickness, and inlet temperature. Hot rolling quality inspection parameters include crown, centerline offset, outlet temperature, wedge shape, width, and straightness.

[0037] Step 2: Preprocess the hot rolling production data set to obtain a preprocessed data set;

[0038] Step 2 is as follows:

[0039] First, outliers in the hot rolling production data set are eliminated according to the 3-sigma principle. Specifically, data that meets [μ-3×σ,μ+3×σ] is selected, where μ is the average value of a single operating variable / parameter and σ is the standard deviation of a single operating variable / parameter, to obtain the eliminated data set; then, the eliminated data set is normalized to obtain a normalized data set; finally, the data in the normalized data set is processed using the sliding window method to obtain a preprocessed data set, which consists of multiple input matrices.

[0040] Among them, the normalization formula is:

[0041] Among them, x is the original data, x max 、x min are the maximum and minimum values ​​in the original data, respectively, x norm is the data after normalization;

[0042] Step 3: Based on the preprocessed data set, a multi-objective network based on progressive hierarchical extraction is trained to obtain a multi-objective hot rolling quality detection model;

[0043] As shown in Figure 2, the multi-objective network based on progressive hierarchical extraction consists of a first-layer progressive hierarchical extraction network, a second-layer progressive hierarchical extraction network, and a multi-objective output network. The first-layer progressive hierarchical extraction network consists of k first private networks, 1 first shared network, and k+1 first gated networks, where k is the number of hot rolling quality detection parameters. The second-layer progressive hierarchical extraction network consists of k second private networks, 1 second shared network, and k second gated networks. The multi-objective output network consists of k output networks.

[0044] The inputs of the multi-target network based on progressive hierarchical extraction are all used as inputs of k first private networks and 1 first shared network. The k first gating networks are respectively connected to the corresponding first private networks, that is, one first gating network is connected one-to-one with a corresponding first private network. The k first gating networks are also respectively connected to the first shared network. The remaining first gating network (that is, the k+1th first gating network) is connected to both the k first private networks and the first shared network. The inputs of the multi-target network based on progressive hierarchical extraction are all connected to the k+1 first gating networks. The outputs of the k+1 first gating networks serve as the outputs of the first-layer progressive hierarchical extraction network and are respectively connected to the k second private networks and 1 second shared network of the second-layer progressive hierarchical extraction network.

[0045] The k second private networks are connected to the corresponding second gating networks, that is, one second private network is connected to the corresponding second gating network one by one, and the second shared network is connected to the k second gating networks. The outputs of the k first gating networks are also connected to the corresponding second gating networks, and the k second gating networks are connected to the k output networks. Each output network consists of two fully connected layers connected in sequence. The output of the kth task is y k is: y k =t k (r k (x)).

[0046] The structures of the k first private networks and the 1 first shared network are the same, both consisting of convolutional neural networks. Gated Recurrent Networks and long short-term memory networks Parallel construction refers to the simultaneous input matrix input to each of the three sub-networks. The convolutional neural network consists of two one-dimensional convolutional layers, one one-dimensional max pooling layer, and one fully connected layer, in sequence. The gated recurrent network consists of one gated recurrent unit layer and one fully connected layer, in sequence. The long short-term memory network consists of one long short-term memory unit layer and one fully connected layer, in sequence. The k+1 first gated networks g have the same structure, consisting of one gated recurrent unit layer and one fully connected layer, in sequence.

[0047] The k second private networks and the one second shared network have the same structure, both consisting of three fully connected layer networks in parallel. The k second gating networks have the same structure, both consisting of two fully connected layers connected in sequence.

[0048] Among them, the process of the first layer of progressive hierarchical extraction network can be expressed as:

[0049] Among them, r k The output of the network is progressively extracted layer by layer for the k-th task.

[0050] In step 4, the input matrix is ​​divided into training set, test set and validation set, with the ratio of 20% training set, 5% test set and 75% validation set. Construct the loss function of the multi-objective network based on progressive hierarchical extraction It is a single objective loss function and the corresponding weight λ k The sum of the products can be expressed as:

[0051] Among them, the single objective loss function θ k is the mean square error, λ k(t) represents the weight of the kth task during the tth training, θ s Represents the parameters of the shared network, y k represents the observed value of the kth quality detection parameter; Represents the predicted value of the kth quality detection parameter; the initial weight of each single-objective loss function is 1, and training is performed according to the dynamic weighted average rule. The weight update rule is:

[0052] Among them, ω k (t) is the loss ratio between the previous round and the previous round, ω i (t) represents the loss ratio of the i-th task in the previous round and the previous round, T is the smoothing coefficient, satisfying It calculates the rate of loss decrease in each round and assigns higher weights to tasks that are easy to decrease, thereby minimizing the overall weight.

[0053] Step 5: Input the pre-processed hot rolling production data to be predicted into the multi-objective hot rolling quality detection model, output the multi-objective hot rolling quality detection results, and calculate the predicted values ​​of multiple hot rolling quality detection parameters based on the multi-objective hot rolling quality detection results. Specifically, the multi-objective hot rolling quality detection results are denormalized (i.e. ), the predicted values ​​of multiple hot rolling quality detection parameters are obtained.

[0054] The pre-processed hot rolling production data to be predicted is obtained by performing the pre-processing in step 2 on the collected hot rolling production data to be predicted.

[0055] In the simulation experiment, hot rolling manufacturing field data were collected, including 59 operating variables and 6 quality inspection parameters, including crown, centerline offset, outlet temperature, wedge, width and straightness. Each group had 60,000 data, 12,000 data were used for training and 45,000 data were used for verification. The root mean square error (RMSE) and mean absolute error (MAE) were used as prediction performance. Among them, convolutional neural network (CNN), long short-term memory network (LSTM), gated recurrent network (GRU) and deep neural network (DNN) were used as comparison algorithms. The prediction results are shown in Table 1.

[0056] Table 1 is a comparison table of prediction results between the present invention and the existing method

[0057] As can be seen from the table, the PLE-MTL proposed in this paper achieves excellent prediction results for all six quality inspection parameters. Figure 3 shows the frequency histograms of the prediction errors of the six quality inspection parameters. Most of the errors are concentrated around 0, which proves that the proposed method has a higher performance in predicting product quality in the hot rolling production process.

[0058] The above embodiments are used to illustrate the present invention rather than to limit the present invention. Any modifications and changes made to the present invention within the spirit of the present invention and the protection scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A multi-objective hot rolling quality detection method based on progressive layered extraction, characterized in that: The following steps are involved: Step 1: Collect hot rolling production process data, determine multiple hot rolling quality detection parameters corresponding to the hot rolling production process data, and thus construct a hot rolling production data set; Step 2: preprocess the hot rolling production data set to obtain a preprocessed data set; Step 3: According to the preprocessed data set, a multi-objective network based on progressive hierarchical extraction is trained to obtain a multi-objective hot rolling quality detection model; Step 4: Input the preprocessed hot rolling production data to be predicted into the multi-objective hot rolling quality detection model, output the multi-objective hot rolling quality detection results, and calculate the predicted values ​​of multiple hot rolling quality detection parameters based on the multi-objective hot rolling quality detection results.

2. A multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 1, characterized in that: In step 1, the hot rolling production process data includes rolling force, rolling speed, bending roll force, roll gap value, roll shifting amount, rolling thickness and inlet temperature.

3. The multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 1 is characterized in that: In step 1, the hot rolling quality inspection parameters include convexity, centerline offset, outlet temperature, wedge shape, width and straightness.

4. The multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 1 is characterized in that: The step 2 is specifically as follows: Firstly, the outliers in the hot rolling production data set are eliminated according to the 3-sigma principle to obtain the eliminated data set; then, the eliminated data set is normalized to obtain the normalized data set; finally, the data in the normalized data set is processed using the sliding window method to obtain the preprocessed data set.

5. The multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 1 is characterized in that: The multi-objective network based on progressive hierarchical extraction is composed of a first-layer progressive hierarchical extraction network, a second-layer progressive hierarchical extraction network and a multi-objective output network; The first-layer progressive hierarchical extraction network is composed of k first private networks, 1 first shared network and k+1 first gating networks; the second-layer progressive hierarchical extraction network is composed of k second private networks, 1 second shared network and k second gating networks; the multi-target output network is composed of k output networks; The input of the multi-objective network based on progressive hierarchical extraction is used as the input of k first private networks and 1 first shared network. The k first gated networks are respectively connected to the corresponding first private networks, and the k first gated networks are also respectively connected to the first shared network. The remaining first gated network is connected to the k first The private network and the first shared network are connected, the input of the multi-objective network based on progressive hierarchical extraction is connected to k+1 first gating networks, the output of the k+1 first gating networks is used as the output of the first layer of progressive hierarchical extraction network and is respectively connected to k second private networks and 1 second shared network of the second layer of progressive hierarchical extraction network; The k second private networks are respectively connected to the corresponding second gating networks, the second shared networks are all connected to the k second gating networks, the outputs of the k first gating networks are also respectively connected to the corresponding second gating networks, and the k second gating networks are respectively connected to the k output networks.

6. A multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 5, characterized in that: The k first private networks and the one first shared network have the same structure, and are all composed of a convolutional neural network, a gated recurrent network, and a long short-term memory network in parallel, wherein the convolutional neural network is composed of two one-dimensional convolutional layers, one one-dimensional maximum pooling layer, and one fully connected layer connected in sequence, the gated recurrent network is composed of one gated recurrent unit and one fully connected layer connected in sequence, and the long short-term memory network is composed of one long short-term memory unit and one fully connected layer connected in sequence; The k second private networks and the one second shared network have the same structure, and are both composed of three fully connected layer networks in parallel.

7. The multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 5 is characterized in that: The k+1 first gating networks have the same structure, and are both composed of 1 layer of gated recurrent unit and 1 layer of fully connected layer connected in sequence; the k second gating networks have the same structure, and are both composed of 2 layers of fully connected layers connected in sequence.

8. The multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 1 is characterized in that: The single-objective loss function of the multi-objective network based on progressive hierarchical extraction during training is the mean square error, the initial weight of each single-objective loss function is 1, and training is performed according to the dynamic weighted average rule.

9. The multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 1 is characterized in that: In the step 4, the pre-processed hot rolling production data to be predicted is obtained by performing the pre-processing in step 2 on the collected hot rolling production data to be predicted.

10. The multi-objective hot rolling quality detection method based on progressive layered extraction according to claim 1 is characterized in that: In step 4, after the multi-objective hot rolling quality detection results are denormalized, the predicted values ​​of multiple hot rolling quality detection parameters are obtained.

Images