A quality traceability method, system and medium based on new energy precursors

Abstract

The present invention provides a quality traceability method, system and medium based on new energy precursors, the method comprising: real-time inspection to obtain the quality inspection results of the prepared new energy precursors, and judge whether the quality inspection results of the new energy precursors are qualified; if not, extract the bad attribute data that causes the unqualified quality inspection results of the new energy precursors; based on the bad attribute data and through a preset traceability detection algorithm, infer the abnormal node parameters that cause the bad attributes; based on the abnormal node parameters and combined with the on-site monitoring data, analyze the corresponding abnormal causes, and formulate corresponding recovery solutions according to the abnormal causes; judge whether the abnormal causes are caused by human factors, if so, start the quality accident accountability procedure and hold the relevant responsible persons accountable. The present invention can timely detect the occurrence of quality problems in the production process of new energy precursors and eliminate the unfavorable factors for the occurrence of quality problems.

Description

Technical Field

[0001] The present invention relates to the technical field of product quality traceability, and in particular to a quality traceability method, system and medium based on new energy precursors. Background Art

[0002] New energy precursor refers to an intermediate material used to prepare new energy positive electrode materials. Since the final performance of the positive electrode material will inherit the morphological and structural characteristics of its precursor, the quality of the precursor directly determines the physical and chemical indicators of the positive electrode sintering product, and is a key raw material for the preparation of new energy positive electrode materials. It is mainly made of nickel salts, cobalt salts, manganese salts or nickel salts, cobalt salts, and aluminum salts. According to the different composition ratios of elements, it can be divided into NCM333, NCM424, NCM442, NCM523, NCM622, NCM811, NCA and other types. The synthesis of the precursor requires the control of the conditions and parameters of the chemical reaction to obtain a high-purity and uniform precursor material. This involves technical requirements for the selection of raw materials, control of the reaction process, and separation and purification of the products.

[0003] The structure and morphology of new energy precursors have an important impact on battery performance, and the quality stability of new energy precursors is directly related to the performance and life of the battery. To ensure the batch consistency and stability of precursor materials, a strict quality control system and testing methods are required.

[0004] With the continuous development of battery technology and the development of the production of new energy precursors, the scale of production equipment is increasing, the system complexity is getting higher and higher, the precision requirements are getting more and more stringent, the level of intelligence is getting higher and higher, and the quality and precision requirements of new energy precursors are also getting higher and higher. Improving the intelligence level of the production line and continuously, stably and efficiently producing products with a high pass rate have become important bargaining chips for enterprises to be competitive in the market and in future development.

[0005] Therefore, in the production process of new energy precursors, it is necessary to detect the occurrence of quality problems in a timely manner, trace the causes of quality problems and control the factors that cause them, and eliminate the adverse factors that cause quality problems, so that the company's production process can be carried out continuously and stably for a long time, thereby improving the company's intelligence level and market competitiveness. Summary of the invention

[0006] In order to solve at least one of the above technical problems, the present invention proposes a quality traceability method, system and medium based on new energy precursors, which can timely detect the occurrence of quality problems in the production process of new energy precursors, trace the causes of quality problems and control the generating factors, eliminate the adverse factors of quality problems, so that the production process of the enterprise can be carried out continuously and stably for a long time, thereby improving the intelligence level and market competitiveness of the enterprise.

[0007] The first aspect of the present invention proposes a quality traceability method based on new energy precursors, the method comprising:

[0008] A quality traceability method based on new energy precursors, the method comprising:

[0009] Real-time inspection to obtain the quality inspection results of the prepared new energy precursors, and determine whether the quality inspection results of the new energy precursors are qualified;

[0010] If not, extract the bad attribute data that causes the new energy precursor quality inspection result to be unqualified;

[0011] Based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters that cause the bad attributes are inferred;

[0012] Analyze the corresponding abnormal causes based on abnormal node parameters and on-site monitoring data, and formulate corresponding recovery solutions based on the abnormal causes;

[0013] Determine whether the cause of the abnormality is due to human factors. If so, initiate the quality accident accountability procedure and hold the relevant responsible persons accountable.

[0014] In this solution, based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters that cause the bad attributes are inferred, including:

[0015] The production process of the preset new energy precursor includes n layers of reaction nodes, and the product obtained from the i+1th layer of reaction nodes is used as the raw material of the ith layer of reaction nodes, and the number of the first layer of reaction nodes is 1, and the number of other layer reaction nodes is greater than or equal to 1, where i is greater than or equal to 1 and less than or equal to n-1;

[0016] Build corresponding node prediction models based on each reaction node;

[0017] Inputting the bad attribute data into the node prediction model of the first layer reaction node, and outputting multiple groups of predicted reaction parameters causing the bad attribute data, each group of predicted reaction parameters including one or two of the bad parameter data of the reaction raw material and the bad parameter data of the reaction condition;

[0018] Obtaining the reaction raw material monitoring data and reaction condition monitoring data at the first layer reaction node at the corresponding historical time point;

[0019] Based on multiple groups of predicted reaction parameters, a group of predicted reaction parameters that are completely consistent with the reaction raw material monitoring data and reaction condition monitoring data at the first-layer reaction node are screened out, and the reaction condition bad parameter data contained in the predicted reaction parameters are listed as abnormal node parameters;

[0020] Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step;

[0021] Determine whether the reaction raw material is generated by the corresponding reaction node of the second layer. If not, list the bad parameter data of the reaction raw material as abnormal node parameters. If yes, proceed to the next step.

[0022] Inputting the reaction raw material bad parameter data into the node prediction model of the corresponding reaction node of the second layer, and outputting multiple groups of predicted reaction parameters causing the reaction raw material bad parameter data;

[0023] Acquire reaction raw material monitoring data and reaction condition monitoring data at the corresponding reaction node in the second layer at the corresponding historical time point;

[0024] Based on the multiple groups of predicted reaction parameters of the reaction raw material bad parameter data, a group of predicted reaction parameters that completely match the reaction raw material monitoring data and the reaction condition monitoring data at the corresponding reaction node of the second layer is screened out, and the reaction condition bad parameter data included in the predicted reaction parameters is listed as abnormal node parameters;

[0025] Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step;

[0026] Determine whether the reaction raw material is generated by the corresponding reaction node of the third layer. If not, list the reaction raw material bad parameter data as abnormal node parameters. If yes, continue to trace the reaction raw material bad parameter data to the upper layer until the traced reaction raw material is not generated by the corresponding reaction node of the upper layer.

[0027] In this solution, after inputting the bad attribute data into the node prediction model of the first-layer reaction node and outputting multiple groups of predicted reaction parameters causing the bad attribute data, the method further includes:

[0028] Acquire multiple historical record data of the first-layer reaction node, each historical record data at least including: historical bad attribute data, and historical actual reaction raw material data and historical actual reaction condition data causing the historical bad attribute data;

[0029] Based on multiple historical record data, extract the historical bad attribute data therein and calculate the extracted data according to a preset processing algorithm to obtain a historical characterization value;

[0030] Extract the bad property data of the current new energy precursor and calculate the extracted data using the same processing algorithm to obtain the current characterization value;

[0031] Based on multiple historical record data, similarity calculation is performed between the historical characterization value corresponding to each historical record data and the current characterization value corresponding to the bad attribute data of the current new energy precursor;

[0032] Adding historical record data with a similarity greater than a first threshold into a revised sample database;

[0033] Based on each historical record data in the modified sample database, the historical bad attribute data is input into the node prediction model, and the historical predicted reaction raw material data and the historical predicted reaction condition data causing the historical bad attribute data are output;

[0034] Based on each historical record data in the correction sample database, the historical actual reaction raw material data and the historical actual reaction condition data are respectively subtracted from the corresponding historical predicted reaction raw material data and the historical predicted reaction condition data to obtain the difference of the reaction raw material data and the difference of the reaction condition data;

[0035] Calculate the average values ​​of the difference values ​​of the reaction raw material data and the difference values ​​of the reaction condition data of all the historical record data in the correction sample database to obtain the correction value of the reaction raw material data and the correction value of the reaction condition data;

[0036] Based on the correction value of the reaction raw material data, the reaction raw material bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction raw material bad parameter data; based on the correction value of the reaction condition data, the reaction condition bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction condition bad parameter data.

[0037] In this scheme, data extraction is performed on the historical bad attribute data and the extracted data is calculated according to a preset processing algorithm to obtain a historical characterization value; data extraction is performed on the bad attribute data of the current new energy precursor and the extracted data is calculated according to the same processing algorithm to obtain a current characterization value, specifically including:

[0038] Extract historical percentage values ​​of K compounds from historical adverse attribute data;

[0039] Sort the K compounds in a preset order, use the serial number of each compound as the horizontal axis value, and use the historical percentage value of each compound as the vertical axis value, and mark the coordinate points corresponding to the historical percentage values ​​of the compounds in K in the two-dimensional coordinate system;

[0040] The coordinate points corresponding to the historical percentage values ​​of the compounds in K are smoothly connected to obtain a historical representation line for representing the historical representation value;

[0041] Extract the current percentage values ​​of K compounds from the current adverse property data of new energy precursors;

[0042] Arrange the K compounds in the same order, use the serial number of each compound as the horizontal coordinate value, and use the current percentage value of each compound as the vertical coordinate value, and mark the coordinate points corresponding to the current percentage values ​​of the compounds in K in the two-dimensional coordinate system;

[0043] The coordinate points corresponding to the current percentage values ​​of the compounds in K are smoothly connected to obtain a current characterization line for representing the current characterization value.

[0044] In this solution, the historical characterization value corresponding to each historical record data is respectively calculated with the current characterization value corresponding to the bad attribute data of the current new energy precursor for similarity, specifically including:

[0045] The first serial number value of the compound in K is obtained as H, and the last serial number value is calculated as H+K;

[0046] Draw the straight lines x=H and x=H+K in the two-dimensional coordinate system;

[0047] Intersect the historical representation line, the horizontal coordinate, the straight line x=H, and the straight line x=H+K to enclose the historical area;

[0048] The current characterization line, the horizontal coordinate, the straight line x=H, and the straight line x=H+K are intersected in pairs to enclose the current region, and the area of ​​the current region is calculated;

[0049] Draw the overlapping area between the historical area and the current area, and calculate the area of ​​the overlapping area;

[0050] Divide the area of ​​the overlapping region by the area of ​​the current region to get the similarity between the two.

[0051] In this solution, the corresponding abnormal cause is analyzed based on the abnormal node parameters and combined with the on-site monitoring data, and the corresponding recovery solution is formulated according to the abnormal cause, including:

[0052] Based on the abnormal node parameters and the association relationship, multiple possible cause locations associated with it are found;

[0053] The probability of abnormal node parameters caused by each possible cause position pair is different, and the probability of abnormal node parameters caused by each possible cause position pair is calculated based on historical data statistics;

[0054] Obtain monitoring data for each possible cause location, and perform anomaly analysis on the monitoring data based on the attention mechanism according to the probability of occurrence of each possible cause location;

[0055] Focus on analyzing the monitoring data to find out the abnormal reasons that lead to abnormal node parameters, and formulate corresponding recovery solutions based on the abnormal reasons.

[0056] The second aspect of the present invention further proposes a quality traceability system based on a new energy precursor, comprising a memory and a processor, wherein the memory comprises a quality traceability method program based on a new energy precursor, and when the quality traceability method program based on a new energy precursor is executed by the processor, the following steps are implemented:

[0057] Real-time inspection to obtain the quality inspection results of the prepared new energy precursors, and determine whether the quality inspection results of the new energy precursors are qualified;

[0058] If not, extract the bad attribute data that causes the new energy precursor quality inspection result to be unqualified;

[0059] Based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters that cause the bad attributes are inferred;

[0060] Analyze the corresponding abnormal causes based on abnormal node parameters and on-site monitoring data, and formulate corresponding recovery solutions based on the abnormal causes;

[0061] Determine whether the cause of the abnormality is due to human factors. If so, initiate the quality accident accountability procedure and hold the relevant responsible persons accountable.

[0062] In this solution, based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters that cause the bad attributes are inferred, including:

[0063] The production process of the preset new energy precursor includes n layers of reaction nodes, and the product obtained from the i+1th layer of reaction nodes is used as the raw material of the ith layer of reaction nodes, and the number of the first layer of reaction nodes is 1, and the number of other layer reaction nodes is greater than or equal to 1, where i is greater than or equal to 1 and less than or equal to n-1;

[0064] Build corresponding node prediction models based on each reaction node;

[0065] Inputting the bad attribute data into the node prediction model of the first layer reaction node, and outputting multiple groups of predicted reaction parameters causing the bad attribute data, each group of predicted reaction parameters including one or two of the bad parameter data of the reaction raw material and the bad parameter data of the reaction condition;

[0066] Obtaining the reaction raw material monitoring data and reaction condition monitoring data at the first layer reaction node at the corresponding historical time point;

[0067] Based on multiple groups of predicted reaction parameters, a group of predicted reaction parameters that are completely consistent with the reaction raw material monitoring data and reaction condition monitoring data at the first-layer reaction node are screened out, and the reaction condition bad parameter data contained in the predicted reaction parameters are listed as abnormal node parameters;

[0068] Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step;

[0069] Determine whether the reaction raw material is generated by the corresponding reaction node of the second layer. If not, list the bad parameter data of the reaction raw material as abnormal node parameters. If yes, proceed to the next step.

[0070] Inputting the reaction raw material bad parameter data into the node prediction model of the corresponding reaction node of the second layer, and outputting multiple groups of predicted reaction parameters causing the reaction raw material bad parameter data;

[0071] Acquire reaction raw material monitoring data and reaction condition monitoring data at the corresponding reaction node in the second layer at the corresponding historical time point;

[0072] Based on the multiple groups of predicted reaction parameters of the reaction raw material bad parameter data, a group of predicted reaction parameters that completely match the reaction raw material monitoring data and the reaction condition monitoring data at the corresponding reaction node of the second layer is screened out, and the reaction condition bad parameter data included in the predicted reaction parameters is listed as abnormal node parameters;

[0073] Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step;

[0074] Determine whether the reaction raw material is generated by the corresponding reaction node of the third layer. If not, list the reaction raw material bad parameter data as abnormal node parameters. If yes, continue to trace the reaction raw material bad parameter data to the upper layer until the traced reaction raw material is not generated by the corresponding reaction node of the upper layer.

[0075] In this solution, after inputting the bad attribute data into the node prediction model of the first layer reaction node and outputting multiple groups of predicted reaction parameters causing the bad attribute data, the quality traceability method program based on the new energy precursor is further implemented when executed by the processor:

[0076] Acquire multiple historical record data of the first-layer reaction node, each historical record data at least including: historical bad attribute data, and historical actual reaction raw material data and historical actual reaction condition data causing the historical bad attribute data;

[0077] Based on multiple historical record data, extract the historical bad attribute data therein and calculate the extracted data according to a preset processing algorithm to obtain a historical characterization value;

[0078] Extract the bad property data of the current new energy precursor and calculate the extracted data using the same processing algorithm to obtain the current characterization value;

[0079] Based on multiple historical record data, similarity calculation is performed between the historical characterization value corresponding to each historical record data and the current characterization value corresponding to the bad attribute data of the current new energy precursor;

[0080] Adding historical record data with a similarity greater than a first threshold into a revised sample database;

[0081] Based on each historical record data in the modified sample database, the historical bad attribute data is input into the node prediction model, and the historical predicted reaction raw material data and the historical predicted reaction condition data causing the historical bad attribute data are output;

[0082] Based on each historical record data in the correction sample database, the historical actual reaction raw material data and the historical actual reaction condition data are respectively subtracted from the corresponding historical predicted reaction raw material data and the historical predicted reaction condition data to obtain the difference of the reaction raw material data and the difference of the reaction condition data;

[0083] Calculate the average values ​​of the difference values ​​of the reaction raw material data and the difference values ​​of the reaction condition data of all the historical record data in the correction sample database to obtain the correction value of the reaction raw material data and the correction value of the reaction condition data;

[0084] Based on the correction value of the reaction raw material data, the reaction raw material bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction raw material bad parameter data; based on the correction value of the reaction condition data, the reaction condition bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction condition bad parameter data.

[0085] The third aspect of the present invention also proposes a computer-readable storage medium, which includes a quality traceability method program based on a new energy precursor. When the quality traceability method program based on a new energy precursor is executed by a processor, the steps of a quality traceability method based on a new energy precursor as described above are implemented.

[0086] The present invention proposes a quality traceability method, system and medium based on new energy precursors. The abnormal node parameters that cause poor properties are inferred through a preset traceability detection algorithm. In the production process of new energy precursors, the occurrence of quality problems can be detected in time, the causes of quality problems can be traced and the generating factors can be controlled, and the adverse factors causing the occurrence of quality problems can be eliminated, so that the production process of new energy precursors can be carried out continuously and stably for a long time, thereby improving the enterprise's intelligence level and market competitiveness.

[0087] Additional aspects and advantages of the present invention will be given in part in the following description, and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 A flow chart showing a quality traceability method based on new energy precursors of the present invention is shown;

[0089] Figure 2 A flowchart of abnormal cause analysis according to an embodiment of the present invention is shown;

[0090] Figure 3 A block diagram of a quality traceability system based on new energy precursors of the present invention is shown. DETAILED DESCRIPTION

[0091] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0092] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0093] Figure 1 A flow chart of a quality traceability method based on new energy precursors of the present invention is shown.

[0094] like Figure 1 As shown, the first aspect of the present invention proposes a quality traceability method based on new energy precursors, the method comprising:

[0095] S102, obtaining the quality inspection result of the prepared new energy precursor in real time, and judging whether the quality inspection result of the new energy precursor is qualified;

[0096] S104, if not, extracting the bad attribute data that caused the new energy precursor quality inspection result to be unqualified;

[0097] S106, inferring abnormal node parameters causing the bad attributes based on the bad attribute data and using a preset traceability detection algorithm;

[0098] S108, analyzing the corresponding abnormal cause based on the abnormal node parameters and combined with the on-site monitoring data, and formulating a corresponding recovery solution according to the abnormal cause;

[0099] S110, determine whether the abnormality is caused by human factors. If so, initiate the quality accident accountability procedure and hold the relevant responsible persons accountable.

[0100] It should be noted that in the production and preparation process of new energy precursors, multiple continuous process nodes may be required, such as chemical reaction nodes, filtration and impurity removal nodes, washing nodes, drying nodes, etc. The changes in the node parameters of each process node will affect the quality of the final new energy precursor. Moreover, each process node may have more than one node parameter. For example, for a chemical reaction node, the node parameters at least include the raw material parameters of each chemical reaction and the various condition parameters required for the reaction. Therefore, any error in any node parameter will directly affect the quality of the product of the chemical reaction.

[0101] The present invention infers abnormal node parameters that cause poor properties through a preset traceability detection algorithm. During the production process of new energy precursors, the occurrence of quality problems can be detected in a timely manner, the causes of the quality problems can be traced and the generating factors can be controlled, and the adverse factors causing the quality problems can be eliminated, thereby enabling the production process of new energy precursors to be sustained and carried out stably for a long time, thereby improving the enterprise's intelligence level and market competitiveness.

[0102] The present invention can timely detect the occurrence of quality problems, trace the causes of quality problems and control the generating factors, while further reducing the output of waste of new energy precursors, reducing the input of raw material costs, and thus saving the production costs of new energy precursors.

[0103] According to an embodiment of the present invention, based on the bad attribute data and through a preset source tracing detection algorithm, the abnormal node parameters causing the bad attribute are inferred, specifically including:

[0104] The production process of the preset new energy precursor includes n layers of reaction nodes, and the product obtained from the i+1th layer of reaction nodes is used as the raw material of the ith layer of reaction nodes, and the number of the first layer of reaction nodes is 1, and the number of other layer reaction nodes is greater than or equal to 1, where i is greater than or equal to 1 and less than or equal to n-1;

[0105] Build corresponding node prediction models based on each reaction node;

[0106] Inputting the bad attribute data into the node prediction model of the first layer reaction node, and outputting multiple groups of predicted reaction parameters causing the bad attribute data, each group of predicted reaction parameters including one or two of the bad parameter data of the reaction raw material and the bad parameter data of the reaction condition;

[0107] Obtaining the reaction raw material monitoring data and reaction condition monitoring data at the first layer reaction node at the corresponding historical time point;

[0108] Based on multiple groups of predicted reaction parameters, a group of predicted reaction parameters that are completely consistent with the reaction raw material monitoring data and reaction condition monitoring data at the first-layer reaction node are screened out, and the reaction condition bad parameter data contained in the predicted reaction parameters are listed as abnormal node parameters;

[0109] Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step;

[0110] Determine whether the reaction raw material is generated by the corresponding reaction node of the second layer. If not, list the bad parameter data of the reaction raw material as abnormal node parameters. If yes, proceed to the next step.

[0111] Inputting the reaction raw material bad parameter data into the node prediction model of the corresponding reaction node of the second layer, and outputting multiple groups of predicted reaction parameters causing the reaction raw material bad parameter data;

[0112] Acquire reaction raw material monitoring data and reaction condition monitoring data at the corresponding reaction node in the second layer at the corresponding historical time point;

[0113] Based on the multiple groups of predicted reaction parameters of the reaction raw material bad parameter data, a group of predicted reaction parameters that completely match the reaction raw material monitoring data and the reaction condition monitoring data at the corresponding reaction node of the second layer is screened out, and the reaction condition bad parameter data included in the predicted reaction parameters is listed as abnormal node parameters;

[0114] Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step;

[0115] Determine whether the reaction raw material is generated by the corresponding reaction node of the third layer. If not, list the reaction raw material bad parameter data as abnormal node parameters. If yes, continue to trace the reaction raw material bad parameter data to the upper layer until the traced reaction raw material is not generated by the corresponding reaction node of the upper layer.

[0116] It should be noted that the generation process of the new energy precursor includes multiple chemical reaction nodes, and multiple chemical reaction nodes are connected through a tree diagram. For example, if the new energy precursor is A, raw materials B1, B2, and B3 need to be prepared by chemical reaction, and for raw material B1, raw materials C1, C2, and C3 need to be prepared by chemical reaction; and for raw material B2, raw materials C4, C5, and C6 need to be prepared by chemical reaction, and the factors affecting each reaction node are not only the raw material parameters, but also the various reaction conditions. Therefore, when the parameter data of the product is poor, there may be a variety of reasons, such as the combination of poor reaction raw material 1 and poor reaction condition 1, the combination of poor reaction raw material 1 and poor reaction condition 3, or poor reaction condition 2 independently causes. The present invention constructs a corresponding node prediction model according to the situation of different nodes, and preliminarily predicts that multiple groups of predicted reaction parameters that may cause poor reaction are obtained. On this basis, the monitoring data obtained by the node is further compared, so as to lock the specific adverse reaction parameters, and in this way, all abnormal node parameters that cause quality problems of new energy precursors are found from the lower layer to the upper layer step by step.

[0117] It can be understood that the production process of new energy precursors includes multiple reaction nodes, and each reaction node includes multiple possible abnormal node parameters. If each node parameter is traced, the efficiency is low. The present invention predicts the node path that may cause quality problems through the node prediction model, and further combines the comparison and verification of actual monitoring data to finally lock the specific abnormal node parameters, thereby improving the quality traceability efficiency in the complex production process of new energy precursors.

[0118] It should be noted that the corresponding historical time point of the present invention is the time point corresponding to the reaction of each node in the upper layer when the raw materials used to prepare the current product react according to the production process timeline of the new energy precursor.

[0119] According to an embodiment of the present invention, after inputting the bad attribute data into the node prediction model of the first-layer reaction node and outputting multiple groups of predicted reaction parameters causing the bad attribute data, the method further includes:

[0120] Acquire multiple historical record data of the first-layer reaction node, each historical record data at least including: historical bad attribute data, and historical actual reaction raw material data and historical actual reaction condition data causing the historical bad attribute data;

[0121] Based on multiple historical record data, extract the historical bad attribute data therein and calculate the extracted data according to a preset processing algorithm to obtain a historical characterization value;

[0122] Extract the bad property data of the current new energy precursor and calculate the extracted data using the same processing algorithm to obtain the current characterization value;

[0123] Based on multiple historical record data, similarity calculation is performed between the historical characterization value corresponding to each historical record data and the current characterization value corresponding to the bad attribute data of the current new energy precursor;

[0124] Adding historical record data with a similarity greater than a first threshold into a revised sample database;

[0125] Based on each historical record data in the modified sample database, the historical bad attribute data is input into the node prediction model, and the historical predicted reaction raw material data and the historical predicted reaction condition data causing the historical bad attribute data are output;

[0126] Based on each historical record data in the correction sample database, the historical actual reaction raw material data and the historical actual reaction condition data are respectively subtracted from the corresponding historical predicted reaction raw material data and the historical predicted reaction condition data to obtain the difference of the reaction raw material data and the difference of the reaction condition data;

[0127] Calculate the average values ​​of the difference values ​​of the reaction raw material data and the difference values ​​of the reaction condition data of all the historical record data in the correction sample database to obtain the correction value of the reaction raw material data and the correction value of the reaction condition data;

[0128] Based on the correction value of the reaction raw material data, the reaction raw material bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction raw material bad parameter data; based on the correction value of the reaction condition data, the reaction condition bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction condition bad parameter data.

[0129] It should be noted that, due to the problem of limited training sample data of the node prediction model itself, certain errors may occur in the reaction raw material bad parameter data and reaction condition bad parameter data predicted by the node prediction model. The present invention corrects the data predicted by the node prediction model by combining the historical record data of the same node, thereby improving the accuracy of the prediction, and further ensuring that in the subsequent matching and comparison process of the reaction raw material bad parameter data and the reaction condition bad parameter data with the reaction raw material monitoring data and the reaction condition monitoring data, efficient identification and judgment are performed, and the wrong matching and comparison screening is not caused by the inaccurate prediction data of the node prediction model.

[0130] According to an embodiment of the present invention, data extraction is performed on the historical bad attribute data and the extracted data is calculated according to a preset processing algorithm to obtain a historical characterization value; data extraction is performed on the bad attribute data of the current new energy precursor and the extracted data is calculated according to the same processing algorithm to obtain a current characterization value, specifically including:

[0131] Extract historical percentage values ​​of K compounds from historical adverse attribute data;

[0132] Sort the K compounds in a preset order, use the serial number of each compound as the horizontal axis value, and use the historical percentage value of each compound as the vertical axis value, and mark the coordinate points corresponding to the historical percentage values ​​of the compounds in K in the two-dimensional coordinate system;

[0133] The coordinate points corresponding to the historical percentage values ​​of the compounds in K are smoothly connected to obtain a historical representation line for representing the historical representation value;

[0134] Extract the current percentage values ​​of K compounds from the current adverse property data of new energy precursors;

[0135] Arrange the K compounds in the same order, use the serial number of each compound as the horizontal coordinate value, and use the current percentage value of each compound as the vertical coordinate value, and mark the coordinate points corresponding to the current percentage values ​​of the compounds in K in the two-dimensional coordinate system;

[0136] The coordinate points corresponding to the current percentage values ​​of the compounds in K are smoothly connected to obtain a current characterization line for representing the current characterization value.

[0137] It should be noted that the compounds in K can include the proportion of the target compound, or the purity proportion, and the proportion of each impurity compound. For example, the higher the proportion of certain impurity compounds, the greater the impact on the quality of the new energy precursor. The present invention analyzes and extracts the percentage of each compound in each product, and calculates it according to a preset processing algorithm to obtain a historical characterization line that can be used to represent the historical characterization value, and a current characterization line for representing the current characterization value, so as to facilitate the subsequent screening of multiple historical record data according to the similarity comparison process, and further improve the accuracy of the correction value.

[0138] According to an embodiment of the present invention, the historical characterization value corresponding to each historical record data is respectively calculated with the current characterization value corresponding to the bad attribute data of the current new energy precursor, specifically including:

[0139] The first serial number value of the compound in K is obtained as H, and the last serial number value is calculated as H+K;

[0140] Draw the straight lines x=H and x=H+K in the two-dimensional coordinate system;

[0141] Intersect the historical representation line, the horizontal coordinate, the straight line x=H, and the straight line x=H+K to enclose the historical area;

[0142] The current characterization line, the horizontal coordinate, the straight line x=H, and the straight line x=H+K are intersected in pairs to enclose the current region, and the area of ​​the current region is calculated;

[0143] Draw the overlapping area between the historical area and the current area, and calculate the area of ​​the overlapping area;

[0144] Divide the area of ​​the overlapping region by the area of ​​the current region to get the similarity between the two.

[0145] It can be understood that since it is difficult to specifically calculate the similarity between the historical characterization line and the current characterization line, the present invention transforms the historical characterization line and the current characterization line, and transforms them into the overlap calculation between the historical area and the current area, so as to accurately calculate the similarity between the two, further ensuring the subsequent accurate screening of historical record data with correction reference value, thereby improving the accuracy of the correction value.

[0146] like Figure 2 As shown in the figure, based on the abnormal node parameters and combined with the on-site monitoring data, the corresponding abnormal cause is analyzed, and the corresponding recovery solution is formulated according to the abnormal cause, including:

[0147] S202, based on the abnormal node parameters and according to the association relationship, find multiple possible cause locations associated therewith;

[0148] S204, presetting different probabilities of each possible cause position pair causing abnormal node parameters, and calculating the probability of each possible cause position pair causing abnormal node parameters based on historical data statistics;

[0149] S206, acquiring monitoring data of each possible cause location, and performing abnormal analysis on the monitoring data based on the attention mechanism according to the occurrence probability of each possible cause location;

[0150] S208, focusing on analyzing the monitoring data to obtain the abnormal cause of the abnormal node parameters, and formulating a corresponding recovery solution according to the abnormal cause.

[0151] It should be noted that the reaction equipment includes multiple key positions. Abnormalities in these key positions will cause abnormal node parameters, such as abnormal reaction temperature, abnormal pH value, etc. Therefore, cameras or sensors need to be installed at these key positions to monitor whether there are problems at the key positions. Once it is determined that a problem occurs at a key position, a recovery solution will be immediately formulated to ensure that the production of new energy precursors can proceed in an orderly and normal manner.

[0152] It should be noted that in the abnormal analysis of monitoring data based on the attention mechanism, due to the different probabilities of occurrence of each possible cause position, in order to further improve the speed of finding the abnormal cause. The present invention introduces an attention mechanism, that is, the possible cause position with a higher probability of occurrence needs to invest more computing power for abnormal cause search and analysis, and the possible cause position with a lower probability of occurrence invests less computing power for abnormal cause search and analysis, thereby being able to achieve a faster focus on the cause of the abnormality based on the limited computing power of the processor. It can be understood that the monitoring data can be a monitoring image, and the processor can recognize and analyze the monitoring image to focus on the location of defects in production equipment and damage to key components.

[0153] According to a specific embodiment of the present invention, obtaining the reaction raw material monitoring data and reaction condition monitoring data at the first layer reaction node at the corresponding historical time point specifically includes:

[0154] A plurality of sensors are set on the production equipment of the first-layer reaction node, and the reaction condition monitoring data at different positions are respectively sensed by the plurality of sensors, such as the reaction temperature;

[0155] Obtaining reaction condition monitoring data for each sensor;

[0156] Subtract the reaction condition monitoring data of each sensor from the reaction condition monitoring data of other sensors one by one;

[0157] Determine whether the absolute value of the difference is greater than a third threshold value, and if so, record the former sensor as abnormal once;

[0158] After the reaction condition monitoring data of all sensors have been compared with each other, the total number of times each sensor is recorded as abnormal is accumulated;

[0159] The sensors whose total times exceed the fourth threshold are kicked out, and the average value of the reaction condition monitoring data of the remaining sensors is calculated to obtain the reaction condition monitoring data at the first-layer reaction node.

[0160] It can be understood that the present invention uses multiple sensors to sense the reaction condition monitoring data of multiple locations, and averages the multiple reaction condition monitoring data to obtain monitoring data that is more in line with the scene. The present invention further kicks out the monitoring data of abnormal sensors through internal difference comparison, further avoiding the interference of abnormal sensors in obtaining monitoring data.

[0161] Figure 3 A block diagram of a quality traceability system based on new energy precursors of the present invention is shown.

[0162] like Figure 3 As shown, the second aspect of the present invention further proposes a quality traceability system 3 based on a new energy precursor, comprising a memory 31 and a processor 32, wherein the memory comprises a quality traceability method program based on a new energy precursor, and when the quality traceability method program based on a new energy precursor is executed by the processor, the following steps are implemented:

[0163] Real-time inspection to obtain the quality inspection results of the prepared new energy precursors, and determine whether the quality inspection results of the new energy precursors are qualified;

[0164] If not, extract the bad attribute data that causes the new energy precursor quality inspection result to be unqualified;

[0165] Based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters that cause the bad attributes are inferred;

[0166] Analyze the corresponding abnormal causes based on abnormal node parameters and on-site monitoring data, and formulate corresponding recovery solutions based on the abnormal causes;

[0167] Determine whether the cause of the abnormality is due to human factors. If so, initiate the quality accident accountability procedure and hold the relevant responsible persons accountable.

[0168] According to an embodiment of the present invention, based on the bad attribute data and through a preset source tracing detection algorithm, the abnormal node parameters causing the bad attribute are inferred, specifically including:

[0169] The production process of the preset new energy precursor includes n layers of reaction nodes, and the product obtained from the i+1th layer of reaction nodes is used as the raw material of the ith layer of reaction nodes, and the number of the first layer of reaction nodes is 1, and the number of other layer reaction nodes is greater than or equal to 1, where i is greater than or equal to 1 and less than or equal to n-1;

[0170] Build corresponding node prediction models based on each reaction node;

[0171] Inputting the bad attribute data into the node prediction model of the first layer reaction node, and outputting multiple groups of predicted reaction parameters causing the bad attribute data, each group of predicted reaction parameters including one or two of the bad parameter data of the reaction raw material and the bad parameter data of the reaction condition;

[0172] Obtaining the reaction raw material monitoring data and reaction condition monitoring data at the first layer reaction node at the corresponding historical time point;

[0173] Based on multiple groups of predicted reaction parameters, a group of predicted reaction parameters that are completely consistent with the reaction raw material monitoring data and reaction condition monitoring data at the first-layer reaction node are screened out, and the reaction condition bad parameter data contained in the predicted reaction parameters are listed as abnormal node parameters;

[0174] Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step;

[0175] Determine whether the reaction raw material is generated by the corresponding reaction node of the second layer. If not, list the bad parameter data of the reaction raw material as abnormal node parameters. If yes, proceed to the next step.

[0176] Inputting the reaction raw material bad parameter data into the node prediction model of the corresponding reaction node of the second layer, and outputting multiple groups of predicted reaction parameters causing the reaction raw material bad parameter data;

[0177] Acquire reaction raw material monitoring data and reaction condition monitoring data at the corresponding reaction node in the second layer at the corresponding historical time point;

[0178] Based on the multiple groups of predicted reaction parameters of the reaction raw material bad parameter data, a group of predicted reaction parameters that completely match the reaction raw material monitoring data and the reaction condition monitoring data at the corresponding reaction node of the second layer is screened out, and the reaction condition bad parameter data included in the predicted reaction parameters is listed as abnormal node parameters;

[0179] Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step;

[0180] Determine whether the reaction raw material is generated by the corresponding reaction node of the third layer. If not, list the reaction raw material bad parameter data as abnormal node parameters. If yes, continue to trace the reaction raw material bad parameter data to the upper layer until the traced reaction raw material is not generated by the corresponding reaction node of the upper layer.

[0181] According to an embodiment of the present invention, after inputting the bad attribute data into the node prediction model of the first layer reaction node and outputting multiple groups of predicted reaction parameters causing the bad attribute data, the quality traceability method program based on the new energy precursor is further implemented when executed by the processor:

[0182] Acquire multiple historical record data of the first-layer reaction node, each historical record data at least including: historical bad attribute data, and historical actual reaction raw material data and historical actual reaction condition data causing the historical bad attribute data;

[0183] Based on multiple historical record data, extract the historical bad attribute data therein and calculate the extracted data according to a preset processing algorithm to obtain a historical characterization value;

[0184] Extract the bad property data of the current new energy precursor and calculate the extracted data using the same processing algorithm to obtain the current characterization value;

[0185] Based on multiple historical record data, similarity calculation is performed between the historical characterization value corresponding to each historical record data and the current characterization value corresponding to the bad attribute data of the current new energy precursor;

[0186] Adding historical record data with a similarity greater than a first threshold into a revised sample database;

[0187] Based on each historical record data in the modified sample database, the historical bad attribute data is input into the node prediction model, and the historical predicted reaction raw material data and the historical predicted reaction condition data causing the historical bad attribute data are output;

[0188] Based on each historical record data in the correction sample database, the historical actual reaction raw material data and the historical actual reaction condition data are respectively subtracted from the corresponding historical predicted reaction raw material data and the historical predicted reaction condition data to obtain the difference of the reaction raw material data and the difference of the reaction condition data;

[0189] Calculate the average values ​​of the difference values ​​of the reaction raw material data and the difference values ​​of the reaction condition data of all the historical record data in the correction sample database to obtain the correction value of the reaction raw material data and the correction value of the reaction condition data;

[0190] Based on the correction value of the reaction raw material data, the reaction raw material bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction raw material bad parameter data; based on the correction value of the reaction condition data, the reaction condition bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction condition bad parameter data.

[0191] The third aspect of the present invention also proposes a computer-readable storage medium, which includes a quality traceability method program based on a new energy precursor. When the quality traceability method program based on a new energy precursor is executed by a processor, the steps of a quality traceability method based on a new energy precursor as described above are implemented.

[0192] The quality traceability method, system and medium based on new energy precursors proposed in the present invention can infer abnormal node parameters that cause poor properties through a preset traceability detection algorithm. In the production process of new energy precursors, the occurrence of quality problems can be detected in time, the causes of quality problems can be traced and the generating factors can be controlled, and the adverse factors causing quality problems can be eliminated, so that the production process of new energy precursors can be carried out continuously and stably for a long time, thereby improving the enterprise's intelligence level and market competitiveness.

[0193] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0194] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0195] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0196] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0197] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0198] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A quality traceability method based on new energy precursors, characterized in that: The method comprises: Real-time inspection to obtain the quality inspection results of the prepared new energy precursors, and determine whether the quality inspection results of the new energy precursors are qualified, wherein the new energy precursors are an intermediate material for preparing new energy positive electrode materials, and are made of nickel salts, cobalt salts, manganese salts or nickel salts, cobalt salts, aluminum salts as raw materials; If not, extract the bad attribute data that causes the new energy precursor quality inspection result to be unqualified; Based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters causing the bad attributes are inferred. The node parameters at least include the raw material parameters of the chemical reaction in the process node and the condition parameters required for the reaction. The abnormal node parameters indicate that the node parameters are abnormal. Analyze the corresponding abnormal causes based on abnormal node parameters and on-site monitoring data, and formulate corresponding recovery solutions based on the abnormal causes; Determine whether the abnormality is caused by human factors. If so, initiate the quality accident accountability procedure and hold the relevant responsible persons accountable; Based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters that cause the bad attributes are inferred, including: The production process of the preset new energy precursor includes n layers of reaction nodes, and the product obtained from the i+1th layer of reaction nodes is used as the raw material of the ith layer of reaction nodes, and the number of the first layer of reaction nodes is 1, and the number of other layer reaction nodes is greater than or equal to 1, where i is greater than or equal to 1 and less than or equal to n-1; Build corresponding node prediction models based on each reaction node; Inputting the bad attribute data into the node prediction model of the first layer reaction node, and outputting multiple groups of predicted reaction parameters causing the bad attribute data, each group of predicted reaction parameters including one or two of the bad parameter data of the reaction raw material and the bad parameter data of the reaction condition; Obtaining the reaction raw material monitoring data and reaction condition monitoring data at the first layer reaction node at the corresponding historical time point; Based on multiple groups of predicted reaction parameters, a group of predicted reaction parameters that are completely consistent with the reaction raw material monitoring data and reaction condition monitoring data at the first-layer reaction node are screened out, and the reaction condition bad parameter data contained in the predicted reaction parameters are listed as abnormal node parameters; Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step; Determine whether the reaction raw material is generated by the corresponding reaction node of the second layer. If not, list the bad parameter data of the reaction raw material as abnormal node parameters. If yes, proceed to the next step. Inputting the reaction raw material bad parameter data into the node prediction model of the corresponding reaction node of the second layer, and outputting multiple groups of predicted reaction parameters causing the reaction raw material bad parameter data; Acquire reaction raw material monitoring data and reaction condition monitoring data at the corresponding reaction node in the second layer at the corresponding historical time point; Based on the multiple groups of predicted reaction parameters of the reaction raw material bad parameter data, a group of predicted reaction parameters that completely match the reaction raw material monitoring data and the reaction condition monitoring data at the corresponding reaction node of the second layer is screened out, and the reaction condition bad parameter data included in the predicted reaction parameters is listed as abnormal node parameters; Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step; Determine whether the reaction raw material is generated by the corresponding reaction node of the third layer. If not, list the reaction raw material bad parameter data as abnormal node parameters. If yes, continue to trace the reaction raw material bad parameter data to the upper layer until the traced reaction raw material is not generated by the corresponding reaction node of the upper layer.

2. A quality traceability method based on new energy precursors according to claim 1, characterized in that: After inputting the bad attribute data into the node prediction model of the first-layer reaction node and outputting a plurality of groups of predicted reaction parameters causing the bad attribute data, the method further includes: Acquire multiple historical record data of the first-layer reaction node, each historical record data at least including: historical bad attribute data, and historical actual reaction raw material data and historical actual reaction condition data causing the historical bad attribute data; Based on multiple historical record data, extract the historical bad attribute data therein and calculate the extracted data according to a preset processing algorithm to obtain a historical characterization value; Extract the bad property data of the current new energy precursor and calculate the extracted data using the same processing algorithm to obtain the current characterization value; Based on multiple historical record data, similarity calculation is performed between the historical characterization value corresponding to each historical record data and the current characterization value corresponding to the bad attribute data of the current new energy precursor; Adding historical record data with a similarity greater than a first threshold into a revised sample database; Based on each historical record data in the modified sample database, the historical bad attribute data is input into the node prediction model, and the historical predicted reaction raw material data and the historical predicted reaction condition data causing the historical bad attribute data are output; Based on each historical record data in the correction sample database, the historical actual reaction raw material data and the historical actual reaction condition data are respectively subtracted from the corresponding historical predicted reaction raw material data and the historical predicted reaction condition data to obtain the difference of the reaction raw material data and the difference of the reaction condition data; Calculate the average values ​​of the difference values ​​of the reaction raw material data and the difference values ​​of the reaction condition data of all the historical record data in the correction sample database to obtain the correction value of the reaction raw material data and the correction value of the reaction condition data; Based on the correction value of the reaction raw material data, the reaction raw material bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction raw material bad parameter data; based on the correction value of the reaction condition data, the reaction condition bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction condition bad parameter data.

3. A quality traceability method based on new energy precursors according to claim 2, characterized in that: The historical bad property data is extracted and the extracted data is calculated according to a preset processing algorithm to obtain a historical characterization value; the bad property data of the current new energy precursor is extracted and the extracted data is calculated according to the same processing algorithm to obtain a current characterization value, specifically including: Extract historical percentage values ​​of K compounds from historical adverse attribute data; Sort the K compounds in a preset order, use the serial number of each compound as the horizontal axis value, and use the historical percentage value of each compound as the vertical axis value, and mark the coordinate points corresponding to the historical percentage values ​​of the compounds in K in the two-dimensional coordinate system; The coordinate points corresponding to the historical percentage values ​​of the compounds in K are smoothly connected to obtain a historical representation line for representing the historical representation value; Extract the current percentage values ​​of K compounds from the current adverse property data of new energy precursors; Arrange the K compounds in the same order, use the serial number of each compound as the horizontal coordinate value, and use the current percentage value of each compound as the vertical coordinate value, and mark the coordinate points corresponding to the current percentage values ​​of the compounds in K in the two-dimensional coordinate system; The coordinate points corresponding to the current percentage values ​​of the compounds in K are smoothly connected to obtain a current characterization line for representing the current characterization value.

4. A quality traceability method based on new energy precursors according to claim 3, characterized in that: The historical characterization value corresponding to each historical record data is respectively calculated with the current characterization value corresponding to the bad attribute data of the current new energy precursor for similarity, specifically including: The first serial number value of the compound in K is obtained as H, and the last serial number value is calculated as H+K; Draw the straight lines x=H and x=H+K in the two-dimensional coordinate system; Intersect the historical representation line, the horizontal coordinate, the straight line x=H, and the straight line x=H+K to enclose the historical area; The current characterization line, the horizontal coordinate, the straight line x=H, and the straight line x=H+K are intersected in pairs to enclose the current region, and the area of ​​the current region is calculated; Draw the overlapping area between the historical area and the current area, and calculate the area of ​​the overlapping area; Divide the area of ​​the overlapping region by the area of ​​the current region to get the similarity between the two.

5. The quality traceability method based on new energy precursors according to claim 1 is characterized in that: Analyze the abnormal cause based on the abnormal node parameters and on-site monitoring data, and formulate corresponding recovery solutions according to the abnormal cause, including: Based on the abnormal node parameters and the association relationship, multiple possible cause locations associated with it are found; The probability of abnormal node parameters caused by each possible cause position pair is different, and the probability of abnormal node parameters caused by each possible cause position pair is calculated based on historical data statistics; Obtain monitoring data for each possible cause location, and perform anomaly analysis on the monitoring data based on the attention mechanism according to the probability of occurrence of each possible cause location; Focus on analyzing the monitoring data to find out the abnormal reasons that lead to abnormal node parameters, and formulate corresponding recovery solutions based on the abnormal reasons.

6. A quality traceability system based on new energy precursors, characterized in that: The invention comprises a memory and a processor, wherein the memory comprises a quality traceability method program based on a new energy precursor, and the quality traceability method program based on a new energy precursor is executed by the processor to implement the following steps: Real-time inspection to obtain the quality inspection results of the prepared new energy precursors, and determine whether the quality inspection results of the new energy precursors are qualified, wherein the new energy precursors are an intermediate material for preparing new energy positive electrode materials, and are made of nickel salts, cobalt salts, manganese salts or nickel salts, cobalt salts, aluminum salts as raw materials; If not, extract the bad attribute data that causes the new energy precursor quality inspection result to be unqualified; Based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters causing the bad attributes are inferred. The node parameters at least include the raw material parameters of the chemical reaction in the process node and the condition parameters required for the reaction. The abnormal node parameters indicate that the node parameters are abnormal. Analyze the corresponding abnormal causes based on abnormal node parameters and on-site monitoring data, and formulate corresponding recovery solutions based on the abnormal causes; Determine whether the abnormality is caused by human factors. If so, initiate the quality accident accountability procedure and hold the relevant responsible persons accountable; Based on the bad attribute data and through the preset traceability detection algorithm, the abnormal node parameters that cause the bad attributes are inferred, including: The production process of the preset new energy precursor includes n layers of reaction nodes, and the product obtained from the i+1th layer of reaction nodes is used as the raw material of the ith layer of reaction nodes, and the number of the first layer of reaction nodes is 1, and the number of other layer reaction nodes is greater than or equal to 1, where i is greater than or equal to 1 and less than or equal to n-1; Build corresponding node prediction models based on each reaction node; Inputting the bad attribute data into the node prediction model of the first layer reaction node, and outputting multiple groups of predicted reaction parameters causing the bad attribute data, each group of predicted reaction parameters including one or two of the bad parameter data of the reaction raw material and the bad parameter data of the reaction condition; Obtaining the reaction raw material monitoring data and reaction condition monitoring data at the first layer reaction node at the corresponding historical time point; Based on multiple groups of predicted reaction parameters, a group of predicted reaction parameters that are completely consistent with the reaction raw material monitoring data and reaction condition monitoring data at the first-layer reaction node are screened out, and the reaction condition bad parameter data contained in the predicted reaction parameters are listed as abnormal node parameters; Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step; Determine whether the reaction raw material is generated by the corresponding reaction node of the second layer. If not, list the bad parameter data of the reaction raw material as abnormal node parameters. If yes, proceed to the next step. Inputting the reaction raw material bad parameter data into the node prediction model of the corresponding reaction node of the second layer, and outputting multiple groups of predicted reaction parameters causing the reaction raw material bad parameter data; Acquire reaction raw material monitoring data and reaction condition monitoring data at the corresponding reaction node in the second layer at the corresponding historical time point; Based on the multiple groups of predicted reaction parameters of the reaction raw material bad parameter data, a group of predicted reaction parameters that completely match the reaction raw material monitoring data and the reaction condition monitoring data at the corresponding reaction node of the second layer is screened out, and the reaction condition bad parameter data included in the predicted reaction parameters is listed as abnormal node parameters; Based on the screened predicted reaction parameters, determine whether the reaction raw material bad parameter data is included, if not, end the process, if yes, proceed to the next step; Determine whether the reaction raw material is generated by the corresponding reaction node of the third layer. If not, list the reaction raw material bad parameter data as abnormal node parameters. If yes, continue to trace the reaction raw material bad parameter data to the upper layer until the traced reaction raw material is not generated by the corresponding reaction node of the upper layer.

7. A quality traceability system based on new energy precursors according to claim 6, characterized in that: After inputting the bad attribute data into the node prediction model of the first-layer reaction node and outputting multiple groups of predicted reaction parameters causing the bad attribute data, the quality traceability method program based on the new energy precursor is further implemented when executed by the processor: Acquire multiple historical record data of the first-layer reaction node, each historical record data at least including: historical bad attribute data, and historical actual reaction raw material data and historical actual reaction condition data causing the historical bad attribute data; Based on multiple historical record data, extract the historical bad attribute data therein and calculate the extracted data according to a preset processing algorithm to obtain a historical characterization value; Extract the bad property data of the current new energy precursor and calculate the extracted data using the same processing algorithm to obtain the current characterization value; Based on multiple historical record data, similarity calculation is performed between the historical characterization value corresponding to each historical record data and the current characterization value corresponding to the bad attribute data of the current new energy precursor; Adding historical record data with a similarity greater than a first threshold into a revised sample database; Based on each historical record data in the modified sample database, the historical bad attribute data is input into the node prediction model, and the historical predicted reaction raw material data and the historical predicted reaction condition data causing the historical bad attribute data are output; Based on each historical record data in the correction sample database, the historical actual reaction raw material data and the historical actual reaction condition data are respectively subtracted from the corresponding historical predicted reaction raw material data and the historical predicted reaction condition data to obtain the difference of the reaction raw material data and the difference of the reaction condition data; Calculate the average values ​​of the difference values ​​of the reaction raw material data and the difference values ​​of the reaction condition data of all the historical record data in the correction sample database to obtain the correction value of the reaction raw material data and the correction value of the reaction condition data; Based on the correction value of the reaction raw material data, the reaction raw material bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction raw material bad parameter data; based on the correction value of the reaction condition data, the reaction condition bad parameter data in the predicted reaction parameters are added to obtain the corrected reaction condition bad parameter data.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a quality traceability method program based on a new energy precursor. When the quality traceability method program based on a new energy precursor is executed by a processor, the steps of a quality traceability method based on a new energy precursor as described in any one of claims 1 to 5 are implemented.

Images