Testing Method for Heat Stabilizer Performance of PVC Products

Abstract

The present invention relates to the technical field of material testing, and particularly to a method for testing the performance of heat stabilizers for PVC products, including: obtaining a multi-dimensional data sequence formed by production data and test data of PVC products, and a standard sequence corresponding to the data sequence of each dimension; obtaining the degree of trend influence of the data sequence of the PVC product under each dimension according to the difference distribution between the data sequences of different dimensions of any batch of PVC products and the corresponding standard sequences; obtaining the degree of influence of environmental conditions of each batch of PVC products according to the clustering characteristics of the degrees of trend influence of the corresponding data sequences of all batches of PVC products in all dimensions; and using a neural network to analyze the degrees of influence of all environmental conditions of the PVC product to evaluate the performance of the stabilizer in the PVC product. The present invention improves the accuracy of the evaluation results of the performance of heat stabilizers in PVC products.

Description

Technical Field

[0001] The present invention relates to the technical field of material detection, and particularly to a method for detecting the performance of heat stabilizers for PVC products. Background Art

[0002] Polyvinyl chloride (PVC) is a commonly used thermoplastic resin. Due to its low price and good performance, it is widely used in many industries. However, industrially produced PVC has some unstable structures, such as secondary chlorine, terminal allyl chloride, etc. These unstable structures will cause the PVC molecular chain to break when heated, resulting in a darker overall color of the product and a decrease in mechanical properties. Heat stabilizer modification is to slow down the degradation rate of PVC by adding an appropriate amount of stabilizer during production, mainly by replacing unstable Cl atoms and destroying the double bond conjugated system to achieve its function.

[0003] The performance detection of heat stabilizers for PVC products is commonly carried out by experiments to detect the content ratio of heat stabilizers and practical effects, and there are various detection methods, including heat aging oven testing, thermogravimetric testing, etc. In the actual production environment, the performance of heat stabilizers should be obtained in a complex production environment, that is, it can effectively respond to the complex factor changes in the production environment and ensure the stable quality of PVC products. Summary of the Invention

[0004] The present invention provides a method for detecting the performance of heat stabilizers for PVC products to solve the existing problems.

[0005] The method for detecting the performance of heat stabilizers for PVC products of the present invention adopts the following technical solutions:

[0006] An embodiment of the present invention provides a method for detecting the performance of heat stabilizers for PVC products, which includes the following steps:

[0007] Obtain data sequences formed by the production data and test data of PVC products respectively, and use each data sequence as the data sequence of PVC products in a corresponding dimension, and obtain the standard sequence corresponding to the data sequence of each dimension;

[0008] According to the difference distribution between the data sequences of different dimensions of any batch of PVC products and the corresponding standard sequences, obtain the trend influence degree of the data sequences of the PVC products in each dimension;

[0009] According to the clustering characteristics of the trend influence degrees of the data sequences corresponding to all dimensions of all batches of PVC products, obtain the environmental condition influence degree of each batch of PVC products;

[0010] Analyze the influence degree of all environmental conditions on PVC products using a neural network, so as to evaluate the performance of stabilizers in PVC products.

[0011] Obtain the trend influence degree of the data sequence of the PVC product in each dimension according to the difference distribution between the data sequences of different dimensions of any batch of PVC products and the corresponding standard sequences. The specific methods included are:

[0012] According to the difference between the data points in the data sequence of any dimension of any batch of PVC products and the data points in the corresponding standard sequence, obtain the trend difference parameter of the data points in the data sequence of any dimension of the PVC product;

[0013] According to the trend difference parameters of all data points within the preset neighborhood range of the data points in the data sequence of any dimension of each batch of PVC products, obtain the trend influence coefficient of the data points in the data sequence of any dimension of each batch of PVC products;

[0014] According to the difference between the trend influence coefficients between all adjacent data points in the data sequence of any dimension, obtain the trend influence degree of the data sequence of the dimension.

[0015] The specific method for obtaining the trend difference parameter is:

[0016] For each data point on the data sequence of each dimension, obtain the difference in value between the data point and the data point in the standard sequence in the same order, and record it as the trend difference parameter of the data point in the data sequence of the corresponding dimension.

[0017] The method for obtaining the trend influence coefficient of the data points in the data sequence of any dimension of each batch of PVC products according to the trend difference parameters of all data points within the preset neighborhood range of the data points in the data sequence of any dimension of each batch of PVC products includes the following specific methods:

[0018] For the th batch of PVC products in the th dimension of the data sequence, for the th data point, the specific calculation method of the corresponding trend influence coefficient is:

[0019]

[0020] Among them, represents the trend influence coefficient of the th data point in the data sequence of the th dimension of the th batch of PVC products, represents the preset range radius, represents the The trend difference parameter corresponding to the th data point within the preset neighborhood range of the th data point on the data sequence of the th production batch of PVC products,

[0021] The method for obtaining the trend influence degree of the data sequence of the dimension according to the differences between the trend influence coefficients between all adjacent data points in the data sequence of any dimension specifically includes:

[0022] For the data sequence of the th batch of PVC products under the th dimension, the specific calculation method for the corresponding trend influence degree is:

[0023]

[0024] Wherein, represents the trend influence degree of the data sequence of the th batch of PVC products under the th dimension, represents the number of data points in the data sequence of the th batch of PVC products in the th dimension, represents the trend influence coefficient of the th data point in the data sequence of the th dimension of the th batch of PVC products, represents the trend influence coefficient of the th data point in the data sequence of the th dimension of the th batch of PVC products, represents taking the absolute value.

[0025] The method for obtaining the environmental condition influence degree of each batch of PVC products according to the clustering characteristics of the trend influence degrees of the corresponding data sequences of all batches of PVC products under all dimensions specifically includes:

[0026] Clustering all batches of PVC products using the trend influence degrees of the corresponding data sequences under all dimensions to obtain several clustering clusters and the maximum clustering center;

[0027] Obtain the standard degree of the data sequence of any dimension of PVC products in any batch. According to the standard degree and the relative distance between the cluster center of the cluster where the PVC product is located and the maximum cluster center, calculate the influence degree of environmental conditions on the PVC products in any dimension of this batch.

[0028] The method of clustering all batches of PVC products by using the trend influence degree of the corresponding data sequences in all dimensions to obtain several clusters and the maximum cluster center specifically includes:

[0029] Obtain the vector composed of the trend influence degrees of the data sequences in all dimensions of any batch of PVC products, denoted as the trend influence vector of the PVC products in the corresponding batch. According to the distances between the trend influence vectors of the PVC products in different batches, and use the density clustering algorithm to cluster all batches of PVC products to obtain several clusters;

[0030] Denote the cluster center of the cluster corresponding to the largest number of data points in all clusters as the maximum cluster center.

[0031] The method of obtaining the standard degree of the data sequence of any dimension of any batch of PVC products specifically includes:

[0032] Obtain the cosine similarity between the data sequence of the th dimension of any batch of PVC products and the standard sequence, denoted as the standard degree of the data sequence of the th dimension of the PVC products in this batch.

[0033] The method of calculating the influence degree of environmental conditions on the PVC products in any dimension of this batch according to the standard degree and the relative distance between the cluster center of the cluster where the PVC product is located and the maximum cluster center specifically includes:

[0034] According to the standard degree and the relative distance between the cluster center of the cluster where the PVC product is located and the maximum cluster center, calculate the environmental condition influence factor of the PVC products in any dimension of this batch;

[0035] Use the maximum-minimum normalization function to normalize the environmental condition influence factor to obtain the environmental condition influence degree.

[0036] The specific method of obtaining the environmental condition influence factor is:

[0037] Take the ratio of the distance between the cluster center of the cluster where the th batch of PVC products is located and the maximum cluster center to the average distance between the cluster centers of all clusters and the maximum cluster center as the The relative distance between the clustering center of the PVC products in each batch and the maximum clustering center;

[0038] According to the relative distance and the th batch of PVC products in the th dimension of the data sequence of the standard degree to obtain the environmental condition influence factor, the relative distance is positively correlated with the environmental condition influence factor, and the standard degree is negatively correlated with the environmental condition influence factor.

[0039] The beneficial effects of the technical solution of the present invention are as follows: By analyzing the differences between the production data of PVC products and the standard sequence batch by batch and dimension by dimension, the embodiments of the present invention can accurately identify the influence of heat stabilizers in each batch of PVC products. Through the clustering analysis of the influence degree of the trend of the multi-dimensional data sequence of PVC products, the PVC products of different batches can be classified according to the influence of environmental conditions, and the differences in performance and influencing factors can be identified. That is, through clustering analysis, batches with relatively consistent environmental condition influences are grouped into one category, which can effectively group the data, avoid subjective interference and noise influence in traditional methods, and thus obtain a more accurate quantification of the environmental condition influence degree, improving the accuracy of the subsequent evaluation results of the performance of heat stabilizers in PVC products. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0041] Figure 1 It is a flowchart of the steps of the method for detecting the performance of heat stabilizers for PVC products of the present invention;

[0042] Figure 2 It is the basic process of the method for analyzing the influence degree of the trend of the data sequence of PVC products in each dimension provided by an embodiment of the present invention;

[0043] Figure 3 It is the basic process of the method for analyzing the influence degree of environmental conditions of each batch of PVC products provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0044] To further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following specifically describes, in conjunction with the accompanying drawings and preferred embodiments, the performance detection method for heat stabilizers used in PVC products according to the present invention, its specific implementation manner, structure, features and effects as follows. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0045] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0046] The following specifically describes the specific solution of the performance detection method for heat stabilizers used in PVC products provided by the present invention in conjunction with the accompanying drawings.

[0047] Please refer to Figure 1 , which shows the step flowchart of the performance detection method for heat stabilizers used in PVC products provided by an embodiment of the present invention. The method includes the following steps:

[0048] Step S001: Obtain the data sequences formed by the production data and test data of the PVC product respectively, and use each data sequence as the data sequence of the PVC product in a corresponding dimension, and obtain the standard sequence corresponding to the data sequence of each dimension.

[0049] It should be noted that since in the production environment of PVC products, the corresponding production process is usually carried out in the form of production batches, the production data and test data of different batches of PVC products of the same specification are extracted.

[0050] Specifically, to implement the performance detection method for heat stabilizers used in PVC products proposed in this embodiment, it is first necessary to collect production data and test data. The specific process is as follows:

[0051] First, use a variety of sensors to obtain production data at the same sampling frequency. The production data includes temperature data, pressure data, and humidity data. The temperature data is sampled by a temperature sensor set on the production line and forms time-series data according to several sampling moments. Similarly, the pressure data of the equipment during the production of the heat stabilizer for PVC products and the humidity data of the production environment are obtained through a pressure sensor and a humidity sensor respectively.

[0052] It should be noted that the standard sequence is the expected standard data under ideal production or test conditions, which is obtained from the specific implementation environment. For example, the standard sequence of the thermogravimetric experiment is the data sequence of the mass change of the PVC product without adding heat stabilizer with the change of temperature.

[0053] Then, the test data includes the sequence data obtained from the thermal aging oven test and thermogravimetric test on the heat stabilizers of PVC products in all batches. For example, the thermogravimetric test results are the data sequence of the sample weight change with temperature change.

[0054] In addition, it should be noted that in this embodiment, for the production data and test data of each batch of PVC products, each type of data is regarded as a dimension respectively, that is, one dimension corresponds to one data sequence.

[0055] Finally, obtain the standard sequence corresponding to the data sequence of each dimension.

[0056] It should be noted that the standard sequence is the expected standard data sequence under ideal production or test conditions, which is obtained from the specific implementation environment. For example, the standard sequence of the thermogravimetric experiment is the data sequence of the mass change of PVC products without adding heat stabilizers with temperature change.

[0057] So far, the production data and test data of the heat stabilizer of PVC products are obtained through the above method.

[0058] Step S002: Obtain the trend influence degree of the data sequence of the PVC product under each dimension according to the difference distribution between the data sequences of different dimensions of any batch of PVC products and the corresponding standard sequences.

[0059] It should be noted that there is a correlation between the production data and the test data. Due to the sampling property of the test data, the reflection of the data information in the real production environment is not accurate enough. Therefore, during the data processing process, the data in the aggregated form should be extracted. The more it conforms to the aggregation characteristics, the more it proves that the data reflects the influence of the condition fluctuations in the production environment on the performance of the heat stabilizer, that is, the improvement direction of the heat stabilizer addition.

[0060] Since the performance detection of the heat stabilizer of PVC products is commonly carried out by experiments to detect the content ratio of the heat stabilizer and the practical effect, and there are various detection methods, including thermal aging oven test, thermogravimetric test, etc., the test results come from the samples of sampling detection, while the production data of the obtained PVC products are more actual environmental data and various quality indicators, especially with a large difference in the embodiment of the sample detection results. For example, the change of some environmental data in the production environment has a greater impact on the quality of PVC products, and the sampling results cannot effectively adjust for environmental differences.

[0061] Taking the thermogravimetric test results as an example, the thermal degradation of PVC mainly occurs in two stages: the first stage is at 180-380 °C (I), mainly releasing HCl and gradually forming a conjugated polyene structure, and the PVC will age and change color; the second stage is at 380-520 °C (II), at this time the carbon skeleton begins to break, and the conjugated polyene structure degrades to form carbon-containing residues. Then, there will be a trend change in the dimensional data of the production and detection of a single PVC product itself. Therefore, there will be a deviation in the results when measuring the performance of heat stabilizers in the PVC product production process through the differences in single-dimensional data values. Especially for the data at the transition part of multiple stages of a single-dimensional trend, it is necessary to analyze from the trend changes on both sides of the data points and adjust the dimensional data value differences to measure the quantization accuracy of heat stabilizer performance.

[0062] Please refer to Figure 2 , which shows the basic process of the method for analyzing the trend influence degree of the data sequence of a PVC product provided by an embodiment of the present invention in each dimension.

[0063] Specifically, in step S201, according to the difference between the data points in the data sequence of any dimension of any batch of PVC products and the data points in the corresponding standard sequence, the trend difference parameter of the data points in the data sequence of any dimension of the PVC products is obtained.

[0064] As an embodiment, the specific method for obtaining the trend difference parameter is:

[0065] For each data point on the data sequence of each dimension, obtain the difference in value between the data point and the data point in the standard sequence in the same order, and record it as the trend difference parameter of the data point in the data sequence of the corresponding dimension.

[0066] It should be noted that the value of the trend difference parameter can be negative. For each data point, the greater the cumulative change in the differences on both sides, the more obvious the trend change feature at the position of the data point.

[0067] In step S202, according to the trend difference parameters of all data points within a preset neighborhood range of the data points in the data sequence of any dimension of each batch of PVC products, the trend influence coefficient of the data points in the data sequence of any dimension of each batch of PVC products is obtained.

[0068] As an embodiment, for the th batch of PVC products in the th dimension, for the th data point in the data sequence, the specific calculation method of the corresponding trend influence coefficient is:

[0069]

[0070] Among them, represents the th batch of PVC products in the th data sequence in the th data point's trend influence coefficient, represents the preset range radius, represents the th production batch of PVC products in the th data sequence of the th data point's preset neighborhood range of the th data point's corresponding trend difference parameter, represents the linear normalization function.

[0071] It should be noted that the represented trend influence coefficient is actually a measure of the trend of the data point deviating from the standard sequence. The more obvious the deviation trend, the greater the possibility that this data point is affected by the heat stabilizer or the data deviation caused by the corresponding production environment conditions, rather than the random fluctuation situation.

[0072] It should be noted that according to experience, can be adjusted according to the actual situation, and no specific limitation is made in this embodiment.

[0073] Step S203, according to the difference between the trend influence coefficients between all adjacent data points in the data sequence of any dimension, obtain the trend influence degree of the data sequence of this dimension.

[0074] It should be noted that for the data sequences under different dimensions, the associated change of the trend influence coefficients of all data points reflects the size of the influence of the heat stabilizer on the corresponding type of this dimension.

[0075] As an embodiment, for the th batch of PVC products in the th data sequence of the

[0076]

[0077] Among them, represents the th batch of PVC products in the th data sequence of the represents the th batch of PVC products in the th data sequence of the represents the th batch of PVC products in the The trend influence coefficient of the th data point in the data sequence of the th batch of PVC products for the th data sequence in the th dimension, indicating to obtain the absolute value.

[0078] It should be noted that the th batch of PVC products for the difference in the trend influence coefficient between two adjacent data points in the data sequence of the

[0079] th dimension reflects the continuity of the trend influence on the dimension data sequence. That is, the more consistent the trend changes between the data points on the sequence, the better the effect of the dimension data on the performance response of the heat stabilizer.

[0080] Up to this point, the degree of trend influence of the data sequence of each dimension of each batch of PVC products is obtained through the above method.

[0081] It should be noted that since the degree of trend influence of the obtained data in different dimensions is the association between the data change and the type obtained from the trend of the respective dimension data, and in fact, there is a certain response difference between the change situation of the environmental data in the production process of PVC products and the difference in the test data. However, due to the limited sampling of the test data, the corresponding relationship between the difference in the test data and the difference in the corresponding production environment data is not obvious. Therefore, it needs to be extracted in the form of clustering.

[0082] Please refer to Figure 3 , which shows the basic process of the method for analyzing the degree of environmental condition influence of each batch of PVC products provided by an embodiment of the present invention.

[0083] Specifically, in step S301, all batches of PVC products are clustered using the degree of trend influence of the corresponding data sequences in all dimensions to obtain several clustering clusters and the largest clustering center.

[0084] As an embodiment, the specific process is as follows:

[0085] First, obtain the vector composed of the trend influence degrees of the data sequences under all dimensions of any batch of PVC products, denoted as the trend influence vector of the corresponding batch of PVC products. According to the distances between the trend influence vectors of different batches of PVC products, and using the density clustering algorithm to cluster all batches of PVC products, several clustering clusters are obtained.

[0086] Then, denote the clustering center corresponding to the largest number of data points in all clustering clusters as the maximum clustering center, and obtain the distance between the clustering center of any clustering cluster and the maximum clustering center.

[0087] It should be noted that among several clustering clusters, when the number of data points in a clustering cluster (since a PVC product is regarded as a data point in the clustering space during the clustering process, the number of data points in the clustering cluster is the number of PVC products included in the clustering cluster) is the largest, the corresponding clustering cluster represents the general situation under the current production conditions, and other clustering clusters can be regarded as significant differences that occur under the fluctuation of some environmental factors.

[0088] It should be noted that the distribution of data points in a clustering cluster reflects the data differences in the actual production process under the same production conditions, that is, the clustering differences in production conditions. When there are also large differences in its test data, it indicates that its production data has a greater impact on the test results.

[0089] Step S302, obtain the standard degree of the data sequence of any dimension of any batch of PVC products, and calculate the environmental condition influence degree of any dimension of the PVC products of this batch according to the standard degree and the relative distance between the clustering center of the clustering cluster where the PVC products are located and the maximum clustering center.

[0090] As an embodiment, the specific method for obtaining the environmental condition influence degree of any batch of PVC products is as follows:

[0091] First, obtain the cosine similarity between the data sequence of the th dimension of any batch of PVC products and the standard sequence, denoted as the standard degree of the data sequence of the th dimension of the PVC products of this batch.

[0092] Then, the specific calculation method of the environmental condition influence degree is:

[0093]

[0094] Among them, represents the environmental condition influence degree of the th dimension of the PVC products of the th batch, Represents the standard degree of the dimensional data sequence of the PVC products of the th batch, represents the distance between the clustering center of the clustering cluster where the PVC products of the th batch are located and the maximum clustering center, represents the average distance between the clustering centers of all clustering clusters and the maximum clustering center, represents the maximum - minimum normalization function.

[0095] It should be noted that represents the relative distance between the clustering center of the clustering cluster where the PVC products of the th batch are located and the maximum clustering center, and the environmental condition influence factor represents the influence performance of the production condition change on the test result. The larger the value, the greater the influence degree of the environmental condition.

[0096] It should be noted that since the ordinals corresponding to the batches, dimensions, and data points in this embodiment are all obtained by counting the discrete data in sequence, therefore, the , and values of all take positive integers.

[0097] Thus far, the influence degree of environmental conditions for different production batches of PVC products is obtained through the above - mentioned method.

[0098] Step S004: Analyze the influence degree of all environmental conditions of PVC products by using a neural network, so as to evaluate the performance of stabilizers in PVC products.

[0099] Specifically, the influence degree of environmental conditions for different production batches of PVC products is obtained, which is used for the construction of the neural network model for accurate analysis of the performance of heat stabilizers in PVC products. The specific steps are as follows:

[0100] 1. Collect the production and test data of multiple batches of PVC products, and clean the data to handle missing values and outliers to ensure data quality. Then, perform standardization processing on the feature data to ensure that different features have the same scale. Finally, divide the data set into a training set and a test set, and ensure that each batch of data has a corresponding weight value for use in subsequent model training.

[0101] 2. Select the multi - layer perceptron (MLP) as the basic architecture of the neural network. The specific steps include adding an input layer, several hidden layers (each layer contains a certain number of neurons and activation functions), and an output layer. This structure can capture the complex patterns and features of the data, so as to effectively evaluate the performance of stabilizers in PVC products.

[0102] 3. During the process of constructing the neural network model, select an appropriate loss function, such as the mean squared error (MSE). In model training, introduce sample weights so that data from different batches can be adjusted according to their weight importance during the calculation of the loss function, making the model more sensitive to data with larger weights.

[0103] 4. When compiling the model, select an appropriate optimizer (such as Adam), and specify the loss function and evaluation metrics (such as mean squared error or R²). Then, use the training set and sample weights to train the model, and monitor the loss and evaluation metrics during the training process in real time. After training is completed, use the test set data to evaluate the performance of the model, and verify the prediction effect of the model by calculating the loss value and other evaluation metrics.

[0104] 5. After preliminary training and evaluation, adjust the hyperparameters (such as the number of layers, the number of neurons, the learning rate, etc.) according to the model performance, and conduct multiple trainings and evaluations to optimize the model. Further verify the stability and generalization ability of the model through methods such as cross-validation. Finally, deploy the trained model to the production environment to be used for real-time evaluation of the performance of stabilizers in PVC products, and optimize the production process in combination with real-time data to improve product quality.

[0105] So far, this embodiment is completed.

[0106] It should be noted that the model used in this embodiment is only used to represent the negative correlation relationship and constrain the results of the model output to be within the interval. Specifically in implementation, it can be replaced with other models with the same purpose. This embodiment only takes the model as an example for description, and does not make specific limitations on it, where refers to the input of the model.

[0107] The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for detecting the performance of heat stabilizers for PVC products, characterized in that: The method comprises the following steps: Obtain data sequences formed by production data and test data of PVC products respectively, and use each data sequence as a data sequence of PVC products in a corresponding dimension, and obtain a standard sequence corresponding to the data sequence in each dimension; According to the difference distribution between the data series of different dimensions of any batch of PVC products and the corresponding standard series, the trend influence degree of the data series of the PVC products in each dimension is obtained; According to the clustering characteristics of the trend influence degree of the corresponding data series of all batches of PVC products in all dimensions, the influence degree of environmental conditions of each batch of PVC products is obtained; Use neural networks to analyze the impact of all environmental conditions on PVC products, thereby evaluating the performance of stabilizers in PVC products; The specific method for obtaining the trend influence degree of the data sequence of the PVC products in each dimension according to the difference distribution between the data sequence of different dimensions of any batch of PVC products and the corresponding standard sequence is as follows: Obtaining trend difference parameters of data points in the data sequence of any dimension of any batch of PVC products according to differences between data points in the data sequence of any dimension of the PVC products and data points in the corresponding standard sequence; Obtaining a trend influence coefficient of a data point in a data sequence of any dimension for each batch of PVC products according to a trend difference parameter of all data points in a preset neighborhood range for the data point in the data sequence of any dimension for each batch of PVC products; According to the differences between the trend influence coefficients of all adjacent data points in a data sequence of any dimension, the trend influence degree of the data sequence of the dimension is obtained; The method of obtaining the trend influence coefficient of the data points in the data sequence of any dimension of each batch of PVC products according to the trend difference parameters of all data points in the preset neighborhood range in the data sequence of any dimension of each batch of PVC products includes the following specific methods: For Batch of PVC products in the In the data sequence under the dimension data points, the specific calculation method of the corresponding trend influence coefficient is: in, Indicates Batch of PVC products in the In the data sequence under the dimension The trend influence coefficient of the data point is Indicates the preset range radius. Indicates The first batch of PVC products The first dimension of the data sequence within a preset neighborhood of data points The trend difference parameter corresponding to the data points is represents the linear normalization function.

2. The method for detecting the performance of heat stabilizers for PVC products according to claim 1, characterized in that: The specific method for obtaining the trend difference parameter is: For each data point in the data sequence of each dimension, the difference in value between the data point and the data points in the same order in the standard sequence is obtained, and recorded as the trend difference parameter of the data point in the data sequence of the corresponding dimension.

3. The method for detecting the performance of heat stabilizers for PVC products according to claim 1, characterized in that: The method of obtaining the trend influence degree of the data sequence of any dimension according to the difference between the trend influence coefficients of all adjacent data points in the data sequence of the dimension includes the following specific methods: For Batch of PVC products in the The specific calculation method for the corresponding trend impact degree of the data series under the dimension is: in, Indicates Batch of PVC products in the The degree of influence of the trend of the data series under the dimension, Indicates Batch of PVC products The number of data points in the data series of dimensions, Indicates Batch of PVC products In the data series of dimensions The trend influence coefficient of the data point is Indicates Batch of PVC products In the data series of dimensions The trend influence coefficient of the data point is Gets the absolute value.

4. The method for detecting the performance of heat stabilizers for PVC products according to claim 1, characterized in that: The method of obtaining the degree of environmental condition influence of each batch of PVC products according to the aggregation characteristics of the trend influence degree of the corresponding data series of all batches of PVC products in all dimensions includes: All batches of PVC products are clustered using the trend influence of the corresponding data series under all dimensions to obtain several clusters and the largest cluster center; The standardization of the data sequence of any dimension of any batch of PVC products is obtained, and the degree of influence of environmental conditions in any dimension of the batch of PVC products is calculated based on the standardization and the relative distance between the cluster center of the cluster where the PVC products are located and the maximum cluster center.

5. The method for detecting the performance of heat stabilizers for PVC products according to claim 4, characterized in that: The method of clustering all batches of PVC products by using the trend influence of corresponding data sequences under all dimensions to obtain several clusters and the maximum cluster center includes the following specific methods: Obtain a vector of trend influence degrees of data sequences in all dimensions of any batch of PVC products, record it as the trend influence vector of the corresponding batch of PVC products, cluster all batches of PVC products according to the distance between trend influence vectors of different batches of PVC products and use density clustering algorithm to obtain several clusters; The cluster center corresponding to the cluster with the largest number of data points in all clusters is recorded as the maximum cluster center.

6. The method for detecting the performance of heat stabilizers for PVC products according to claim 4, characterized in that: The specific method for obtaining the standard degree of the data sequence of any dimension of any batch of PVC products includes: Get the first The cosine similarity between the data sequence of the dimension and the standard sequence is recorded as the first The standard degree of the data series in one dimension.

7. The method for detecting the performance of heat stabilizers for PVC products according to claim 4, characterized in that: The method of calculating the degree of influence of environmental conditions on the batch of PVC products in any dimension according to the standard degree and the relative distance between the cluster center of the cluster where the PVC products are located and the maximum cluster center includes: The environmental condition influencing factor of the batch of PVC products in any dimension is calculated according to the standard degree and the relative distance between the cluster center of the cluster where the PVC products are located and the maximum cluster center; The maximum and minimum normalization functions are used to normalize the environmental condition influencing factors to obtain the degree of environmental condition influence.

8. The method for detecting the performance of heat stabilizers for PVC products according to claim 7, characterized in that: The specific method for obtaining the environmental condition influencing factor is: The first The ratio of the distance between the cluster center of the cluster where the PVC products of the batches belong and the largest cluster center to the average distance between the cluster center of all clusters and the largest cluster center is used as the first The relative distance between the cluster center of the cluster where the batches of PVC products are located and the maximum cluster center; According to the relative distance and Batch of PVC products The standard degree of the data sequence of the dimensions is used to obtain the environmental condition influencing factor, the relative distance is positively correlated with the environmental condition influencing factor, and the standard degree is negatively correlated with the environmental condition influencing factor.

Images