A tracing method and system based on regression analysis

Abstract

The present invention discloses a tracing method and system based on regression analysis, which belongs to the field of tracing methods. The method comprises the following steps: Step 1: clarify the waste tracing target, and identify key variables according to the characteristics of the waste and the tracing requirements, wherein the key variables include independent variables and dependent variables that affect the source of the waste; Step 2: data collection and preprocessing, that is, removing duplicate, invalid or abnormal data from the collected data to ensure data quality; Step 3: selecting a regression analysis model, importing the preprocessed data into the selected regression analysis model for analysis, and outputting the analysis results; Step 4: combining the regression analysis results, checking and verifying the current waste source, if the verification passes, proceeding to the next step, if the verification fails, returning to the previous step; Step 5: formulating a waste recycling priority list based on the regression analysis results. The present invention can select a suitable regression analysis model according to the needs, and is easy to use.

Description

Technical Field

[0001] The present invention relates to the field of traceability methods, and in particular to a traceability method and system based on regression analysis. Background Art

[0002] Waste recycling traceability refers to the use of a series of technical means and management measures to track and record the source, collection, transportation, and treatment of waste throughout the process to ensure the legality of the waste and the compliance of the treatment process. Its purpose is to improve the recycling rate of resources, reduce environmental pollution, and protect public health and safety. Waste recycling traceability is a complex and important process that requires the comprehensive use of a variety of technical means and management measures. By continuously improving traceability technology and processes, strengthening supervision and cooperation, we can promote the healthy development of the waste recycling industry and achieve effective recycling and utilization of resources.

[0003] In order to improve the accuracy of existing waste recycling traceability, regression analysis is added to the traceability method. Regression analysis is to describe the relationship between independent variables (explanatory variables, input variables) and dependent variables (response variables, output variables) by establishing a mathematical model. This relationship is usually represented by the function Y=f(X), where X is the input variable that affects the prediction result, Y is the output variable to be predicted, and the function f represents the regression model.

[0004] The existing traceability method has the following problems: it is impossible to select a suitable regression analysis model according to the needs, and it is not convenient to use. Therefore, those skilled in the art provide a traceability method and system based on regression analysis to solve the problems raised in the above background technology. Summary of the invention

[0005] The purpose of the present invention is to provide a tracing method and system based on regression analysis, which can select a suitable regression analysis model according to needs and is easy to use, so as to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A source tracing method based on regression analysis includes the following steps:

[0008] Step 1: Clarify the waste traceability goal and identify key variables based on the characteristics of the waste and traceability requirements. Key variables include independent variables and dependent variables that affect the source of waste.

[0009] Step 2: Data collection and preprocessing, that is, removing duplicate, invalid or abnormal data from the collected data to ensure data quality;

[0010] Step 3: Select a regression analysis model, import the preprocessed data into the selected regression analysis model for analysis, and output the analysis results;

[0011] Step 4: Combined with the regression analysis results, check and verify the current waste source. If the verification passes, proceed to the next step. If the verification fails, return to the previous step.

[0012] Step 5: Develop a waste recycling priority list based on the regression analysis results.

[0013] As a further solution of the present invention: in step 3, the specific process of selecting the regression analysis model is:

[0014] Obtain all regression analysis models from the big data and mark them as Ai, i = 1···n, n is a positive integer;

[0015] Mark all waste types to be traced as Bi, i = 1···k, k is a positive integer;

[0016] Establish a screening analysis model to select a suitable regression analysis model for each type of waste Bi;

[0017] Based on the screened regression analysis models, the single regression analysis model with the best adaptability is selected;

[0018] Analyze the quantity of various types of waste to be traced this time, and determine whether to use a single regression analysis model or multiple regression analysis models based on the analysis results.

[0019] As a further solution of the present invention: the specific process of the screening analysis model to screen out a suitable regression analysis model for each type of waste Bi is:

[0020] Collect the records of Bi waste traceability based on regression analysis in the past, and plan the records using the same regression analysis model into the same cluster. Each cluster is marked as Ci, i=1···n;

[0021] Get the analysis time Ti and accuracy Qi corresponding to each record in cluster Ci, and calculate the comprehensive model performance Vi = (Ti / Qi) of the corresponding record based on Ti and Qi;

[0022] Calculate the average value of the model comprehensive performance Vi corresponding to each record in cluster Ci to obtain the average model comprehensive performance vi corresponding to cluster Ci;

[0023] The average model comprehensive performance vi corresponding to each cluster Ci is compared, the cluster Ci corresponding to the maximum vi is selected, and then the regression analysis model corresponding to the selected cluster Ci is used as the regression analysis model suitable for Bi type waste.

[0024] As a further solution of the present invention: the specific selection process of the regression analysis model with the best adaptability is:

[0025] Obtain the three clusters with the largest vi in ​​the cluster Ci corresponding to the Bi type of waste, and mark the regression analysis models corresponding to the three clusters as a1i, a2i, and a3i, where i=1···k, and k is a positive integer;

[0026] The a1i, a2i, and a3i corresponding to each type of waste are collected together, and the regression analysis model with the most repetitions is selected and marked as G. If G is a single model, then G is output as the regression analysis model with the best adaptability. If G is multiple models, then proceed to the next step.

[0027] Label the multiple models of G as gi, i = 1···n, and obtain the sum of the comprehensive performance of the cluster average models corresponding to each type of waste in gi, labeled as vvi;

[0028] Filter out the maximum value from vvi and mark it as vv 最大 , and then vv 最大 The corresponding model is output as the regression analysis model with the best adaptability.

[0029] As a further solution of the present invention: the specific process of judging whether to select a single regression analysis model or multiple regression analysis models according to the analysis results is:

[0030] The single regression analysis model with the best adaptability is marked as G 单一 ;

[0031] Extract a1i, a2i, a3i corresponding to each type of waste. If there is any one of a1i, a2i, a3i that matches G 单一 If the same regression analysis model is used, this type of waste is planned to the first sequence. If there is no waste in a1i, a2i, and a3i that is related to G 单一 For the same regression analysis model, this type of waste is planned into the second sequence;

[0032] Get the sum of the quantities of each type of waste in the second sequence, marked as D 总 , D 总 With the preset value D 预设 Compare, if D 总 >D 预设 , then multiple regression analysis models are selected, that is, each type of waste is individually selected with an appropriate regression analysis model. 总 ≤D 预设 , a single regression analysis model is selected, that is, each type of waste is uniformly equipped with the single regression analysis model with the best adaptability.

[0033] As a further solution of the present invention: In step 4, the specific process of checking and verifying the current waste source in combination with the regression analysis results is as follows:

[0034] Extract the analyzed waste sources from the analysis results output by the regression analysis model and mark them as X;

[0035] The real source Y and circulation channel Z of waste are collected manually by questionnaire survey and telephone consultation;

[0036] If X and Y are from the same source, the verification result is output as passed. If X and Y are from different sources, proceed to the next step.

[0037] Calculate the accuracy W of the regression analysis model, and then compare W with the preset value w. If W>w, then output the verification result as passed; if W≤w, then output the verification result as failed.

[0038] As a further solution of the present invention: the specific process of calculating the accuracy W of the regression analysis model is:

[0039] Label each node on the flow path Z as zi, i = 1···n, n is a positive integer;

[0040] If X matches any item zi, go to the next step. If X does not match any item zi, output W=0.

[0041] Mark the node of Y on the flow path Z as zy, and then mark the node matched by X on the flow path Z as zx;

[0042] Get the number of all nodes between zy and zx and mark them as R. If R=0, output W=100%. If R≠0, go to the next step.

[0043] If zx is located upstream of zy, then W=100%-(R*5%); if zx is located downstream of zy, then W=100%-(R*8%).

[0044] As a further solution of the present invention: in step 5, the specific process of formulating the waste recycling priority list based on the regression analysis results is:

[0045] The type, quantity, quality, and source of waste were used as independent variables, and the reliability score of the source of waste was used as the dependent variable;

[0046] Use the collected data to fit the determined regression analysis model and optimize the model;

[0047] Output the analysis result of the regression analysis model, that is, the reliability score of the scrap source, and mark the reliability score of each scrap source as Pi, i=1···n, n is a positive integer;

[0048] Sort various waste sources in descending order of Pi to obtain a reliability list. In the reliability list, the higher the ranking of the waste source, the higher its reliability;

[0049] Output the reliability list as a scrap recycling priority list.

[0050] As a further solution of the present invention: the optimization of the model includes checking the model's fit, prediction accuracy, and stability, and avoiding overfitting and underfitting problems through cross-validation and regularization techniques.

[0051] The present application also discloses a traceability system based on regression analysis, which adopts a traceability method based on regression analysis, including:

[0052] The variable identification module is used to clarify the waste traceability target and identify key variables based on the characteristics of the waste and traceability requirements. The key variables include independent variables and dependent variables that affect the source of the waste.

[0053] Data collection and processing module, used to collect data and remove duplicate, invalid or abnormal data from the collected data to ensure data quality;

[0054] The regression analysis module is used to select a regression analysis model, import the preprocessed data into the selected regression analysis model for analysis, and output the analysis results;

[0055] The investigation and verification module is used to investigate and verify the current waste source in combination with the regression analysis results;

[0056] A list generation module is used to develop a waste recycling priority list based on the regression analysis results.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] 1. The present application can select a suitable regression analysis model according to the needs, which is easy to use. Among them, the present application selects a single regression analysis model or multiple regression analysis models according to the needs, and finds a balance between reducing the model calculation pressure and improving the traceability accuracy. It is easy to use and helps to discover the commonalities and differences between different waste products.

[0059] 2. This application investigates and verifies the current waste sources based on the constructed regression analysis model and the regression analysis results. This verification process not only verifies the accuracy of the predicted waste sources, but also can accurately calculate the difference between the predicted results and the actual results, thereby facilitating further optimization and adjustment by the staff. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 It is a flowchart of a traceability method based on regression analysis;

[0061] Figure 2 This is a structural block diagram of a traceability system based on regression analysis. DETAILED DESCRIPTION

[0062] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0063] As mentioned in the background technology of this application, it has been found through research that the existing traceability method cannot select a suitable regression analysis model according to demand, is not convenient to use, and has certain defects.

[0064] In order to solve the above-mentioned defects, the present application discloses a tracing method and system based on regression analysis, which can select a suitable regression analysis model according to needs and is easy to use.

[0065] The following will describe in detail how the solution of the present application solves the above technical problems in conjunction with the accompanying drawings.

[0066] See also Figure 1 In an embodiment of the present invention, a traceability method based on regression analysis includes the following steps: Step 1: clarify the waste traceability target, and identify key variables according to the characteristics of the waste and the traceability requirements, wherein the key variables include independent variables and dependent variables that affect the source of the waste, and the independent variables and dependent variables can be set as needed, such as common independent variables such as production batches, suppliers, transportation routes, etc., and common dependent variables such as scrap rate, waste type, etc.; Step 2: data collection and preprocessing, that is, removing duplicate, invalid or abnormal data from the collected data to ensure data quality; Step 3: select a regression analysis model, import the preprocessed data into the selected regression analysis model for analysis, and output the analysis results; Step 4: combine the regression analysis results to check and verify the current waste source, if the verification passes, proceed to the next step, if the verification fails, return to the previous step; Step 5: formulate a waste recycling priority list based on the regression analysis results. This application can select a suitable regression analysis model according to demand, which is easy to use.

[0067] In this embodiment, in step three, the specific process of selecting a regression analysis model is: obtain all regression analysis models from big data and mark them as Ai, i=1···n, n is a positive integer; mark all types of waste to be traced as Bi, i=1···k, k is a positive integer; establish a screening analysis model to screen out a suitable regression analysis model for each type of waste Bi; based on the screened regression analysis model, select the single regression analysis model with the best adaptability; analyze the number of various types of waste to be traced this time, and determine whether to use a single regression analysis model or multiple regression analysis models based on the analysis results. This application selects a single regression analysis model or multiple regression analysis models as needed, finds a balance between reducing the model calculation pressure and improving the traceability accuracy, and is easy to use while helping to discover the commonalities and differences between different wastes.

[0068] In this embodiment, the specific process of selecting the appropriate regression analysis model for each type of waste Bi is as follows: collecting the records of Bi type waste traceability based on regression analysis in the past, and planning the records using the same regression analysis model into the same cluster, each cluster is marked as Ci, i=1···n; obtaining the analysis time Ti and accuracy Qi corresponding to each record in cluster Ci, and calculating the model comprehensive performance Vi=(Ti / Qi) of the corresponding record based on Ti and Qi; calculating the average value of the model comprehensive performance Vi corresponding to each record in cluster Ci to obtain the average model comprehensive performance vi corresponding to cluster Ci; comparing the average model comprehensive performance vi corresponding to each cluster Ci, selecting the cluster Ci corresponding to the maximum vi, and then using the regression analysis model corresponding to the selected cluster Ci as the regression analysis model suitable for Bi type waste. This setting can quickly determine the most suitable regression analysis model for each type of waste, laying a foundation for the subsequent selection of multiple regression analysis models.

[0069] In this embodiment, the specific selection process of the regression analysis model with the best adaptability is as follows: obtain the three clusters with the largest vi in ​​the cluster Ci corresponding to the Bi type of waste, and mark the regression analysis models corresponding to the three clusters as a1i, a2i, and a3i, where i=1···k, and k is a positive integer; collect a1i, a2i, and a3i corresponding to each type of waste together, and select the regression analysis model with the largest number of repetitions, and mark it as G; if G is a single model, output G as the regression analysis model with the best adaptability; if G is multiple models, proceed to the next step; mark the multiple models of G as gi, i=1···n, obtain the sum of the comprehensive performance of the cluster average models corresponding to each type of waste of gi, and mark it as vvi; select the maximum value from vvi and mark it as vv 最大 , and then vv 最大The corresponding model is output as the most adaptable regression analysis model. This setting can quickly determine the most suitable regression analysis model for all types of waste, laying the foundation for the subsequent selection of a single regression analysis model.

[0070] In this embodiment, the specific process of determining whether to use a single regression analysis model or multiple regression analysis models according to the analysis results is as follows: marking the selected single regression analysis model with the best adaptability as G 单一 ; Extract a1i, a2i, a3i corresponding to each type of waste. If there is any one of a1i, a2i, a3i that matches G 单一 If the same regression analysis model is used, this type of waste is planned to the first sequence. If there is no waste in a1i, a2i, and a3i that is related to G 单一 If the model is the same as the regression analysis model, the waste of this type is planned into the second sequence; the sum of the quantity of each type of waste in the second sequence is obtained, which is marked as D 总 , D 总 With the preset value D 预设 Compare, if D 总 >D 预设 , then multiple regression analysis models are selected, that is, each type of waste is individually selected with an appropriate regression analysis model. 总 ≤D 预设 , a single regression analysis model is selected, that is, each type of waste product uniformly adopts a single regression analysis model with the best adaptability. According to the needs, this application selects a single regression analysis model under the premise that the difference between the waste type and weight is small. In this way, the same model can be selected to reduce the calculation pressure under the premise of ensuring the accuracy of the model analysis results. In addition, the same model can also be easily compared and evaluated, which is helpful to discover the commonalities and differences between different waste products. When the difference between the waste type and weight is large, this application adopts multiple regression analysis models. Although it increases the calculation pressure of the model, it ensures the accuracy of the model analysis results. Constructing a special regression analysis model for each type of waste can more accurately capture its traceability information. In addition, multiple models can verify and complement each other to improve the accuracy and reliability of traceability.

[0071] In this embodiment, in step 4, the specific process of checking and verifying the current waste source in combination with the regression analysis results is as follows: extract the analyzed waste source from the analysis results output by the regression analysis model, marked as X; manually collect the real source Y and circulation path Z of the waste by questionnaire survey and telephone consultation; if X and Y are from the same source, output the verification result as passed; if X and Y are from different sources, proceed to the next step; calculate the accuracy W of the regression analysis model, and then compare W with the preset value w. If W>w, output the verification result as passed; if W≤w, output the verification result as failed. This setting can effectively ensure that the constructed regression analysis model has sufficient accuracy.

[0072] In this embodiment, the specific process of calculating the accuracy W of the regression analysis model is: mark each node on the circulation path Z as zi, i=1···n, n is a positive integer; if X matches any item zi, then proceed to the next step, if X does not match any item zi, then output W=0; mark the node of Y on the circulation path Z as zy, and then mark the node matched by X on the circulation path Z as zx; obtain the number of all nodes between zy and zx and mark them as R, if R=0, then output W=100%, if R≠0, then proceed to the next step; if zx is located upstream of zy, then W=100%-(R*5%), if zx is located downstream of zy, then W=100%-(R*8%). This application investigates and verifies the current waste source based on the constructed regression analysis model and the regression analysis results. The verification process not only verifies the accuracy of the predicted waste source, but also accurately calculates the difference between the predicted result and the actual result, thereby facilitating further optimization and adjustment by the staff.

[0073] In this embodiment, in step five, the specific process of formulating a waste recycling priority list based on the regression analysis results is: taking the type, quantity, quality, and source of waste as independent variables, and then taking the reliability score of the waste source as the dependent variable; using the collected data to fit the determined regression analysis model, and optimizing the model; outputting the analysis result of the regression analysis model, that is, the reliability score of the waste source, marking the reliability score of each type of waste source as Pi, i=1···n, n is a positive integer; sorting the various types of waste sources in order from large to small according to Pi, to obtain a reliability list, in which the higher the ranking of the waste source is, the higher its reliability is; outputting the reliability list as a waste recycling priority list. This setting can quickly determine the reliability scores of various waste sources, and then sort them to obtain a waste recycling priority list, thereby facilitating waste recycling parties to formulate more efficient recycling strategies according to the priority level.

[0074] In this embodiment, optimizing the model includes checking the fit, prediction accuracy, stability of the model, and avoiding overfitting and underfitting problems through cross-validation and regularization techniques. Cross-validation is a model evaluation technique widely used in the fields of statistical analysis and machine learning. In particular, when establishing regression analysis models or other prediction models, cross-validation techniques are of great significance for verifying the performance of the model and avoiding overfitting and underfitting problems. Cross-validation is a method of verifying the performance of the model by dividing the data set into multiple subsets and conducting repeated experiments. In each experiment, a subset is used as a training set to train the model, and the performance of the model is tested on the remaining data subsets. In this way, the generalization ability of the model on different data sets can be evaluated, and the best model parameters can be found to obtain the best prediction performance. Regularization is a technique in the field of machine learning and statistical modeling to prevent model overfitting and enhance the generalization ability of the model. Regularization limits the complexity of the model and manages the degrees of freedom of the model parameters by introducing additional constraints or penalty terms in the loss function of the model. Its purpose is to control the complexity of the model and avoid overfitting the model on the training data, thereby improving the generalization ability of the model on unseen data.

[0075] See also Figure 2 The present application also discloses a traceability system based on regression analysis, which adopts a traceability method based on regression analysis, including: a variable identification module, which is used to clarify the traceability target of waste products, and identify key variables according to the characteristics of the waste products and the traceability requirements, wherein the key variables include independent variables and dependent variables that affect the source of waste products; a data collection and processing module, which is used to collect data and remove duplicate, invalid or abnormal data from the collected data to ensure data quality; a regression analysis module, which is used to select a regression analysis model, import the pre-processed data into the selected regression analysis model for analysis, and output the analysis results; a screening and verification module, which is used to screen and verify the current waste source in combination with the regression analysis results; a list generation module, which is used to formulate a waste recycling priority list based on the regression analysis results.

[0076] The present invention can select a suitable regression analysis model according to the needs, which is easy to use. Among them, the present application selects a single regression analysis model or multiple regression analysis models according to the needs, finds a balance between reducing the model calculation pressure and improving the traceability accuracy, and is easy to use while helping to discover the commonalities and differences between different waste products. The present application investigates and verifies the current waste source based on the constructed regression analysis model and the regression analysis results. The verification process not only verifies the accuracy of the predicted waste source, but also can accurately calculate the difference between the predicted results and the actual results, thereby facilitating further optimization and adjustment by the staff.

[0077] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

[0078] What is described above is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A tracing method based on regression analysis, characterized in that: The following steps are involved: Step 1: Clarify the waste traceability goal and identify key variables based on the characteristics of the waste and traceability requirements. Key variables include independent variables and dependent variables that affect the source of waste. Step 2: Data collection and preprocessing, that is, removing duplicate, invalid or abnormal data from the collected data to ensure data quality; Step 3: Select a regression analysis model, import the preprocessed data into the selected regression analysis model for analysis, and output the analysis results; Step 4: Combined with the regression analysis results, check and verify the current waste source. If the verification passes, proceed to the next step. If the verification fails, return to the previous step. Step 5: Develop a waste recycling priority list based on the regression analysis results; In step 3, the specific process of selecting the regression analysis model is as follows: Obtain all regression analysis models from the big data and mark them as Ai, i = 1···n, n is a positive integer; Mark all waste types to be traced as Bi, i = 1···k, k is a positive integer; Establish a screening analysis model to select a suitable regression analysis model for each type of waste Bi; Based on the screened regression analysis models, the single regression analysis model with the best adaptability is selected; Analyze the quantity of various types of waste to be traced, and select a single regression analysis model or multiple regression analysis models based on the analysis results; The specific process of the screening analysis model to screen out a suitable regression analysis model for each type of waste Bi is as follows: Collect the records of Bi waste traceability based on regression analysis in the past, and plan the records using the same regression analysis model into the same cluster. Each cluster is marked as Ci, i=1···n; Get the analysis time Ti and accuracy Qi corresponding to each record in cluster Ci, and calculate the comprehensive model performance Vi = (Ti / Qi) of the corresponding record based on Ti and Qi; Calculate the average value of the model comprehensive performance Vi corresponding to each record in cluster Ci to obtain the average model comprehensive performance vi corresponding to cluster Ci; The average model comprehensive performance vi corresponding to each cluster Ci is compared, the cluster Ci corresponding to the maximum vi is selected, and then the regression analysis model corresponding to the selected cluster Ci is used as the regression analysis model suitable for Bi type waste; The specific selection process of the regression analysis model with the best adaptability is: Obtain the three clusters with the largest vi in ​​the cluster Ci corresponding to the Bi type of waste, and mark the regression analysis models corresponding to the three clusters as a1i, a2i, and a3i, where i=1···k, and k is a positive integer; The a1i, a2i, and a3i corresponding to each type of waste are collected together, and the regression analysis model with the most repetitions is selected and marked as G. If G is a single model, then G is output as the regression analysis model with the best adaptability. If G is multiple models, then proceed to the next step. Label the multiple models of G as gi, i = 1···n, and obtain the sum of the comprehensive performance of the cluster average models corresponding to each type of waste in gi, labeled as vvi; Filter out the maximum value from vvi and mark it as vv 最大 , and then vv 最大 The corresponding model is output as the regression analysis model with the best adaptability.

2. The source tracing method based on regression analysis according to claim 1 is characterized in that: The specific process of determining whether to use a single regression analysis model or multiple regression analysis models based on the analysis results is as follows: The single regression analysis model with the best adaptability is marked as G 单一 ; Extract a1i, a2i, a3i corresponding to each type of waste. If there is any one of a1i, a2i, a3i that matches G 单一 If the same regression analysis model is used, this type of waste is planned to the first sequence. If there is no waste in a1i, a2i, and a3i that is related to G 单一 For the same regression analysis model, this type of waste is planned into the second sequence; Get the sum of the quantities of each type of waste in the second sequence, marked as D 总 , D 总 With the preset value D 预设 Compare, if D 总 >D 预设 , then multiple regression analysis models are selected, that is, each type of waste is individually selected with an appropriate regression analysis model. 总 ≤D 预设 , a single regression analysis model is selected, that is, each type of waste is uniformly equipped with the single regression analysis model with the best adaptability.

3. The source tracing method based on regression analysis according to claim 2 is characterized in that: In step 4, the specific process of checking and verifying the current waste source in combination with the regression analysis results is as follows: Extract the analyzed waste sources from the analysis results output by the regression analysis model and mark them as X; The real source Y and circulation channel Z of waste are collected manually by questionnaire survey and telephone consultation; If X and Y are from the same source, the verification result is output as passed. If X and Y are from different sources, proceed to the next step. Calculate the accuracy W of the regression analysis model, and then compare W with the preset value w. If W>w, then output the verification result as passed; if W≤w, then output the verification result as failed.

4. The source tracing method based on regression analysis according to claim 3 is characterized in that: The specific process of calculating the accuracy W of the regression analysis model is: Label each node on the flow path Z as zi, i = 1···n, n is a positive integer; If X matches any item zi, go to the next step. If X does not match any item zi, output W=0. Mark the node of Y on the flow path Z as zy, and then mark the node matched by X on the flow path Z as zx; Get the number of all nodes between zy and zx and mark them as R. If R=0, output W=100%. If R≠0, go to the next step. If zx is located upstream of zy, then W=100%-(R*5%); if zx is located downstream of zy, then W=100%-(R*8%).

5. The method for tracing the source based on regression analysis according to claim 4 is characterized in that: In step 5, the specific process of formulating a waste recycling priority list based on the regression analysis results is as follows: The type, quantity, quality, and source of waste were used as independent variables, and the reliability score of the source of waste was used as the dependent variable; Use the collected data to fit the determined regression analysis model and optimize the model; Output the analysis result of the regression analysis model, that is, the reliability score of the scrap source, and mark the reliability score of each scrap source as Pi, i=1···n, n is a positive integer; Sort various waste sources in descending order of Pi to obtain a reliability list. In the reliability list, the higher the ranking of the waste source, the higher its reliability; Output the reliability list as a scrap recycling priority list.

6. The method for tracing the source based on regression analysis according to claim 5, characterized in that: The optimization of the model includes checking the model's fit, prediction accuracy, and stability, as well as avoiding overfitting and underfitting problems through cross-validation and regularization techniques.

7. A traceability system based on regression analysis, using the traceability method based on regression analysis as claimed in any one of claims 1 to 6, characterized in that: include: The variable identification module is used to clarify the waste traceability target and identify key variables based on the characteristics of the waste and traceability requirements. The key variables include independent variables and dependent variables that affect the source of the waste. Data collection and processing module, used to collect data and remove duplicate, invalid or abnormal data from the collected data to ensure data quality; The regression analysis module is used to select a regression analysis model, import the preprocessed data into the selected regression analysis model for analysis, and output the analysis results; The investigation and verification module is used to investigate and verify the current waste source in combination with the regression analysis results; A list generation module is used to develop a waste recycling priority list based on the regression analysis results.

Images