Impact polypropylene microstructure prediction method and system, and storage medium

Abstract

The present invention provides an impact polypropylene microstructure prediction method and system, and a storage medium. The impact polypropylene microstructure prediction method comprises the following steps: acquiring process parameters for processing impact polypropylene, the process parameters comprising a first raw material concentration and a homopolymerization time of at least one first raw material in a homopolymerization stage, and a second raw material concentration and a copolymerization time of at least one second raw material in a copolymerization stage; on the basis of the process parameters, determining a predicted value of a molecular weight characteristic of an impact polypropylene product; and on the basis of the difference between a target value and the predicted value of the molecular weight characteristic, adjusting the process parameters to obtain the impact polypropylene product having the target value. In the present invention, a large amount of historical data is used to train a neural network model, and prediction is performed, so as to more accurately adjust the process parameters for processing the impact polypropylene and improve the adaptability to a variety of process parameters, thereby improving the quality and performance of the impact polypropylene product.

Description

Method, system and storage medium for predicting microstructure of impact polypropylene TECHNICAL FIELD

[0001] The present application relates to the field of computerized chemical engineering, and in particular, to a method for predicting microstructure of impact polypropylene, a system for predicting microstructure of impact polypropylene, and a computer-readable storage medium. BACKGROUND

[0002] Impact polypropylene is a complex multi-component and multi-phase high molecular alloy system, and the matrix is polypropylene. By introducing an appropriate amount of impact modifier into the polypropylene matrix, the impact resistance and low-temperature toughness can be significantly improved while maintaining the excellent performance of polypropylene, and it has been widely used in the fields of automobiles, home appliances and home furnishing. With the continuous improvement of product quality requirements, high-performance impact polypropylene products have become the goal pursued by manufacturers. Researchers have continuously studied the process parameters and mechanical properties of impact polypropylene, and found that different polymerization process parameters will form different molecular weight characteristics, which in turn affect the mechanical properties such as strength and toughness of impact polypropylene. However, traditional preparation methods, such as mathematical models or chemical process simulation using Aspen software, have limited adaptability, especially for impact polypropylene, which has the characteristics of multi-component and complex phase structure. It is challenging to use mechanism models to predict its performance. Moreover, the various interactions and molecular arrangements involved in the polymerization process make it difficult to completely capture these complexities with a system of differential equations. Especially when considering factors such as polymer chain orientation, copolymer composition, and the influence of additives, the complexity and computational load of the model are further increased.

[0003] In order to overcome the above-mentioned defects existing in the prior art, there is an urgent need in the art for an improved method for predicting the microstructure of impact polypropylene, which can more accurately adjust the process parameters for processing impact polypropylene and improve the adaptability to various process parameters, thereby improving the quality and performance of impact polypropylene products. SUMMARY

[0004] The following presents a simplified summary of one or more aspects in order to provide a basic understanding of such aspects. This summary is not an extensive overview of all contemplated aspects, and is intended to neither identify key or critical elements of all aspects nor delineate the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form as a prelude to the more detailed description that is presented later.

[0005] In order to overcome the above-mentioned defects existing in the prior art, the application provides a method for predicting the microstructure of impact polypropylene, a system for predicting the microstructure of impact polypropylene and a computer readable storage medium, which can train and predict based on a neural network model and a large amount of historical data, so as to more accurately adjust the process parameters of the impact polypropylene, improve the adaptability to various process parameters, and thus improve the quality and performance of the impact polypropylene product.

[0006] Specifically, the method for predicting the microstructure of impact polypropylene according to the first aspect of the application comprises the following steps: obtaining process parameters for processing impact polypropylene. The process parameters include the first raw material concentration and the homopolymerization time of at least one first raw material in the homopolymerization stage, and the second raw material concentration and the copolymerization time of at least one second raw material in the copolymerization stage; determining the predicted value of the molecular weight characteristics of the impact polypropylene product according to the process parameters; and adjusting the process parameters according to the difference between the target value and the predicted value of the molecular weight characteristics, so as to obtain the impact polypropylene product meeting the target value.

[0007] Further, in some embodiments of the application, the at least one first raw material includes propylene and hydrogen. The at least one second raw material includes propylene, ethylene and hydrogen. The target value and the predicted value of the molecular weight characteristics correspond to at least one of the first number average molecular weight Mn_PP, the first weight average molecular weight Mw_PP, the first molecular weight distribution index PDI_PP, the first mass percentage PP_per of polypropylene, the second number average molecular weight Mn_iPP, the second weight average molecular weight Mw_iPP, the second molecular weight distribution index PDI_iPP of isotactic polypropylene, the third number average molecular weight Mn_aPP, the third weight average molecular weight Mw_aPP, the third molecular weight distribution index PDI_aPP of atactic polypropylene, the fourth number average molecular weight Mn_EPR, the fourth weight average molecular weight Mw_EPR, the fourth molecular weight distribution index PDI_EPR, the fourth mass percentage EPR_per of ethylene-propylene random copolymer, the first ethylene mole content E_mole_EPR, the fifth number average molecular weight Mn_EPS, the fifth weight average molecular weight Mw_EPS, the fifth molecular weight distribution index PDI_EPS, the fifth mass percentage EPS_per of ethylene-propylene multi-block copolymer, the second ethylene mole content E_mole_EPS, and the third ethylene mole content E_in_IPC in the polymer.

[0008] Further, in some embodiments of the application, the step of determining the predicted value of the molecular weight characteristics of the impact polypropylene product according to the process parameters comprises: inputting the process parameters into a pre-trained BP neural network model to determine the predicted value of the molecular weight characteristics of the impact polypropylene product via the BP neural network model.

[0009] Further, in some embodiments of the present application, the step of training the BP neural network model comprises: obtaining a plurality of sets of historical sample data of the processed impact polypropylene to construct a historical sample data set. Each of the historical sample data comprises process parameters and real values of molecular weight characteristics of the corresponding historical sample; building a BP neural network model to be trained. The BP neural network model is composed of an input layer, two hidden layers and an output layer; and inputting each set of training sample data in the historical sample data set into the BP neural network model in turn to obtain corresponding predicted values of the molecular weight characteristics, and adjusting learning parameters of the BP neural network model according to differences between the predicted values and real values of the molecular weight characteristics of the corresponding training sample, so as to train the BP neural network model.

[0010] Further, in some embodiments of the present application, before inputting each training sample in the historical sample data set into the BP neural network model in turn, the prediction method further comprises the following step: normalizing each historical sample data in the historical sample data set to normalize the numerical value between -1 and 1:

[0011] wherein, X reduced,p，i is a normalized variable of the pth historical sample data of the ith neuron. X p，i is an original variable of the pth historical sample data of the ith neuron. X min，i is a minimum value of the variable of the ith neuron in the historical sample data set. X max,i is a maximum value of the variable of the ith neuron in the historical sample data set.

[0012] Further, in some embodiments of the present application, the step of inputting each set of training sample data in the historical sample data set into the BP neural network model in turn to obtain corresponding predicted values of the molecular weight characteristics comprises: dividing the historical sample data set into a training set and a test set according to a predetermined proportion at random; defining the historical sample data in the training set as training sample data; and inputting each set of training sample data in the training set into the BP neural network model in turn to obtain corresponding predicted values of the molecular weight characteristics.

[0013] Further, in some embodiments of the present application, the following steps are further included: defining the historical sample data in the test set as test sample data; in response to completing the training of the BP neural network model, inputting each set of test sample data in the test set into the BP neural network model in turn to obtain corresponding test values of the molecular weight characteristics; and verifying the effectiveness of the BP neural network model according to a fitting degree of a regression curve of each test value and a real value of the corresponding test sample.

[0014] Further, in some embodiments of the present application, the number of neurons N i of the input layer is determined according to the input dimension of each group of the training sample data. The number of neurons N o of the output layer is determined according to the output dimension of the molecular weight characteristic prediction value. The number of neurons of the hidden layer is determined according to the number of neurons N i of the input layer, the number of neurons N o of the output layer, and the number of groups N s of the training sample data.

[0015] wherein the value of a depends on the type of the molecular weight characteristic to be predicted, and ranges from 4 to 10.

[0016] Further, in some embodiments of the present application, the step of sequentially inputting each group of the training sample data in the historical sample data set into the BP neural network model to obtain the corresponding molecular weight characteristic prediction value comprises: initializing the weight w j，i connecting each neuron i, j in the BP neural network model to a normal distribution:

[0017] wherein U is a function of normal distribution, n in is the number of rows in the training sample data matrix, and n out is the number of columns in the training sample data matrix; initializing the threshold value B i of each neuron i in the BP neural network model to 0; calculating the net input variable of each layer of neural network according to the weight connecting each neuron and the threshold value of the neurons of each layer of neural network:

[0018] wherein, is the net input variable of the pth data of the jth neuron of the k+1th layer of neural network, is the threshold value of the jth neuron of the k+1th layer of neural network, is the weight connecting the jth neuron of the k+1th layer of neural network and the ith neuron of the kth layer of neural network, is the output variable of the pth data of the ith neuron of the kth layer of neural network, and the output variable of the jth neuron of the k+1th layer of neural network is calculated from the net input variable of the jth neuron of the k+1th layer of neural network according to the transfer calculation of the self-defined activation function:

[0019] wherein, is the net input variable of the pth data of the jth neuron of the k+1th layer of neural network, is the output variable of the pth data of the jth neuron of the k+1th neural network, and a1 is 1.5.

[0020] Further, in some embodiments of the present application, the step of adjusting the learning parameters of the BP neural network model according to the difference between the predicted value and the true value of the molecular weight characteristics of the corresponding training sample to train the BP neural network model comprises: calculating the mean square error between the predicted value and the true value of the molecular weight characteristics of each training sample:

[0021] wherein, is the true value of the jth neuron of the pth training sample. is the predicted value of the jth neuron of the pth training sample; and using the Adam optimizer to perform back propagation on the BP neural network model based on the mean square error until the training of the BP neural network model is completed.

[0022] Further, in some embodiments of the present application, the step of using the Adam optimizer to perform back propagation on the BP neural network model based on the mean square error until the training of the BP neural network model is completed comprises: setting the maximum iteration number t max , an indication parameter θ1, and a patience parameter to prevent overfitting of training. The value of the patience parameter is determined according to the formula patience θ = max(input_n, output-n) / 2. input_n is the number of input nodes. output_n is the number of output nodes. The initial value of the indication parameter θ1 is 0; in response to the current number of rounds t max not reaching the maximum iteration number t t , and the indication parameter θ1 being less than or equal to the patience parameter, combining the first moment estimate m t and the second moment estimate v t of the Adam optimizer to update the weights of the connections between the neurons in the BP neural network model and the thresholds of the neurons:

[0023] wherein, w t , b t is the updated weight and threshold of the tth round. w t-1 , b t-1 is the weight and threshold of the t-1th round. a2 is the learning rate. θ2 is the anti-zero constant. The expression of the first moment estimate m t of the Adam optimizer is:

[0024] m t = β1mt-1 + (1 - β1)h t

[0025] wherein m t-1 and m t represent the first moment estimates before and after iteration, respectively, and β1is an exponential weighted average parameter. h t represents the gradient of the weight ω or the threshold value b.

[0026] The second moment estimate v t of the Adam optimizer has the expression:

[0027] v t = β2v t-1 + (1 - β2)h t 2

[0028] wherein v t-1 and v t are the second moment estimates before and after iteration, respectively, and β2is an exponential weighted average parameter. h t 2 represents the square of the gradient of the weight ω or the threshold value b; it is determined whether the mean square error of the current round is increased, and if so, the indicator θ1is added by 1, otherwise the indicator θ1is set to zero; and in response to the current round number t reaching the maximum iteration round number t max , or the indicator θ1being greater than the patience parameter, it is determined that the training of the BP neural network model is completed.

[0029] In addition, the prediction system of the microstructure of the impact polypropylene according to the second aspect of the present application comprises a memory and a processor. The memory has computer instructions stored thereon. The processor is connected to the memory and is configured to execute the computer instructions stored on the memory to implement the prediction method of the microstructure of the impact polypropylene according to the first aspect of the present application.

[0030] In addition, the computer readable storage medium according to the third aspect of the present application has computer instructions stored thereon. When the computer instructions are executed by a processor, the prediction method of the microstructure of the impact polypropylene according to the first aspect of the present application is implemented. BRIEF DESCRIPTION OF DRAWINGS

[0031] The above features and advantages of the present application can be better understood by reading the following detailed description of embodiments of the present application in conjunction with the drawings, in which various elements of the drawings are not necessarily drawn to scale and in which like or similar elements are identified with the same or similar reference numerals throughout the several views. Embodiments of the present application will be described and explained with additional specificity and detail through the use of the accompanying drawings in which:

[0032] FIG. 1 shows a structural schematic diagram of a BP neural network model according to some embodiments of the present application.

[0033] FIG. 2A shows a regression plot of the second number average molecular weight Mn_iPP according to some embodiments of the application.

[0034] FIG. 2B shows a regression plot of the second weight average molecular weight Mw_iPP according to some embodiments of the application.

[0035] FIG. 3A shows a regression plot of the fourth number average molecular weight Mn_EPR of ethylene-propylene random copolymer according to some embodiments of the application.

[0036] FIG. 3B shows a regression plot of the fifth number average molecular weight Mn_EPS of ethylene-propylene multi-block copolymer according to some embodiments of the application.

[0037] FIG. 4 shows a flow diagram of a method for predicting the microstructure of impact polypropylene according to some embodiments of the application. DETAILED DESCRIPTION

[0038] The specific embodiments of the present application will now be described in detail with specific reference being made to the figures. The following description of the embodiments is merely exemplary in nature and is in no way intended to limit the scope of the application, its application, or its uses. While the application can be susceptible to embodiments in various specific contexts, the application can be understood more readily by reference to the following detailed description and the accompanying drawings. As will become readily apparent to those skilled in the art from the following description, the embodiments do not restrict the application to the specific embodiments described herein, but rather the scope of the application is to be determined from the appended claims. While the application is amenable to various modifications and alternative forms, specific embodiments have been shown by way of example in the drawings and are described in detail below. It should be understood, however, that the intention is not to limit the application to the particular embodiments described or illustrated. On the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the application as defined by the appended claims.

[0039] In the description of the present application, it is necessary to explain that, unless explicitly defined and limited, the terms "mounting", "connection", "connecting" should be understood in a broad sense, for example, it can be fixed connection, or detachable connection, or integrally connected; it can be mechanical connection, or electrical connection; it can be directly connected, or indirectly connected through intermediate medium, or the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0040] In addition, "up", "down", "left", "right", "top", "bottom", "horizontal", "vertical" used in the following description should be understood as the orientation shown in the section and the related drawings. The relative terms are only for the convenience of description, and they do not mean that the devices described should be manufactured or operated in a particular orientation, so they should not be understood as a limitation on the application.

[0041] It is to be understood that, although terms such as "first", "second", "third", etc. can be used herein to describe various components, regions, layers and / or sections, these components, regions, layers and / or sections should not be limited by these terms, and these terms are only used to distinguish different components, regions, layers and / or sections. Therefore, the first components, regions, layers and / or sections discussed below can be referred to as the second components, regions, layers and / or sections without departing from some embodiments of the present application.

[0042] As described above, the conventional preparation method, such as mathematical model-based or chemical process simulation using Aspen software, has a limited adaptability problem, especially for the impact polypropylene with multiple components and complex phase structure characteristics, using mechanism model to predict its performance is challenging. And the various interactions and molecular arrangement involved in the polymerization process make it difficult to completely capture these complexities in the differential equation system. Especially when considering the orientation of the polymer chain, the composition of the copolymer, the influence of additives and other factors, the complexity of the model and the amount of calculation are further increased.

[0043] In order to overcome the above-mentioned defects existing in the prior art, the present application provides a method for predicting the microstructure of impact polypropylene, a system for predicting the microstructure of impact polypropylene and a computer readable storage medium, which can be based on a neural network model and trained and predicted using a large amount of historical data, for more accurately adjusting the process parameters of the impact polypropylene and improving the adaptability to various process parameters, thereby improving the quality and performance of the impact polypropylene product.

[0044] In some non-limiting embodiments, the above-mentioned method for predicting the microstructure of impact polypropylene according to the first aspect of the present application can be implemented based on the above-mentioned system for predicting the microstructure of impact polypropylene according to the second aspect of the present application. Specifically, the system for predicting the microstructure of impact polypropylene can be configured with a memory and a processor. The memory includes but is not limited to the above-mentioned computer readable storage medium according to the third aspect of the present application, on which computer instructions are stored. The processor is connected to the memory and is configured to execute the computer instructions stored on the memory to implement the above-mentioned method for predicting the microstructure of impact polypropylene according to the first aspect of the present application.

[0045] Specifically, the processor can train the BP neural network model offline in advance. The processor can first acquire a plurality of historical sample data of the impact polypropylene to construct a historical sample data set. In this case, each historical sample data includes the process parameters and the real value of the molecular weight characteristics of the corresponding historical sample.

[0046] Please refer to FIG. 1. FIG. 1 shows a structural schematic diagram of a BP neural network model according to some embodiments of the present application.

[0047] As shown in FIG. 1, the processor can build a BP neural network model to be trained. Here, the BP neural network model is composed of an input layer 11, two hidden layers 12, and an output layer 13.

[0048] After that, in some preferred embodiments, the processor can first normalize each historical sample data in the historical sample data set to normalize its value between -1 and 1:

[0049] where X reduced,p，i is the normalized variable of the i-th neuron for the p-th historical sample data, X p，i is the original variable of the i-th neuron for the p-th historical sample data, X min,i is the minimum value of the variable of the i-th neuron in the historical sample data set, and X max，i is the maximum value of the variable of the i-th neuron in the historical sample data set.

[0050] After that, the processor can randomly divide the historical sample data set into a training set and a test set according to a preset ratio (for example, 8:2). After that, the processor can define the historical sample data in the training set as training sample data, and input each group of training sample data in the training set into the BP neural network model in turn to obtain the corresponding molecular weight characteristic prediction value.

[0051] In addition, in the process of determining the structure of the above-mentioned BP neural network model, the number of neurons N i in the input layer is determined according to the input dimension of each group of training sample data, the number of neurons N o in the output layer is determined according to the output dimension of the molecular weight characteristic prediction value, and the number of neurons in the hidden layer is determined according to the number of neurons N i in the input layer, the number of neurons N o in the output layer, and the number of groups N s of training sample data:

[0052] where the value of a depends on the type of molecular weight characteristic to be predicted, and the value range is 4-10. Here, when predicting the molecular weight characteristics of both homopolymer and copolymer, a can be set to 4, when predicting the molecular weight characteristics of homopolymer, a can be set to 8, and when predicting the molecular weight characteristics of copolymer, a can be set to 10.

[0053] Further, after inputting each group of training sample data in the training set into the BP neural network model in turn, the processor can initialize the weight w j，i connecting each neuron i, j in the BP neural network model to a normal distribution:

[0054] wherein U is a normal distribution function, n in is the number of rows in the training sample data matrix, n out is the number of columns in the training sample data matrix.

[0055] Then, the processor can initialize the threshold value B i of each neuron i in the BP neural network model to 0, and then calculate the net input variable of the BP neural network model according to the weights of the connections between neurons and the threshold values of the neurons of each layer of neural network:

[0056] wherein, is the net input variable of the pth data of the jth neuron of the k+1th layer of neural network, is the threshold value of the jth neuron of the k+1th layer of neural network, is the weight of the connection between the jth neuron of the k+1th layer of neural network and the ith neuron of the kth layer of neural network, is the output variable of the pth data of the ith neuron of the kth layer of neural network, and the output variable of the jth neuron of the k+1th layer of neural network is calculated from the net input variable of the jth neuron of the k+1th layer of neural network according to the transfer calculation of the self-defined activation function:

[0057] wherein, is the net input variable of the pth data of the jth neuron of the k+1th layer of neural network, is the output variable of the pth data of the jth neuron of the k+1th layer of neural network, and a1 takes a value of 1.5.

[0058] In this way, the processor can further process the variables in the hidden layer and the output layer using the same above formula, and then obtain the output variable of the BP neural network model.

[0059] After that, the processor can adjust the learning parameters of the BP neural network model according to the difference between the predicted value and the true value of the molecular weight characteristics of the corresponding training sample, to train the BP neural network model.

[0060] Specifically, the processor can calculate the mean square error between the predicted value and the true value of the molecular weight characteristics of each training sample:

[0061] wherein, is the true value of the pth training sample of the jth neuron, is the predicted value of the pth training sample of the jth neuron.

[0062] Then, the processor can perform back propagation based on mean square error on the BP neural network model using the Adam optimizer until the training of the BP neural network model is completed.

[0063] Specifically, in the process of back propagation on the BP neural network, the processor can first set the maximum iteration number t max , an indication parameter θ1, and a patience parameter for preventing overfitting of the training. Here, the value of the patience parameter is determined according to the formula patience θ = max(input_n, output_n) / 2, input_n is the number of input nodes, output_n is the number of output nodes, and the initial value of the indication parameter θ1 is 0. Here, in the process of predicting the molecular weight characteristics of the homopolymer, the patience parameter patience θ may be set to 5. In the process of predicting the molecular weight characteristics of the copolymer, the patience parameter patience θ may be set to 6.

[0064] Then, in response to the current iteration number t max not reaching the maximum iteration number t t and the indication parameter θ1 being less than or equal to the patience parameter, the processor can combine the first moment estimate m t and the second moment estimate v t of the Adam optimizer together to update the weights of the connections between the neurons in the BP neural network model and the thresholds of the neurons:

[0065] where w t and b t-1 are the updated weights and thresholds at the tth iteration, w t-1 and b -8 are the weights and thresholds at the (t-1)th iteration. α2 is the learning rate, which is 0.0025. θ2 is the anti-zero constant, which is 10 t .

[0066] Here, the expression of the first moment estimate m t of the Adam optimizer is:

[0067] m t-1 = β1m t + (1-β1)h t-1 t where m t and m t represent the first moment estimates before and after iteration, β1 is the exponential weighted average parameter, and h t represents the gradient of the weight ω or the threshold b.

[0069] Here, the second moment estimation v t of the Adam optimizer is expressed as:

[0070] v t = β2v t-1 + (1-β2)h t 2

[0071] where v t-1 and v t are the second moment estimations before and after iteration, β2is an exponential weighted average parameter, h t 2 represents the gradient square of the weight ω or the threshold b.

[0072] After that, the processor can determine whether the mean square error of the current round is increased, and if so, add 1 to the indication parameter θ1, otherwise set the indication parameter θ1to zero.

[0073] After that, in response to the current round number t reaching the maximum iteration round number t max , or the indication parameter θ1being greater than the patience parameter, the processor can determine that the training of the BP neural network model is completed.

[0074] In addition, in some preferred embodiments, the processor can also define the historical sample data in the test set as test sample data. In response to completing the training of the BP neural network model, the processor can input each group of test sample data in the test set into the BP neural network model in turn to obtain the corresponding molecular weight characteristic test value. After that, the processor can verify the effectiveness of the BP neural network model according to the fitting degree of the regression curve of each test value and the true value of the corresponding test sample.

[0075] Specifically, the processor can use three indicators, mean squared error (MSE), mean absolute error (MAE), and goodness of fit (r 2 ) to evaluate the performance:

[0076] where y i represents the true value of the test sample, represents the average value of the true value of the test sample, represents the true value of the test sample.

[0077] Herein, the target values and the predicted values of the molecular weight characteristics correspond to at least one of a first number average molecular weight Mn PP, a first weight average molecular weight Mw PP, a first molecular weight distribution index PDI PP, a first mass percentage PP per of the polypropylene, a second number average molecular weight Mn iPP, a second weight average molecular weight Mw iPP, a second molecular weight distribution index PDI iPP of the isotactic polypropylene, a third number average molecular weight Mn aPP, a third weight average molecular weight Mw aPP, a third molecular weight distribution index PDI aPP of the atactic polypropylene, a fourth number average molecular weight Mn EPR, a fourth weight average molecular weight Mw EPR, a fourth molecular weight distribution index PDI EPR, a fourth mass percentage EPR per of the ethylene-propylene random copolymer, a first ethylene mole content E mole EPR, a fifth number average molecular weight Mn EPS, a fifth weight average molecular weight Mw EPS, a fifth molecular weight distribution index PDI EPS, a fifth mass percentage EPS per of the ethylene-propylene multi-block copolymer, a second ethylene mole content E mole EPS, and a third ethylene mole content E in IPC in the polymer.

[0078] For details, please refer to Table 1, FIG. 2A and FIG. 2B. Table 1 shows a performance index table of the homopolymer molecular weight characteristics provided according to some embodiments of the present application. FIG. 2A shows a regression graph of the second number average molecular weight Mn iPP provided according to some embodiments of the present application. FIG. 2B shows a regression graph of the second weight average molecular weight Mw iPP provided according to some embodiments of the present application.

[0079] Table 1 Performance index table of homopolymer molecular weight characteristics

[0080] As shown in Table 1, when the value of a is 8, the BP neural network model predicts the molecular weight characteristics of the homopolymer. The mean square error of the BP neural network model is 0.00034036, the mean absolute error is 0.01100553, and the goodness of fit is 0.9966. As shown in FIG. 2A and FIG. 2B, the BP neural network model has low error and high accuracy for the molecular weight characteristics of the homopolymer.

[0081] For details, please refer to Table 2, FIG. 3A and FIG. 3B. Table 2 shows a performance index table of the copolymer molecular weight characteristics provided according to some embodiments of the present application. FIG. 3A shows a regression graph of the fourth number average molecular weight Mn EPR of the ethylene-propylene random copolymer provided according to some embodiments of the present application. FIG. 3B shows a regression graph of the fifth number average molecular weight Mn EPS of the ethylene-propylene multi-block copolymer provided according to some embodiments of the present application.

[0082] Table 2 Performance index table of copolymer molecular weight characteristics

[0083] As shown in Table 2, when the value of a is 10, the BP neural network model predicts the molecular weight characteristics of the copolymer. The mean square error of the BP neural network model is 0.00064456, the mean absolute error is 0.01680668, and the goodness of fit is 0.9601. As shown in FIGS. 3A and 3B, the BP neural network model has low error and high accuracy for the molecular weight characteristics of the copolymer.

[0084] Specifically, refer to FIG. 4. FIG. 4 shows a flowchart of a method for predicting the microstructure of the impact polypropylene according to some embodiments of the present application.

[0085] As shown in FIG. 4, during the online processing of the impact polypropylene, the processor can first obtain process parameters for processing the impact polypropylene. Here, the process parameters include the first raw material concentration of at least one first raw material and the homopolymerization time in the homopolymerization stage, and the second raw material concentration of at least one second raw material and the copolymerization time in the copolymerization stage.

[0086] Further, in some embodiments, the at least one first raw material includes propylene and hydrogen, and the at least one second raw material includes propylene, ethylene, and hydrogen.

[0087] Then, the processor can determine the predicted value of the molecular weight characteristics of the impact polypropylene product according to the process parameters.

[0088] Specifically, the processor can input the process parameters into a pre-trained BP neural network model to determine the predicted value of the molecular weight characteristics of the impact polypropylene product via the BP neural network model.

[0089] Further, the processor can adjust the process parameters according to the difference between the target value and the predicted value of the molecular weight characteristics to obtain the impact polypropylene product with the load target value.

[0090] In summary, the method for predicting the microstructure of the impact polypropylene, the system for predicting the microstructure of the impact polypropylene, and the computer readable storage medium provided by the present application can all be trained and predicted based on a neural network model and using a large amount of historical data, which can be used to more accurately adjust the process parameters for processing the impact polypropylene and improve the adaptability to various process parameters, thereby improving the quality and performance of the impact polypropylene product.

[0091] Although the above-described methods are illustrated and described as a series of acts for the sake of simplicity, it should be understood and appreciated that the methods are not limited by the order of acts, as some acts may, in accordance with one or more embodiments, occur in different orders and / or concurrently with other acts from that shown and described herein. And / or, depending on the embodiment, various elements of the methods could be implemented in hardware, software, or a combination of both.

[0092] Those skilled in the art will appreciate that information, signals, and data can be represented using any of a variety of different technologies and techniques. For example, data, instructions, commands, information, signals, bits, symbols, and chips that can be referenced throughout the above description can be represented by voltages, currents, electromagnetic waves, magnetic fields or particles, optical fields or particles, or any combination thereof.

[0093] Those skilled in the art will further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans can implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

[0094] Although the processor described in the above embodiments can be implemented by a combination of software and hardware, it is understood that the processor can be implemented in software or hardware. For hardware implementation, the processor can be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, micro-controllers, microprocessors, other electronic units, or a selective combination thereof. For software implementation, the processor can be implemented by separate software modules such as procedures and functions, each of which performs one or more of the functions and operations described herein in conjunction with the relevant hardware, if needed.

[0095] The various illustrative logical blocks, modules, and circuits described in connection with the embodiments disclosed herein can be implemented or performed with a general purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general purpose processor can be a microprocessor, but in the alternative, the processor can be any conventional processor, controller, microcontroller, or state machine. A processor can also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration.

[0096] The steps of a method or algorithm described in connection with the embodiments disclosed herein can be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module can reside in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium can be integral to the processor. The processor and the storage medium can reside in an ASIC. The ASIC can reside in a user terminal. In the alternative, the processor and the storage medium can reside as discrete components in a user terminal.

[0097] In one or more exemplary embodiments, the functions described can be implemented in hardware, software, firmware, or any combination thereof. If implemented in software as a computer program product, the functions can be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media can be any available media that can be accessed by a computer. By way of example, and not limitation, such computer-readable media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, if the software is transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. Disk and disc, as used herein, include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray® disc where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.

[0098] The previous description of the disclosure is provided to enable any person skilled in the art to make or use the disclosure. Various modifications to the disclosure will be readily apparent to those skilled in the art, and the generic principles defined herein can be applied to other variations without departing from the spirit or scope of the disclosure. Thus, the disclosure is not intended to be limited to the examples described herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of predicting the microstructure of an impact polypropylene, characterized in that, The method comprises the following steps: obtaining process parameters of processing impact polypropylene, wherein the process parameters comprise a first raw material concentration and a homopolymerization time of at least one first raw material in a homopolymerization stage, and a second raw material concentration and a copolymerization time of at least one second raw material in a copolymerization stage; determining a predicted value of a molecular weight characteristic of the impact polypropylene product according to the process parameters; and adjusting the process parameters according to a difference between a target value and the predicted value of the molecular weight characteristic, so as to obtain the impact polypropylene product loaded with the target value.

2. The prediction method of claim 1, wherein, The at least one first raw material comprises propylene and hydrogen, and the at least one second raw material comprises propylene, ethylene and hydrogen, The target value and the predicted value of the molecular weight characteristic correspond to at least one of a first number average molecular weight Mn_PP, a first weight average molecular weight Mw_PP, a first molecular weight distribution index PDI_PP, a first mass percentage PP_per, a second number average molecular weight Mn_iPP of isotactic polypropylene, a second weight average molecular weight Mw_iPP, a second molecular weight distribution index PDI_iPP, a third number average molecular weight Mn_aPP of atactic polypropylene, a third weight average molecular weight Mw_aPP, a third molecular weight distribution index PDI_aPP, a fourth number average molecular weight Mn_EPR of ethylene-propylene random copolymer, a fourth weight average molecular weight Mw_EPR, a fourth molecular weight distribution index PDI_EPR, a fourth mass percentage EPR_per, a first ethylene mole content E_mole_EPR, a fifth number average molecular weight Mn_EPS of ethylene-propylene multi-block copolymer, a fifth weight average molecular weight Mw_EPS, a fifth molecular weight distribution index PDI_EPS, a fifth mass percentage EPS_per, a second ethylene mole content E_mole_EPS, and a third ethylene mole content E_in_IPC in the polymer.

3. The prediction method of claim 1 or 2, characterized in that, The step of determining the predicted value of the molecular weight characteristic of the impact polypropylene product according to the process parameters comprises: inputting the process parameters into a pre-trained BP neural network model, so as to determine the predicted value of the molecular weight characteristic of the impact polypropylene product via the BP neural network model.

4. The prediction method of claim 3, wherein, The step of training the BP neural network model comprises: obtaining a plurality of sets of historical sample data of processing impact polypropylene, so as to construct a historical sample data set, wherein each historical sample data set comprises process parameters and a real value of a molecular weight characteristic of a corresponding historical sample; building a BP neural network model to be trained, wherein the BP neural network model is composed of an input layer, two hidden layers and an output layer; and inputting each set of training sample data in the historical sample data set into the BP neural network model in sequence, so as to obtain a corresponding predicted value of the molecular weight characteristic, and adjusting learning parameters of the BP neural network model according to a difference between the predicted value and a real value of the molecular weight characteristic of the corresponding training sample, so as to train the BP neural network model.

5. The prediction method of claim 4, wherein, Before inputting each training sample in the historical sample data set into the BP neural network model in sequence, the prediction method further comprises the following steps: normalizing each historical sample data in the historical sample data set to normalize its value between -1 and 1: wherein X reduced,p，i is the normalized variable of the pth historical sample data of the ith neuron, X p，i is the original variable of the pth historical sample data of the ith neuron, X min,i is the minimum value of the variable of the ith neuron in the historical sample data set, and X max,i is the maximum value of the variable of the ith neuron in the historical sample data set.

6. The prediction method of claim 4, wherein, The step of sequentially inputting each training sample data in the historical sample data set into the BP neural network model to obtain a corresponding molecular weight characteristic prediction value comprises: The historical sample data set is randomly divided into a training set and a test set according to a preset proportion; The historical sample data in the training set is defined as training sample data; and Each group of training sample data in the training set is sequentially inputted into the BP neural network model to obtain a corresponding molecular weight characteristic prediction value.

7. The prediction method of claim 6, wherein, The steps further comprise: The historical sample data in the test set is defined as test sample data; In response to completion of the training of the BP neural network model, each group of test sample data in the test set is sequentially inputted into the BP neural network model to obtain a corresponding molecular weight characteristic test value; and The effectiveness of the BP neural network model is verified according to the goodness of fit of a regression curve of each test value and a true value of the corresponding test sample.

8. The prediction method of claim 4, wherein, The number of neurons in the input layer N i The number of neurons N in the output layer is determined based on the input dimension of the training sample data in each group. o The number of neurons in the hidden layer is determined based on the output dimension of the predicted molecular weight characteristics, and the number of neurons in the hidden layer is based on the number of neurons N in the input layer. i The number of neurons N in the output layer o and the number of training sample data groups N s Sure: The value of α depends on the type of the molecular weight characteristic to be predicted, and ranges from 4 to 10.

9. The prediction method of claim 4, wherein, The step of sequentially inputting each group of training sample data in the historical sample data set into the BP neural network model to obtain a corresponding molecular weight characteristic prediction value comprises: The weights w of the connections between the neurons i, j in the BP neural network model are initialized to a normal distribution: j，i initialized to a normal distribution: where U is a normal distribution function, n in is the number of rows in the training sample data matrix, n out is the number of columns in the training sample data matrix. setting the threshold value B of each neuron i in the BP neural network model to i initialized to 0; and According to the weights of the inter-neuron connections of each layer neural network, and the threshold values of the neurons of each layer neural network, the net input variable of each layer neural network is calculated: wherein, is a net input variable for the pth piece of data for the jth neuron of the k+1th neural network, is a threshold value of the jth neuron of the (k+1)th neural network, Wk+1jkis the weight of the connection between the jth neuron of the (k+1)th neural network and the ith neuron of the kth neural network, is an output variable of the pth piece of data of the ith neuron of the kth neural network, and the output variable of the jth neuron of the k+1th neural network is calculated by the net input variable of the jth neuron of the k+1th neural network according to a self-defined activation function transfer: wherein a net input variable for the pth data of the jth neuron of the k+1th neural network, The output variable of the pth data of the jth neuron of the k+1th layer neural network is 1.

5.

10. The prediction method of claim 9, wherein, The step of adjusting the learning parameter of the BP neural network model according to the difference between the predicted value and the true value of the corresponding training sample to train the BP neural network model comprises: calculating a mean squared error of the predicted values and the true values of the molecular weight property for each of the training samples: wherein, is the true value of the pth training sample of the jth neuron, is the predicted value of the jth neuron of the pth training sample; and The BP neural network model is back propagated based on the mean square error using an Adam optimizer until the training of the BP neural network model is completed.

11. The prediction method of claim 10, wherein, The step of back propagating the BP neural network model based on the mean square error using an Adam optimizer until the training of the BP neural network model is completed comprises: Setting the maximum iteration round number t for training the BP neural network model max , an indication parameter θ1, and a patience parameter for preventing overfitting of training, wherein the value of the patience parameter is determined according to the formula patience θ = max(input_n, output_n) / 2, input_n is the number of input nodes, output_n is the number of output nodes, and the initial value of the indication parameter θ1 is 0; in response to the current round t not reaching the maximum iteration round t max , and the indication parameter θ1 being less than or equal to the patience parameter, combining the first moment estimation m t and the second moment estimation v t of the Adam optimizer to update the weights of connections between neurons in the BP neural network model and the thresholds of the neurons: where w t , b t are the updated weights and thresholds at the t-th round, w t-1 , b t-1 are the weights and thresholds at the (t-1)-th round, a2is the learning rate, and 2is the anti-zero constant. The expression of the first moment estimation m t of the Adam optimizer is: m t = β1m t-1 + (1 - β1)h t where m t-1 and m t represent the first moment estimates before and after iteration, respectively, β1is the exponential weighted average parameter, h t denotes the gradient of the weight ω or the threshold b, The expression of the second moment estimation v of the Adam optimizer t is: v t = β2v t-1 + (1 - β2)h t 2 where v t-1 and v t are the pre- and post-iteration second moment estimates, respectively, β2is an exponential weighting parameter, h t 2 represent the gradient square of the weight ω or the threshold b. It is judged whether the mean square error of the current round is increased, and if so, the indicator parameter θ1 is added by 1, otherwise the indicator parameter θ1 is set to zero; and in response to the current round t reaching the maximum iteration round t max or the indication parameter θ1 is greater than the patience parameter, it is determined that the training of the BP neural network model is completed.

12. A system for predicting the microstructure of an impact polypropylene, characterized in that, It comprises: a memory having computer instructions stored thereon; and a processor connected to the memory and configured to execute the computer instructions stored on the memory to implement the prediction method of the microstructure of the impact polypropylene according to any one of claims 1 to 11.

13. A computer readable storage medium having stored thereon computer instructions, wherein, The computer instructions are executed by the processor to implement the prediction method of the microstructure of the impact polypropylene according to any one of claims 1 to 11. The computer instructions are executed by the processor to implement the prediction method of the microstructure of the impact polypropylene according to any one of claims 1 to 11.

Images