Recommended methods for production process parameters and related equipment and program media

Abstract

The present application discloses a method, device, and program medium for recommending production process parameters based on abnormal operating conditions. The method includes: if no process parameter version is available in the production process under abnormal operating conditions or the index evaluation value of the process parameter version does not meet the standard, determining the process parameter recommendation value according to the abnormal operating condition recommendation algorithm; obtaining the actual values ​​of historical product performance parameters and the historical product performance parameter prediction values ​​of multiple rounds under the abnormal operating conditions stored in a database; calculating the product performance parameter prediction deviation value based on the historical product performance parameter actual values ​​and the historical product performance parameter prediction values; inputting the process parameter recommendation value into a multi-objective optimization model to obtain the current product performance parameter prediction value; and obtaining the compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value. The present application embodiment can improve the accuracy of the prediction algorithm and ensure production quality.

Description

Technical Field

[0001] The present application relates to automation technology, which is applied to fields such as chemical production, and in particular to a method, device and program medium for recommending production process parameters based on abnormal working conditions. Background Art

[0002] The usual benchmark operating condition database usually stores relatively simple production process and production conditions that cannot be adjusted in the short term (material conditions, external environment (temperature, humidity, etc.) conditions). The above simple conditions can also be called standard conditions. Since there may be as many as hundreds of process parameters in the production process, when any process parameter (such as steam pressure, production gas pressure, equipment vacuum, etc.) exceeds the pre-set tolerance range of the standard value due to uncontrollable factors such as equipment load, and cannot be restored for a long time, in theory, the process parameter will become a special non-standard operating condition parameter, which will cause abnormal conditions in the production process. At this time, if production continues according to the pre-set standard value, the production result index will not be achieved, which will lead to too low production efficiency of the product. Summary of the Invention

[0003] The embodiments of the present application provide a method, device, and program medium for recommending production process parameters based on abnormal operating conditions, which can timely adjust the process parameter values ​​of later production equipment to adapt to changes in abnormal operating conditions, thereby improving prediction accuracy, ensuring production quality, and meeting product production result indicators.

[0004] In a first aspect, an embodiment of the present application provides a method for recommending production process parameters based on abnormal working conditions, the method comprising:

[0005] If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard, the process parameter recommended value is determined according to the abnormal working condition recommendation algorithm;

[0006] Obtaining actual values ​​of historical product performance parameters and predicted values ​​of historical product performance parameters for multiple rounds under the abnormal operating conditions stored in a database;

[0007] Calculating a product performance parameter prediction deviation value based on the actual value of the historical product performance parameter and the predicted value of the historical product performance parameter, wherein the predicted value of the historical product performance parameter is obtained by prediction using a multi-objective optimization model;

[0008] Inputting the recommended process parameter values ​​into the multi-objective optimization model to obtain predicted values ​​of current product performance parameters;

[0009] Based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, the compensated current product performance parameter prediction value is obtained, wherein the compensated current product performance parameter prediction value is used to evaluate whether to use the process parameter recommended value for production.

[0010] In the prior art, any process parameter of the production equipment (such as steam pressure, production gas pressure, equipment vacuum, etc.) may cause abnormal working conditions in the production process due to uncontrollable factors such as equipment load, causing the process parameter to exceed the tolerance range preset by the standard value, and when it cannot be restored for a long time, in theory, the process parameter will become a special non-standard working condition parameter. At this time, production still needs to continue. If production continues according to the pre-set standard value, the production result index will not be achieved. However, the present application can obtain the actual value of the historical product performance parameter under abnormal working conditions, combine the actual value of each historical product performance parameter of the product and the predicted value of each historical product performance parameter of the product to calculate the product performance parameter prediction deviation value, determine the process parameter recommendation value through the abnormal working condition recommendation algorithm, and then input the process parameter recommendation value into the multi-objective optimization model to obtain the current product performance parameter prediction value, and then compensate the current product performance parameter prediction value according to the product performance parameter prediction deviation value to obtain the compensated current each product performance parameter prediction value, thereby improving the accuracy of the prediction algorithm, ensuring the product quality of the production process, and meeting the product production result index.

[0011] In a possible implementation, if there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard, before determining the recommended process parameter value according to the abnormal working condition recommendation algorithm, the method further includes:

[0012] Determine whether an abnormal operating condition occurs in the current production process, wherein the abnormal operating condition is an abnormal operating condition that occurs in the production process when production is under normal standard conditions;

[0013] If the abnormal working condition occurs, obtaining the working condition code of the abnormal working condition;

[0014] According to the working condition code, query in the database whether there is an available process parameter version or whether the index evaluation value of the process parameter version meets the standard;

[0015] If there is no available process parameter version in the production process under abnormal working conditions or the indicator evaluation value of the process parameter version does not meet the standard, the step of determining the recommended value of the process parameter according to the abnormal working condition recommendation algorithm is executed.

[0016] In the above method, suppose that a process parameter in the current production process is abnormal, for example, the process parameter X 16 If an exception occurs, obtain the process parameter X16 The process parameter fault code is 3, and the database code address of the abnormal process parameter is 100 (the database code address of the abnormal process parameter is used to query the abnormal detailed information), then the working condition fault code is 1, 1, 1, 16, 3, 1, 100. Assuming that the benchmark working condition code is 100100100, the abnormal working condition code is (100100100, 1, 1, 1, 16, 3, 1, 100). Then, according to the abnormal working condition code (100100100, 1, 1, 1, 16, 3, 1, 100), the database is queried to see whether there is an available process parameter version or process parameter. Whether the index evaluation value of the version meets the standard. If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard (for example, the available process parameter version used in the historical production process is not queried through the abnormal working condition code, or an available process parameter version is queried, but the index evaluation value corresponding to the available process parameter version is 58 points, which is not within the standard range of 60-80 points), in order to ensure that the process parameter values ​​of the later production equipment can be adjusted in time to adapt to changes in abnormal working conditions, it is necessary to execute the above steps of determining the recommended process parameter values ​​according to the abnormal working condition recommendation algorithm.

[0017] In another possible implementation, after obtaining the compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, the method further includes:

[0018] Determining an index evaluation value based on the compensated current product performance parameter prediction value, wherein the index evaluation value is used to evaluate the comprehensive closeness between each of the multiple product performance parameter prediction values ​​corresponding to the process parameter and the corresponding product performance parameter target value;

[0019] Determining an optimal version of abnormal operating conditions among the process parameter versions, wherein the optimal version of abnormal operating conditions is the process parameter version with the largest indicator evaluation value;

[0020] If the indicator evaluation value is greater than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, production is carried out according to the process parameter value corresponding to the current indicator evaluation value;

[0021] If the indicator evaluation value is less than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, production is performed according to the optimal version of the abnormal working condition.

[0022] In the above method, there is at least one available process parameter version. Each time the abnormal working condition recommendation algorithm is optimized, the priority level of each product performance parameter is determined based on the target value of each product performance parameter of multiple products obtained by the current multi-objective optimization model, wherein the priority level is used to indicate the degree of influence of the product performance parameter on the quality of the product. The multiple product performance parameters are then divided according to the priority level to obtain multiple optimization sets, and then the indicator evaluation value of each optimization set in the multiple optimization sets is determined based on the optimization value. Finally, the comprehensive index evaluation value of the current multi-objective optimization is determined according to the index evaluation value of each optimization set, and the multiple process parameter versions are sorted according to the comprehensive index evaluation value to obtain the sorting result, that is, the process parameter version corresponding to the largest index evaluation value is obtained as the optimal version of the abnormal working condition. After each optimization of the abnormal working condition recommendation algorithm is completed, the current index evaluation value is compared with the maximum index evaluation value corresponding to the optimal version of the abnormal working condition. If the current index evaluation value is less than the maximum index evaluation value corresponding to the optimal version of the abnormal working condition, production is carried out according to the optimal version of the abnormal working condition (for example, the current index evaluation value is 75 points, and the maximum index evaluation value corresponding to the optimal version of the abnormal working condition is 86 points, then it means that the production result index obtained by producing according to the optimal version of the abnormal working condition achieves a better target). If the index evaluation value is greater than the index evaluation value corresponding to the optimal version of the abnormal working condition (for example, the current index evaluation value is 90 points, and the maximum index evaluation value corresponding to the optimal version of the abnormal working condition is 86 points), then it means that the abnormal working condition recommendation algorithm is optimized successfully. At this time, production is carried out according to the process parameter value corresponding to the current index evaluation value, which can meet the production result index of the product.

[0023] In yet another possible implementation, the invention further includes:

[0024] If the abnormal operating condition has not appeared in the database, determining a first reference value of the process parameter under normal operating conditions, wherein the first reference value is a target value of the process parameter in the benchmark operating condition scorecard under the normal operating conditions;

[0025] determining a second reference value of the adjustable process parameter;

[0026] If the adjustable process parameter has the process parameter recommended value, the second reference value is the process parameter recommended value of the adjustable process parameter;

[0027] If the process parameter recommended value does not exist for the adjustable process parameter, the second reference value is the target value of the adjustable process parameter, wherein the target value of the adjustable process parameter is the target value of the adjustable process parameter in the benchmark operating condition scorecard;

[0028] Determine the abnormal prediction value of non-adjustable process parameters;

[0029] The process parameter recommended value, the first reference value, the second reference value and the abnormal prediction value of the non-adjustable process parameter are respectively input into the multi-objective optimization model to obtain multiple current product performance index prediction values.

[0030] In the above method, if no abnormal operating conditions have appeared in the database, the types of process parameters may include process parameters that are adjustable under abnormal operating conditions and are undergoing predictive value calculation, process parameters under normal operating conditions, process parameters that are adjustable under abnormal operating conditions but are not undergoing predictive value calculation, process parameters that are not adjustable under abnormal operating conditions, and when abnormal operating conditions occur, the recommended value of the process parameter that is undergoing predictive value calculation and has completed the recommended value algorithm or the default value of the process parameter that is undergoing predictive value calculation and has not completed the recommended value algorithm (target value of the benchmark operating condition scorecard). At this time, the process parameter that is adjustable but is not undergoing predictive value calculation is judged. If a recommended value exists, the recommended value is used; if no recommended value exists, the target value is used.

[0031] Among them, when calculating the product performance index prediction values ​​corresponding to multiple process parameters within the adjustment range of a certain adjustable process parameter that is currently undergoing prediction value calculation, it is necessary to first determine the calculation initial values ​​of the remaining process parameters except for the adjustable process parameter that is currently undergoing prediction value calculation. It can be understood that if there is a recommended value for the adjustable process parameter that is not currently undergoing prediction value calculation, the recommended value is used as the calculation initial value of the adjustable process parameter that is not currently undergoing prediction value calculation; if there is no recommended value for the remaining adjustable process parameter that is not currently undergoing prediction value calculation, the target value is used as the calculation initial value of the adjustable process parameter that is not currently undergoing prediction value calculation; the process parameters under normal conditions use the target value in the benchmark operating condition scorecard as the calculation initial value; the process parameters that cannot be adjusted under abnormal conditions use the abnormal prediction value as the calculation initial value. This method takes into account that the product performance index prediction value is affected by all the process parameters of the product. When calculating the prediction value of the recommended value of the process parameter, the values ​​of the remaining process parameters are also pre-set, which can make the prediction value more accurate.

[0032] In another possible implementation, after determining the index evaluation value according to the compensated current product performance parameter prediction value, the method further includes:

[0033] If the index evaluation value obtained in the current round of production is less than the preset index evaluation reference value, the current product performance parameter prediction deviation value is recalculated based on the current product performance parameter actual value and the historical product performance parameter actual value;

[0034] Determining the compensated current product performance parameter prediction value according to the current product performance parameter prediction deviation value and the current product performance parameter prediction value;

[0035] The abnormal working condition recommendation algorithm is optimized according to the current product performance parameter prediction value after compensation, until the index evaluation value is controlled within the preset index evaluation reference value range during multiple consecutive productions, and the optimization ends. In the above method, if the determined index evaluation value is less than the preset index evaluation reference value, it means that the product performance index of multiple historical production rounds does not meet the standard (for example, one or more of the performance parameters such as the primary amine content, iodine value, and primary amide of the amine is less than the standard range interval of the corresponding performance parameter), the abnormal working condition recommendation algorithm can be optimized and trained according to the actual value of each product performance parameter currently produced and its corresponding product performance parameter prediction value, and then according to the current performance parameter prediction value of each product after compensation, a new mathematical model is generated, and the previous recommendation algorithm optimization is repeated until the index evaluation value after multiple trainings is stable within the preset threshold range (for example, the index evaluation value remains between 75 and 80 points, and the preset index evaluation reference value range is 60 to 80 points, so the optimized performance index of the current production product meets the reference standard), and the algorithm is no longer optimized. This solution evaluates the comprehensive closeness between each of the multiple product performance parameter prediction values ​​corresponding to the process parameters and the corresponding product performance parameter target value, derives the recommended value of the process parameter, and uses the recommended value of the process parameter for production, which can effectively improve the quality of the produced products and the accuracy of the prediction algorithm.

[0036] In the second aspect, an embodiment of the present application provides a device for recommending production process parameters based on abnormal working conditions, which includes a selection unit, an acquisition unit, a calculation unit and an input unit. The device is used to implement the method described in the first aspect or any possible implementation method of the first aspect.

[0037] It should be noted that the processor included in the apparatus described in the second aspect above may be a processor specifically used to execute these methods (referred to as a dedicated processor for ease of distinction), or may be a processor that executes these methods by calling a computer program, such as a general-purpose processor. Optionally, the at least one processor may include both a dedicated processor and a general-purpose processor.

[0038] Optionally, the computer program may be stored in a memory. For example, the memory may be a non-transitory memory, such as a read-only memory (ROM), which may be integrated with the processor on the same device or provided on separate devices. The embodiments of the present application do not limit the type of memory or the configuration of the memory and the processor.

[0039] In a possible implementation manner, the at least one memory is located outside the recommendation device.

[0040] In yet another possible implementation, the at least one memory is located within the recommendation device.

[0041] In another possible implementation manner, part of the at least one memory is located within the recommendation device, and another part of the memory is located outside the recommendation device.

[0042] In this application, the processor and the memory may also be integrated into one device, that is, the processor and the memory may also be integrated together.

[0043] In a third aspect, an embodiment of the present application provides a device for recommending production process parameters based on abnormal working conditions, the device comprising a processor and a memory; a computer program is stored in the memory; when the processor executes the computer program, the computing device executes the method described in any one of the first or first aspects above.

[0044] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed on at least one processor, the method described in any one of the first aspects is implemented.

[0045] In a fifth aspect, the present application provides a computer program product, comprising computer instructions that, when executed on at least one processor, implement the method described in any one of the first aspects. The computer program product may be a software installation package. When the aforementioned method is required, the computer program product may be downloaded and executed on a computing device.

[0046] The beneficial effects of the technical methods provided in the second to fifth aspects of this application can refer to the beneficial effects of the technical solution of the first aspect, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The following is a brief introduction to the drawings used in describing the embodiments.

[0048] Figure 1 1 is a schematic diagram of the architecture of a production process parameter recommendation system based on abnormal working conditions provided in an embodiment of the present application;

[0049] Figure 2 This is a flow chart of a method for recommending production process parameters based on abnormal working conditions, provided in an embodiment of the present application;

[0050] Figure 3 3 is a structural diagram of a device 30 for recommending production process parameters based on abnormal working conditions provided in an embodiment of the present application;

[0051] Figure 4 4 is a structural diagram of a device 40 for recommending production process parameters based on abnormal working conditions provided in an embodiment of the present application. DETAILED DESCRIPTION

[0052] The embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0053] To facilitate understanding, a brief introduction to the technical terms involved in the embodiments of this application is first given.

[0054] 1. Process parameters refer to all index parameters in the production process that are used to control the production process in order to ensure the quality, efficiency and cost of the products produced and to meet user needs. These include the control parameters of the production equipment (such as temperature, pressure, speed, etc.), production time and feed amount, etc.

[0055] The above-mentioned explanations of related terms can be applied to the following embodiments.

[0056] The following is an introduction to the architecture of the embodiments of the present application.

[0057] See Figure 1 , Figure 1 This is an architectural diagram of a system for recommending production process parameters based on abnormal working conditions provided in an embodiment of the present application. The system includes a server 101, and the server 101 includes a data acquisition device 102, a database 103, a training device 104, a data processing module 105, and an output module 106.

[0058] Server 101 can be a server or a server cluster composed of multiple servers, which can be a computer or a host computer. Server 101 is mainly used to train models according to production conditions and obtain a multi-objective optimization model of the production conditions. When abnormal process parameters occur, the training model is used to perform a prediction algorithm, adjust the normal adjustable process parameters, and give recommended values ​​(optimal values) of the normal adjustable process parameters, so that the adjusted production result indicators reach the optimal target.

[0059] The data acquisition device 102 is used to obtain the actual value of each product performance parameter of the product in the historical production round under abnormal working conditions, the target value of the non-optimizable process parameter, the process parameters under normal working conditions, the process parameters under abnormal working conditions and the working condition code of the abnormal working condition stored in the database.

[0060] The training device 104 obtains a multi-objective optimization model / rule based on the process parameter training under abnormal working conditions in the database 103, wherein the multi-objective optimization model / rule can be a mathematical model or an algorithmic model, etc. Through the algorithmic prediction of the multi-objective optimization model / rule, the predicted value of each historical product performance parameter / current product performance parameter of the product can be obtained.

[0061] It should be noted that, in actual applications, the data in the database 103 may not all be collected by the data acquisition device 102, but may also be sent by other devices. It should also be noted that the training device 104 may not necessarily train the multi-objective optimization model / rules based entirely on the training data set in the database 103, but may also obtain the training data set from the cloud or other places for model training. The above description should not be used as a limitation on the embodiments of the present application.

[0062] In an embodiment of the present application, the data processing module 105 is used to calculate the product performance parameter prediction deviation value based on the actual value of each product performance parameter of the product and the predicted value of each product performance parameter of the product, and can also be used to determine the recommended value of the process parameter based on the target value of the process parameter that cannot be optimized under abnormal working conditions.

[0063] The output module 106 is used to input the recommended process parameter values ​​into the multi-objective optimization model / rules to obtain the predicted value of each historical product performance parameter / current product performance parameter of the product.

[0064] The method of the embodiment of the present application is described in detail below.

[0065] See Figure 2 , Figure 2 This is a flow chart of a method for recommending production process parameters based on abnormal working conditions provided by an embodiment of the present application. Figure 1 It can be implemented based on the system architecture shown in the figure, and can also be implemented based on other system architectures. Specifically, it can be applied to servers, such as Figure 2 The method at least includes steps S201 to S205.

[0066] Step S201: If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard, the server determines the recommended value of the process parameter according to the abnormal working condition recommendation algorithm.

[0067] Specifically, process parameters are used to control the production process, ensuring the quality, efficiency, and cost of the products produced. They are all production process parameters that can meet user requirements, including equipment control parameters (temperature, pressure, speed), production steps, production time, feed rate, etc. For example, the parameters that affect the quality, efficiency, and cost of a type of product can be classified into a process parameter version, which is the parameter version used under abnormal operating conditions. For example, a process parameter version A1 stored in the database includes multiple parameters that affect the current production process, such as temperature (70°C), hydrogen pressure (700Pa), time (480min), etc., and process parameter version A2 also includes multiple parameters that affect the current production process, such as temperature (75°C), hydrogen pressure (850Pa), time (480min), etc. It is understandable that the values ​​of the parameters in process parameter version A1 and process parameter version A2 may be different or the same. Then, if there is no available process parameter version in the production process under abnormal working conditions or the indicator evaluation value of the process parameter version does not meet the standard, in order to ensure that the process parameter values ​​of the later production equipment can be adjusted in time to adapt to the changes in abnormal working conditions, it is necessary to determine the recommended process parameter values ​​based on the abnormal working condition recommendation algorithm in order to perform subsequent operations.

[0068] Optionally, before determining the recommended values ​​of process parameters according to the abnormal working condition recommendation algorithm, the server determines whether abnormal working conditions occur in the current production process. If abnormal working conditions occur, the working condition code of the abnormal working condition is obtained; according to the working condition code, the server queries in the database whether there is an available process parameter version or whether the indicator evaluation value of the process parameter version meets the standard; if there is no available process parameter version in the production process under the abnormal working condition or the indicator evaluation value of the process parameter version does not meet the standard, the above steps of determining the recommended values ​​of process parameters according to the abnormal working condition recommendation algorithm are executed.

[0069] Specifically, abnormal working conditions refer to abnormal working conditions that occur during the production process when production is carried out under normal standard working conditions. If there are P1~P n There are m production processes, each of which has m process parameters X. Suppose that one process parameter in the current production process is abnormal, for example, the process parameter X of production process P1 is 16 If an exception occurs, obtain the process parameter X 16The process parameter fault code is 3, and the database code address of the abnormal process parameter is 100 (the database code address of the abnormal process parameter is used to query the abnormal detailed information), then the working condition fault code is 1, 1, 1, 16, 3, 1, 100. Assuming that the benchmark working condition code is 100100100, the abnormal working condition code is (100100100, 1, 1, 1, 16, 3, 1, 100). Then, according to the abnormal working condition code (100100100, 1, 1, 1, 16, 3, 1, 100), the database is queried to see whether there is an available process parameter version or process parameter. Whether the index evaluation value of the version meets the standard. If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard (for example, the available process parameter version used in the historical production process is not queried through the abnormal working condition code, or an available process parameter version is queried, but the index evaluation value corresponding to the available process parameter version is 58 points, which is not within the standard range of 60-80 points), in order to ensure that the process parameter values ​​of the later production equipment can be adjusted in time to adapt to changes in abnormal working conditions, it is necessary to execute the above steps of determining the recommended process parameter values ​​according to the abnormal working condition recommendation algorithm.

[0070] Optionally, if no abnormal operating condition has occurred in the database, determining a first reference value of the process parameter under normal operating conditions, wherein the first reference value is a target value of the process parameter in a benchmark operating condition scorecard under normal operating conditions;

[0071] determining a second reference value of the adjustable process parameter;

[0072] If the adjustable process parameter has a recommended process parameter value, the second reference value is the recommended process parameter value of the adjustable process parameter;

[0073] If there is no recommended process parameter value for the adjustable process parameter, the second reference value is the target value for the adjustable process parameter, wherein the target value for the adjustable process parameter is the target value for the adjustable process parameter in the benchmark operating condition scorecard;

[0074] Determine the abnormal prediction value of non-adjustable process parameters;

[0075] The process parameter value, the first reference value, the second reference value and the abnormal prediction value of the non-adjustable process parameter are respectively input into the multi-objective optimization model to obtain multiple current product performance index prediction values.

[0076] Specifically, if no abnormal operating conditions have occurred in the database, the types of process parameters may include process parameters that are adjustable under abnormal operating conditions and are undergoing predictive value calculations, process parameters under normal operating conditions, process parameters that are adjustable under abnormal operating conditions but are not undergoing predictive value calculations, process parameters that are not adjustable under abnormal operating conditions, and recommended values ​​for process parameters that are undergoing predictive value calculations for which the recommended value algorithm has been completed or default values ​​(target values ​​of the benchmark operating condition scorecard) for process parameters that are undergoing predictive value calculations for which the recommended value algorithm has not been completed when abnormal operating conditions occur. At this time, a judgment is made on the process parameter that is adjustable but is not undergoing predictive value calculations. If a recommended value exists, the recommended value is used; if no recommended value exists, the target value is used.

[0077] Among them, when calculating the product performance index prediction values ​​corresponding to multiple process parameters within the adjustment range of a certain adjustable process parameter that is currently undergoing prediction value calculation, it is necessary to first determine the calculation initial values ​​of the remaining process parameters except for the adjustable process parameter that is currently undergoing prediction value calculation. It can be understood that if there is a recommended value for the adjustable process parameter that is not currently undergoing prediction value calculation, the recommended value is used as the calculation initial value of the adjustable process parameter that is not currently undergoing prediction value calculation; if there is no recommended value for the remaining adjustable process parameter that is not currently undergoing prediction value calculation, the target value is used as the calculation initial value of the adjustable process parameter that is not currently undergoing prediction value calculation; the process parameters under normal conditions use the target value in the benchmark operating condition scorecard as the calculation initial value; the process parameters that cannot be adjusted under abnormal conditions use the abnormal prediction value as the calculation initial value. This method takes into account that the product performance index prediction value is affected by all the process parameters of the product. When calculating the prediction value of the recommended value of the process parameter, the values ​​of the remaining process parameters are also pre-set, which can make the prediction value more accurate.

[0078] Step S202: The server obtains actual values ​​of historical product performance parameters and predicted values ​​of historical product performance parameters for multiple rounds under abnormal working conditions stored in a database.

[0079] Specifically, regarding the relevant content of the database used in the production process, the blockchain is a distributed shared account book and database with the characteristics of decentralization, non-tampering, full traceability, traceability, collective maintenance, openness and transparency. Blockchain technology is a new distributed infrastructure and computing method that uses block chain data structure to verify and store data, uses distributed node consensus algorithm to generate and update data, uses cryptography to ensure the security of data transmission and access, and uses smart contracts composed of automated script code to program and operate data. In the embodiment of the present application, obtaining the data parameters stored in the database can be achieved by using blockchain technology. For example, obtaining the historical product performance parameter values ​​of multiple rounds under abnormal working conditions can be the same blockchain, obtaining the current product performance parameter values ​​under abnormal working conditions can also be the same blockchain, and obtaining the historical product performance parameter values ​​and the current product performance parameter values ​​can be different blockchains.

[0080] Specifically, the method includes sharding the performance parameter data of products produced in multiple rounds under abnormal working conditions and / or the performance parameter data of current products under abnormal working conditions stored in the database, and then homomorphically encrypting the data corresponding to at least one shard to obtain ciphertext data. The ciphertext data is synchronized to the computing node device through the same blockchain (such as a blockchain containing all historical performance parameter values) to obtain the encryption key of at least one shard, and then generating a hash based on the encryption key of at least one shard or the data corresponding to the shard, and then adding the hash to the ledger and shard metadata. The ledger stores detailed information related to the transaction, such as shard location, shard hash, and rental cost, so as to link the transaction to the stored shard. Finally, all transactions in the same blockchain ledger are recorded through the storage system, and all data of the same blockchain and different blockchains are synchronized between all nodes.

[0081] Specific steps for obtaining the actual values ​​of historical product performance parameters for multiple rounds under abnormal working conditions stored in the database: For example, first, the server needs to query the historical production rounds under the current abnormal working conditions in the database with the coding address 100. If a production round consists of process steps P1-P6, P1-P5 represents the normal production process in a round, and P6 is the process where an abnormality occurs in the round. Five historical rounds represent five production situations in which P1-P5 are normal and P6 is abnormal. If the current abnormal working condition occurs in the production process, the historical production round is 5. At this time, the server can obtain the actual values ​​of historical product performance parameters for multiple rounds. For example, the actual values ​​of historical product performance parameters corresponding to the first round of production are expressed as N1—primary amine content 95%—primary amide 0.12%—iodine value 45g, the actual values ​​of historical product performance parameters corresponding to the second round of production are expressed as N2—amine content 97%—primary amide 0.14%—iodine value 42g, and the actual values ​​of historical product performance parameters corresponding to the third round of production are expressed as N3—amine content 97%—primary amide 0.14%—iodine value 42g. The actual values ​​of the parameters correspond to N3—amine content 96.2%—primary amide 0.18%—iodine value 48g, the actual values ​​of the historical product performance parameters corresponding to the fourth round of production correspond to N4—amine content 95.7%—primary amide 0.15%—iodine value 46g, and the actual values ​​of the historical product performance parameters corresponding to the fifth round of production correspond to N5—amine content 94.8%—primary amide 0.16%—iodine value 46g. The actual values ​​of the historical product performance parameters of the five rounds of historical production processes obtained above are the actual values ​​of the historical product performance parameters of multiple rounds under abnormal working conditions stored in the database.

[0082] Among them, the predicted values ​​of historical product performance parameters are obtained through a multi-objective optimization model. Specifically, the process parameters of the historical production process (such as X1, X2, X3, ..., X m ) into the equation system Y1-Y n =F1(X1, X2, X3, ..., X m )-F n (X1, X2, X3, ..., X m), that is, the predicted value of each product performance parameter in the historical production process is obtained. When m=1, X1 can be a process parameter (such as production time) that is adjustable under abnormal conditions and is undergoing prediction value calculation; when m=2, X2 can be a process parameter (such as hydrogen pressure) under normal conditions; when m=3, X3 can be a process parameter (such as hydrogenation amount) that is adjustable under abnormal conditions but is not undergoing prediction value calculation. If there is a recommended value for the process parameter that is adjustable but is not undergoing prediction value calculation, the recommended value will be used as the initial value for the process parameter that is adjustable but is not undergoing prediction value calculation. If the remaining adjustable but not undergoing prediction value calculations are If there is no recommended value for the calculated process parameter, the target value is used as the initial value for the process parameter that is adjustable but not being calculated for the predicted value; similarly, X4 can be a non-adjustable process parameter under abnormal conditions (such as water addition amount), X5 can be a process parameter under normal conditions (such as hydrogen flow time), X6 can be an adjustable process parameter under abnormal conditions but not being calculated for the predicted value (such as hydrolysis temperature), X7 can be a process parameter under normal conditions (such as hydrolysis oil-water ratio), X8 can be a process parameter under normal conditions (such as hydrolysis time), and so on. The following process parameters X m The above process parameters can be used for analogy.

[0083] Step S203: The server calculates the product performance parameter prediction deviation value based on the historical product performance parameter actual value and the historical product performance parameter prediction value.

[0084] Specifically, after the server obtains the actual values ​​of the historical product performance parameters of the product (such as the actual value of the primary amine content of the amine is 95%, the actual value of the primary amide is 0.12%, and the actual value of the iodine value is 45g) and the predicted values ​​of the historical product performance parameters of the product (such as the predicted value of the primary amine content of the amine is 97%, the predicted value of the primary amide is 0.15%, and the predicted value of the iodine value is 48g), the historical product performance parameter prediction deviation corresponding to each product performance parameter is calculated according to the equation (∑(actual value of historical product performance parameter Y2-predicted value of historical product performance parameter Y1)) / (number of historical production runs n) For example, if statistics show that there are 5 historical production rounds of amines, and the actual value of the average primary amine content of the amines produced in the 5 rounds is 95.6%, and the predicted value of the average primary amine content is 96%, then the predicted deviation value DY of the product performance parameter corresponding to the primary amine content is 95.6%-96%=-0.4%. For example, if the actual value of the average primary amide of the amines produced in the 5 rounds is 0.125%, and the predicted value of the average primary amide is 0.145%, then the predicted deviation value DY of the historical product performance parameter corresponding to the primary amide is 0.125%-0.145%=-0.02%.

[0085] Step S204: The server inputs the recommended process parameter values ​​into the multi-objective optimization model to obtain the predicted values ​​of the current product performance parameters.

[0086] Specifically, after the server obtains the recommended values ​​of process parameters (such as hydrogen pressure 800Pa, temperature 70℃, and production time 480min), it inputs the recommended values ​​of process parameters into the multi-objective optimization model and obtains the current predicted values ​​of each product performance parameter through a specific algorithm. For example, if the recommended values ​​of process parameters are hydrogen pressure 800Pa, temperature 70℃, and production time 480min, the above process parameters (X1, X2, X3, ..., X4) are also input through the algorithm. m ) into the equation system Y1-Y n =F1(X1, X2, X3, ..., X m )-F n (X1, X2, X3, ..., X m ) to obtain the solution result, that is, the predicted value of each product performance parameter in the current production process (such as the predicted value of primary amine content of output amine is 96%, the predicted value of primary amide is 0.15%), due to the process parameters X1, X2, X3, ..., X m ) The definition and description of the process parameters are the same as those of the aforementioned historical production process and will not be repeated here.

[0087] Step S205: The server obtains the compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value.

[0088] Specifically, in order to improve the prediction accuracy of the current product performance parameter prediction value, a multi-objective optimization model (i.e., a modified compensation prediction model) can be used to compensate the current product performance parameter prediction value. The compensated current product performance parameter prediction value is used to evaluate whether to use the process parameter recommended value for production. Since the steps for calculating each product performance parameter of the product are the same, the following embodiment only describes the primary amine content in the performance parameter as an example. The calculation process of the other product performance parameter prediction values ​​can be obtained with reference to the primary amine content. The current product performance parameter prediction value is compensated using the product performance parameter prediction deviation value corresponding to the primary amine content, so that the compensated current product performance parameter prediction value is more accurate, which can improve the accuracy of the prediction algorithm. For example, if the primary amine content prediction value of the amine obtained through 5 rounds of historical production is 96%, the product performance parameter prediction deviation value of -0.4% is used to compensate the current product performance parameter prediction value of 96%, and the primary amine content prediction value of the compensated amine is 95.6%.

[0089] Optionally, after obtaining the compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, the indicator evaluation value is determined based on the compensated current product performance parameter prediction value;

[0090] Then determine the optimal version of abnormal working conditions among the process parameter versions;

[0091] If the indicator evaluation value is greater than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, production is carried out according to the process parameter value corresponding to the current indicator evaluation value;

[0092] If the indicator evaluation value is less than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, production will be carried out according to the optimal version of the abnormal working condition.

[0093] Specifically, the index evaluation value is used to evaluate the comprehensive closeness of each product performance parameter prediction value in the multiple product performance parameter prediction values ​​corresponding to the process parameters to the corresponding product performance parameter target value, wherein the comprehensive closeness can be understood as comparing the closeness of each product performance prediction value with each product performance target value, and finally taking the multiple comparison results obtained as the comprehensive closeness. The closer each product performance prediction value is to each product performance target value, the higher the corresponding index evaluation value. For example, if the target value corresponding to the primary amine content of the amine (product performance index) is 96.1% and its predicted value is 95.6%, the comparison result between the target value and the predicted value of the primary amine content of the amine is 0.5%; for another example, if the target value corresponding to the primary amide content of the amine (product performance index) is 0.12% and its predicted value is 0.15%, the comparison result between the target value and the predicted value of the primary amide content of the amine is 0.03%. From the above, it can be seen that the closer the comparison result between the target value and the predicted value of the primary amide content of the amine is, the higher the index evaluation value corresponding to the primary amide content of the amine is. The specific steps for the server to determine the index evaluation value based on the predicted value of the current product performance parameter after compensation are mainly to first obtain the deviation value between the current target value of each product and the predicted value of the product performance parameter based on the difference between the target value corresponding to the performance index parameter of each product produced by the current production process and the current predicted value of each product performance parameter after compensation. For example, the target value corresponding to the primary amine content of amine (product performance index) is 96.1%, and the predicted value of the primary amine content of amine obtained after compensation is 95.6%. At this time, the difference between the target value and the predicted value of the primary amine content of amine is 0.5%.

[0094] In this solution, further, there is at least one available process parameter version. Each time the abnormal working condition recommendation algorithm is optimized, the priority level of each product performance parameter is determined based on the target value of each product performance parameter of multiple products obtained by the current multi-objective optimization model, wherein the priority level is used to indicate the degree of influence of the product performance parameter on the quality of the product. The multiple product performance parameters are then divided according to the priority level to obtain multiple optimization sets, and then the indicator evaluation value of each optimization set in the multiple optimization sets is determined based on the optimization value. Finally, the comprehensive index evaluation value of the current multi-objective optimization is determined according to the index evaluation value of each optimization set, and the multiple process parameter versions are sorted according to the comprehensive index evaluation value to obtain the sorting result, that is, the process parameter version corresponding to the largest index evaluation value is obtained as the optimal version of the abnormal working condition. After each optimization of the abnormal working condition recommendation algorithm is completed, the current index evaluation value is compared with the maximum index evaluation value corresponding to the optimal version of the abnormal working condition. If the current index evaluation value is less than the maximum index evaluation value corresponding to the optimal version of the abnormal working condition, production is carried out according to the optimal version of the abnormal working condition (for example, the current index evaluation value is 75 points, and the maximum index evaluation value corresponding to the optimal version of the abnormal working condition is 86 points, then it means that the production result index obtained by producing according to the optimal version of the abnormal working condition achieves a better target). If the index evaluation value is greater than the index evaluation value corresponding to the optimal version of the abnormal working condition (for example, the current index evaluation value is 90 points, and the maximum index evaluation value corresponding to the optimal version of the abnormal working condition is 86 points), then it means that the abnormal working condition recommendation algorithm is optimized successfully. At this time, production is carried out according to the process parameter value corresponding to the current index evaluation value, which can meet the production result index of the product.

[0095] Optionally, after determining the index evaluation value based on the compensated current product performance parameter prediction value, if the index evaluation value obtained in the current production round is less than the preset index evaluation reference value, the current product performance parameter prediction deviation value is recalculated based on the current product performance parameter actual value and the historical product performance parameter actual value;

[0096] Then, according to the current product performance parameter prediction deviation value and the current product performance parameter prediction value, the compensated current product performance parameter prediction value is determined;

[0097] Finally, the abnormal working condition recommendation algorithm is optimized according to the predicted value of the current product performance parameter after compensation until the index evaluation value is controlled within the preset index evaluation reference value range during multiple consecutive productions, and the optimization is completed.

[0098] Specifically, if the determined index evaluation value is less than the preset index evaluation reference value (for example, the index evaluation value is 58 points, but the preset index evaluation reference value range is 60 to 80 points), it means that the performance index of the currently produced product does not meet the standard (for example, one or more of the performance parameters such as the primary amine content, iodine value, and primary amide of the amine are less than the standard range of the corresponding performance parameters). The current product performance parameter prediction deviation value can be recalculated based on the actual value of the performance parameter of each product produced in multiple historical rounds and the actual value of the current product performance parameter. Then, based on the current product performance parameter prediction deviation value and the current product performance parameter prediction value, the compensated current performance parameter prediction value of each product is determined. Finally, the abnormal working condition recommendation algorithm is optimized and trained based on the compensated current performance parameter prediction value of each product to generate a new mathematical model, and the previous recommendation algorithm optimization is repeated until the index evaluation value after multiple trainings is stable within the preset threshold range (for example, the index evaluation value remains between 75 and 80 points, and the preset index evaluation reference value range is 60 to 80 points, so the optimized performance index of the currently produced product meets the reference standard), and the algorithm is no longer optimized. This application evaluates the comprehensive closeness between each of the multiple product performance parameter prediction values ​​corresponding to the process parameters and the corresponding product performance parameter target value, derives a recommended value for the process parameter, and uses the recommended value for the process parameter for production, which can effectively improve the quality of the produced products and the accuracy of the prediction algorithm.

[0099] In the prior art, any process parameter of the production equipment (such as steam pressure, production gas pressure, equipment vacuum, etc.) may cause abnormal working conditions in the production process due to uncontrollable factors such as equipment load, causing the process parameter to exceed the tolerance range preset by the standard value, and when it cannot be restored for a long time, in theory, the process parameter will become a special non-standard working condition parameter. At this time, production still needs to continue. If production continues according to the pre-set standard value, the production result index will not be achieved. However, the present application can obtain the actual value of the historical product performance parameter under abnormal working conditions, combine the actual value of each historical product performance parameter of the product and the predicted value of each historical product performance parameter of the product to calculate the product performance parameter prediction deviation value, determine the process parameter recommendation value through the abnormal working condition recommendation algorithm, and then input the process parameter recommendation value into the multi-objective optimization model to obtain the current product performance parameter prediction value, and then compensate the current product performance parameter prediction value according to the product performance parameter prediction deviation value to obtain the compensated current each product performance parameter prediction value, thereby improving the accuracy of the prediction algorithm, ensuring the product quality of the production process, and meeting the product production result index.

[0100] The above describes in detail the method of the embodiment of the present application. The following provides an apparatus of the embodiment of the present application.

[0101] It can be understood that the multiple devices provided in the embodiments of the present application, such as the recommendation device, include hardware structures, software modules, or a combination of hardware structures and software structures corresponding to executing each function in order to implement the functions in the above method embodiments.

[0102] Those skilled in the art should easily appreciate that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the embodiments of the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different device implementations to implement the aforementioned method embodiments in different usage scenarios, and different implementations of the devices should not be considered to exceed the scope of the embodiments of the present application.

[0103] The embodiments of the present application may divide the device into functional modules. For example, each functional module may be divided according to each function, or two or more functions may be integrated into one functional module. The above-mentioned integrated modules may be implemented in the form of hardware or software functional modules. It should be noted that the division of modules in the embodiments of the present application is schematic and is only a logical functional division. In actual implementation, other division methods may be used.

[0104] For example, in the case of dividing the functional modules of the device in an integrated manner, this application cites several possible processing devices.

[0105] See Figure 3 , Figure 3 : This is a schematic diagram of a device 30 for recommending production process parameters based on abnormal working conditions provided by an embodiment of the present application. The device 30 is a server, or a device in the server, such as a chip, a software module, an integrated circuit, etc. The device 30 is used to implement the aforementioned method for recommending production process parameters based on abnormal working conditions, for example Figure 2 The method for recommending production process parameters based on abnormal working conditions.

[0106] In a possible implementation, the recommendation device 30 may include a determination unit 301 , an acquisition unit 302 , a calculation unit 303 , an input unit 304 , and a compensation unit 305 .

[0107] If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard, the determining unit 301 is used to determine the recommended value of the process parameter according to the abnormal working condition recommendation algorithm;

[0108] The acquisition unit 302 is configured to acquire actual values ​​of historical product performance parameters and predicted values ​​of historical product performance parameters for multiple rounds under the abnormal working conditions stored in a database;

[0109] The calculation unit 303 is configured to calculate a product performance parameter prediction deviation value based on the actual value of the historical product performance parameter and the predicted value of the historical product performance parameter, wherein the predicted value of the historical product performance parameter is obtained by prediction using a multi-objective optimization model;

[0110] The input unit 304 is used to input the recommended process parameter values ​​into the multi-objective optimization model to obtain the predicted values ​​of the current product performance parameters;

[0111] The compensation unit 305 is used to obtain the compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, wherein the compensated current product performance parameter prediction value is used to evaluate whether to use the process parameter recommended value for production.

[0112] In the prior art, any process parameter of the production equipment (such as steam pressure, production gas pressure, equipment vacuum, etc.) may cause abnormal working conditions in the production process due to uncontrollable factors such as equipment load, causing the process parameter to exceed the tolerance range preset by the standard value, and when it cannot be restored for a long time, in theory, the process parameter will become a special non-standard working condition parameter. At this time, production still needs to continue. If production continues according to the pre-set standard value, the production result index will not be achieved. However, the present application can obtain the actual value of the historical product performance parameter under abnormal working conditions, combine the actual value of each historical product performance parameter of the product and the predicted value of each historical product performance parameter of the product to calculate the product performance parameter prediction deviation value, determine the process parameter recommendation value through the abnormal working condition recommendation algorithm, and then input the process parameter recommendation value into the multi-objective optimization model to obtain the current product performance parameter prediction value, and then compensate the current product performance parameter prediction value according to the product performance parameter prediction deviation value to obtain the compensated current each product performance parameter prediction value, thereby improving the accuracy of the prediction algorithm, ensuring the product quality of the production process, and meeting the product production result index.

[0113] Another possible implementation further includes:

[0114] The determining unit 301 is further configured to determine whether an abnormal operating condition occurs in the current production process, wherein the abnormal operating condition is an abnormal operating condition that occurs in the production process during production under normal standard conditions;

[0115] The acquisition unit 302 is further configured to acquire a working condition code of the abnormal working condition if the abnormal working condition occurs;

[0116] A query unit, configured to query in a database whether there is an available process parameter version or whether an indicator evaluation value of the process parameter version meets the standard according to the working condition code;

[0117] If there is no available process parameter version in the production process under abnormal working conditions or the indicator evaluation value of the process parameter version does not meet the standard, the execution unit is used to execute the step of determining the recommended process parameter value according to the abnormal working condition recommendation algorithm.

[0118] In the embodiment of the present application, it is assumed that a process parameter in the current production process is abnormal, for example, the process parameter X of the production process P1 is abnormal. 16 If an exception occurs, obtain the process parameter X 16 The process parameter fault code is 3, and the database code address of the abnormal process parameter is 100 (the database code address of the abnormal process parameter is used to query the abnormal detailed information), then the working condition fault code is 1, 1, 1, 16, 3, 1, 100. Assuming that the benchmark working condition code is 100100100, the abnormal working condition code is (100100100, 1, 1, 1, 16, 3, 1, 100). Then, according to the abnormal working condition code (100100100, 1, 1, 1, 16, 3, 1, 100), the database is queried to see whether there is an available process parameter version or process parameter. Whether the index evaluation value of the version meets the standard. If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard (for example, the available process parameter version used in the historical production process is not queried through the abnormal working condition code, or an available process parameter version is queried, but the index evaluation value corresponding to the available process parameter version is 58 points, which is not within the standard range of 60-80 points), in order to ensure that the process parameter values ​​of the later production equipment can be adjusted in time to adapt to changes in abnormal working conditions, it is necessary to execute the above steps of determining the recommended process parameter values ​​according to the abnormal working condition recommendation algorithm.

[0119] In another possible implementation, after obtaining the compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, the method further includes:

[0120] The determining unit 301 is further configured to determine an index evaluation value based on the compensated current product performance parameter prediction value, wherein the index evaluation value is used to evaluate the comprehensive closeness between each of the multiple product performance parameter prediction values ​​corresponding to the process parameter and the corresponding product performance parameter target value;

[0121] The determining unit 301 is further configured to determine an optimal version for abnormal working conditions among the process parameter versions, wherein the optimal version for abnormal working conditions is the process parameter version with the largest indicator evaluation value;

[0122] If the indicator evaluation value is greater than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, the production unit is configured to perform production according to the process parameter value corresponding to the current indicator evaluation value;

[0123] If the indicator evaluation value is less than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, the production unit is further used to produce according to the optimal version of the abnormal working condition.

[0124] In an embodiment of the present application, there is at least one available process parameter version. Each time the abnormal working condition recommendation algorithm is optimized, the priority level of each product performance parameter is determined based on the target value of each product performance parameter of multiple products obtained by the current multi-objective optimization model, wherein the priority level is used to indicate the degree of influence of the product performance parameter on the quality of the product. The multiple product performance parameters are then divided according to the priority level to obtain multiple optimization sets, and then the index evaluation value of each optimization set in the multiple optimization sets is determined based on the optimization value. Finally, the comprehensive index evaluation value of the current multi-objective optimization is determined based on the index evaluation value of each optimization set. Multiple process parameter versions are sorted according to the comprehensive index evaluation value to obtain the sorting result, that is, the process parameter version corresponding to the largest index evaluation value is obtained as the optimal version for the abnormal working condition. After each optimization of the abnormal working condition recommendation algorithm is completed, the current index evaluation value is compared with the largest index evaluation value corresponding to the optimal version for the abnormal working condition. If the current index evaluation value is less than the largest index evaluation value corresponding to the optimal version for the abnormal working condition, production is carried out according to the optimal version for the abnormal working condition (for example, if the current index evaluation value is 75 points and the largest index evaluation value corresponding to the optimal version for the abnormal working condition is 86 points, it means that the production result index obtained by producing according to the optimal version for the abnormal working condition achieves a better target). If the index evaluation value is greater than the index evaluation value corresponding to the optimal version for the abnormal working condition (for example, if the current index evaluation value is 90 points and the largest index evaluation value corresponding to the optimal version for the abnormal working condition is 86 points), it means that the abnormal working condition recommendation algorithm is optimized successfully. At this time, production is carried out according to the process parameter value corresponding to the current index evaluation value, which can make the adjusted production result index achieve a better target.

[0125] In another possible implementation, the following further comprises:

[0126] If the abnormal operating condition has not appeared in the database, the determining unit 301 is further configured to determine a first reference value of the process parameter under normal operating conditions, wherein the first reference value is a target value of the process parameter in the benchmark operating condition scorecard under normal operating conditions;

[0127] The determining unit 301 is further configured to determine a second reference value of the adjustable process parameter;

[0128] If the adjustable process parameter has the process parameter recommended value, the second reference value is the process parameter recommended value of the adjustable process parameter;

[0129] If the process parameter recommended value does not exist for the adjustable process parameter, the second reference value is the target value of the adjustable process parameter, wherein the target value of the adjustable process parameter is the target value of the adjustable process parameter in the benchmark operating condition scorecard;

[0130] The determining unit 301 is further configured to determine an abnormality prediction value of an unadjustable process parameter;

[0131] The input unit 304 is further configured to input the process parameter value, the first reference value, the second reference value, and the abnormality prediction value of the non-adjustable process parameter into the multi-objective optimization model to obtain multiple current product performance index prediction values.

[0132] In an embodiment of the present application, if no abnormal operating conditions have occurred in the database, the types of process parameters may include process parameters that are adjustable under abnormal operating conditions and are undergoing predictive value calculation, process parameters under normal operating conditions, process parameters that are adjustable under abnormal operating conditions but are not undergoing predictive value calculation, process parameters that are not adjustable under abnormal operating conditions, and when abnormal operating conditions occur, the recommended value of the process parameter that is undergoing predictive value calculation and has completed the recommended value algorithm or the default value (target value of the benchmark operating condition scorecard) of the process parameter that is undergoing predictive value calculation and has not completed the recommended value algorithm. At this time, the process parameter that is adjustable but is not undergoing predictive value calculation is judged, and if a recommended value exists, the recommended value is used; if no recommended value exists, the target value is used.

[0133] Among them, when calculating the product performance index prediction values ​​corresponding to multiple process parameters within the adjustment range of a certain adjustable process parameter that is currently undergoing prediction value calculation, it is necessary to first determine the calculation initial values ​​of the remaining process parameters except for the adjustable process parameter that is currently undergoing prediction value calculation. It can be understood that if there is a recommended value for the adjustable process parameter that is not currently undergoing prediction value calculation, the recommended value is used as the calculation initial value of the adjustable process parameter that is not currently undergoing prediction value calculation; if there is no recommended value for the remaining adjustable process parameter that is not currently undergoing prediction value calculation, the target value is used as the calculation initial value of the adjustable process parameter that is not currently undergoing prediction value calculation; the process parameters under normal conditions use the target value in the benchmark operating condition scorecard as the calculation initial value; the process parameters that cannot be adjusted under abnormal conditions use the abnormal prediction value as the calculation initial value. This method takes into account that the product performance index prediction value is affected by all the process parameters of the product. When calculating the prediction value of the recommended value of the process parameter, the values ​​of the remaining process parameters are also pre-set, which can make the prediction value more accurate.

[0134] In another possible implementation, after determining the index evaluation value according to the compensated current product performance parameter prediction value, the method further includes:

[0135] If the index evaluation value obtained in the current round of production is less than the preset index evaluation reference value, the current product performance parameter prediction deviation value is recalculated based on the current product performance parameter actual value and the historical product performance parameter actual value;

[0136] Determining the compensated current product performance parameter prediction value according to the current product performance parameter prediction deviation value and the current product performance parameter prediction value;

[0137] The abnormal operating condition recommendation algorithm is optimized according to the compensated current product performance parameter prediction value until the index evaluation value is controlled within the preset index evaluation reference value range during multiple consecutive productions, and the optimization is terminated.

[0138] In an embodiment of the present application, if the determined index evaluation value is less than the preset index evaluation reference value, it means that the product performance indicators of multiple rounds of historical production do not meet the standards (for example, one or more of the performance parameters such as the primary amine content, iodine value, and primary amide of the amine are less than the standard range interval of the corresponding performance parameters). The abnormal working condition recommendation algorithm can be optimized and trained based on the actual value of each product performance parameter currently produced and its corresponding product performance parameter prediction value, and then based on the current predicted value of each product performance parameter after compensation, a new mathematical model is generated, and the previous recommendation algorithm optimization is repeated until the index evaluation value after multiple trainings is stable within the preset threshold range (for example, the index evaluation value remains between 75 and 80 points, and the preset index evaluation reference value range is 60 to 80 points, so the optimized current product performance index meets the reference standard), and the algorithm is no longer optimized. This solution evaluates the comprehensive closeness of each product performance parameter prediction value in the multiple product performance parameter prediction values ​​corresponding to the process parameters to the corresponding product performance parameter target value, obtains a process parameter recommendation value, and uses the process parameter recommendation value for production, which can effectively improve the quality of the produced products and the accuracy of the prediction algorithm.

[0139] See Figure 4 , Figure 4: This is a structural diagram of a device 40 for recommending production process parameters based on abnormal working conditions provided in an embodiment of the present application. The recommendation device 40 can be a server (e.g., one or more servers, etc.), or a component inside the server (e.g., a chip, a software module, or a hardware module, etc.). The recommendation device 40 may include at least one processor 401. Optionally, it may also include at least one memory 403. Further optionally, the recommendation device 40 may also include a communication interface 402. Further optionally, it may also include a bus 404, wherein the processor 401, the communication interface 402, and the memory 403 are connected via the bus 404.

[0140] Among them, the processor 401 is a module that performs arithmetic operations and / or logical operations, and can specifically be a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor (MPU), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), a complex programmable logic device (CPLD), a coprocessor (assisting the central processing unit to complete corresponding processing and applications), a microcontroller unit (MCU), and other processing modules, or a combination of multiple thereof.

[0141] The communication interface 402 can be used to provide information input or output for the at least one processor. And / or, the communication interface 402 can be used to receive data sent externally and / or send data externally. It can be a wired link interface such as an Ethernet cable, or a wireless link interface (Wi-Fi, Bluetooth, general wireless transmission, vehicle-mounted short-range communication technology, and other short-range wireless communication technologies). Optionally, the communication interface 402 can also include a transmitter (such as a radio frequency transmitter, antenna, etc.) or a receiver coupled to the interface.

[0142] Memory 403 is used to provide storage space for storing data such as the operating system and computer programs. Memory 403 can be one or a combination of random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or compact disc read-only memory (CD-ROM).

[0143] At least one processor 401 in the recommendation device 40 is configured to execute the aforementioned method, for example Figure 2 The method described in the embodiment.

[0144] Optionally, processor 401 may be a processor specifically configured to execute these methods (referred to as a dedicated processor for ease of distinction), or may be a processor that executes these methods by invoking a computer program, such as a general-purpose processor. Optionally, the at least one processor may include both a dedicated processor and a general-purpose processor. Optionally, when the recommendation device 40 includes at least one processor 401, the computer program may be stored in memory 403.

[0145] Optionally, the at least one processor 401 in the recommendation device 40 is configured to execute a calling computer instruction to perform the following operations:

[0146] If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard, the process parameter recommended value is determined according to the abnormal working condition recommendation algorithm;

[0147] Obtaining actual values ​​of historical product performance parameters and predicted values ​​of historical product performance parameters for multiple rounds under the abnormal operating conditions stored in a database;

[0148] Calculating a product performance parameter prediction deviation value based on the actual value of the historical product performance parameter and the predicted value of the historical product performance parameter, wherein the predicted value of the historical product performance parameter is obtained by prediction using a multi-objective optimization model;

[0149] Inputting the recommended process parameter values ​​into the multi-objective optimization model to obtain predicted values ​​of current product performance parameters;

[0150] Based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, the compensated current product performance parameter prediction value is obtained, wherein the compensated current product performance parameter prediction value is used to evaluate whether to use the process parameter recommended value for production.

[0151] In the prior art, any process parameter of the production equipment (such as steam pressure, production gas pressure, equipment vacuum, etc.) may cause abnormal working conditions in the production process due to uncontrollable factors such as equipment load, causing the process parameter to exceed the tolerance range preset by the standard value, and when it cannot be restored for a long time, in theory, the process parameter will become a special non-standard working condition parameter. At this time, production still needs to continue. If production continues according to the pre-set standard value, the production result index will not be achieved. However, the present application can obtain the actual value of the historical product performance parameter under abnormal working conditions, combine the actual value of each historical product performance parameter of the product and the predicted value of each historical product performance parameter of the product to calculate the product performance parameter prediction deviation value, determine the process parameter recommendation value through the abnormal working condition recommendation algorithm, and then input the process parameter recommendation value into the multi-objective optimization model to obtain the current product performance parameter prediction value, and then compensate the current product performance parameter prediction value according to the product performance parameter prediction deviation value to obtain the compensated current each product performance parameter prediction value, thereby improving the accuracy of the prediction algorithm, ensuring the product quality of the production process, and meeting the product production result index.

[0152] Optionally, the processor 401 is further configured to:

[0153] Determine whether an abnormal operating condition occurs in the current production process, wherein the abnormal operating condition is an abnormal operating condition that occurs in the production process when production is under normal standard conditions;

[0154] If the abnormal working condition occurs, obtaining the working condition code of the abnormal working condition;

[0155] According to the working condition code, query in the database whether there is an available process parameter version or whether the index evaluation value of the process parameter version meets the standard;

[0156] If there is no available process parameter version in the production process under abnormal working conditions or the indicator evaluation value of the process parameter version does not meet the standard, the step of determining the recommended value of the process parameter according to the abnormal working condition recommendation algorithm is executed.

[0157] In the embodiment of the present application, it is assumed that a process parameter in the current production process is abnormal, for example, the process parameter X of the production process P1 is abnormal. 16 If an exception occurs, obtain the process parameter X 16The process parameter fault code is 3, and the database code address of the abnormal process parameter is 100 (the database code address of the abnormal process parameter is used to query the abnormal detailed information), then the working condition fault code is 1, 1, 1, 16, 3, 1, 100. Assuming that the benchmark working condition code is 100100100, the abnormal working condition code is (100100100, 1, 1, 1, 16, 3, 1, 100). Then, according to the abnormal working condition code (100100100, 1, 1, 1, 16, 3, 1, 100), the database is queried to see whether there is an available process parameter version or process parameter. Whether the index evaluation value of the version meets the standard. If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard (for example, the available process parameter version used in the historical production process is not queried through the abnormal working condition code, or an available process parameter version is queried, but the index evaluation value corresponding to the available process parameter version is 58 points, which is not within the standard range of 60-80 points), in order to ensure that the process parameter values ​​of the later production equipment can be adjusted in time to adapt to changes in abnormal working conditions, it is necessary to execute the above steps of determining the recommended process parameter values ​​according to the abnormal working condition recommendation algorithm.

[0158] Optionally, the processor 401 is further configured to:

[0159] Determining an index evaluation value based on the compensated current product performance parameter prediction value, wherein the index evaluation value is used to evaluate the comprehensive closeness between each of the multiple product performance parameter prediction values ​​corresponding to the process parameter and the corresponding product performance parameter target value;

[0160] Determining an optimal version of abnormal operating conditions among the process parameter versions, wherein the optimal version of abnormal operating conditions is the process parameter version with the largest indicator evaluation value;

[0161] If the indicator evaluation value is greater than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, production is carried out according to the process parameter value corresponding to the current indicator evaluation value;

[0162] If the indicator evaluation value is less than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, production is performed according to the optimal version of the abnormal working condition.

[0163] In an embodiment of the present application, there is at least one available process parameter version. Each time the abnormal working condition recommendation algorithm is optimized, the priority level of each product performance parameter is determined based on the target value of each product performance parameter of multiple products obtained by the current multi-objective optimization model, wherein the priority level is used to indicate the degree of influence of the product performance parameter on the quality of the product. The multiple product performance parameters are then divided according to the priority level to obtain multiple optimization sets, and then the index evaluation value of each optimization set in the multiple optimization sets is determined based on the optimization value. Finally, the comprehensive index evaluation value of the current multi-objective optimization is determined based on the index evaluation value of each optimization set. Multiple process parameter versions are sorted according to the comprehensive index evaluation value to obtain the sorting result, that is, the process parameter version corresponding to the largest index evaluation value is obtained as the optimal version for the abnormal working condition. After each optimization of the abnormal working condition recommendation algorithm is completed, the current index evaluation value is compared with the largest index evaluation value corresponding to the optimal version for the abnormal working condition. If the current index evaluation value is less than the largest index evaluation value corresponding to the optimal version for the abnormal working condition, production is carried out according to the optimal version for the abnormal working condition (for example, if the current index evaluation value is 75 points and the largest index evaluation value corresponding to the optimal version for the abnormal working condition is 86 points, it means that the production result index obtained by producing according to the optimal version for the abnormal working condition achieves a better target). If the index evaluation value is greater than the index evaluation value corresponding to the optimal version for the abnormal working condition (for example, if the current index evaluation value is 90 points and the largest index evaluation value corresponding to the optimal version for the abnormal working condition is 86 points), it means that the abnormal working condition recommendation algorithm is optimized successfully. At this time, production is carried out according to the process parameter value corresponding to the current index evaluation value, which can make the adjusted production result index achieve a better target.

[0164] Optionally, the processor 401 is further configured to:

[0165] If the abnormal operating condition has not appeared in the database, determining a first reference value of the process parameter under normal operating conditions, wherein the first reference value is a target value of the process parameter in the benchmark operating condition scorecard under the normal operating conditions;

[0166] determining a second reference value of the adjustable process parameter;

[0167] If the adjustable process parameter has the process parameter recommended value, the second reference value is the process parameter recommended value of the adjustable process parameter;

[0168] If the process parameter recommended value does not exist for the adjustable process parameter, the second reference value is the target value of the adjustable process parameter, wherein the target value of the adjustable process parameter is the target value of the adjustable process parameter in the benchmark operating condition scorecard;

[0169] Determine the abnormal prediction value of non-adjustable process parameters;

[0170] The process parameter recommended value, the first reference value, the second reference value and the abnormal prediction value of the non-adjustable process parameter are respectively input into the multi-objective optimization model to obtain multiple current product performance index prediction values.

[0171] Specifically, if no abnormal operating conditions have occurred in the database, the types of process parameters may include process parameters that are adjustable under abnormal operating conditions and are undergoing predictive value calculations, process parameters under normal operating conditions, process parameters that are adjustable under abnormal operating conditions but are not undergoing predictive value calculations, process parameters that are not adjustable under abnormal operating conditions, and recommended values ​​for process parameters that are undergoing predictive value calculations for which the recommended value algorithm has been completed or default values ​​(target values ​​of the benchmark operating condition scorecard) for process parameters that are undergoing predictive value calculations for which the recommended value algorithm has not been completed when abnormal operating conditions occur. At this time, a judgment is made on the process parameter that is adjustable but is not undergoing predictive value calculations. If a recommended value exists, the recommended value is used; if no recommended value exists, the target value is used.

[0172] Among them, when calculating the product performance index prediction values ​​corresponding to multiple process parameters within the adjustment range of a certain adjustable process parameter that is currently undergoing prediction value calculation, it is necessary to first determine the calculation initial values ​​of the remaining process parameters except for the adjustable process parameter that is currently undergoing prediction value calculation. It can be understood that if there is a recommended value for the adjustable process parameter that is not currently undergoing prediction value calculation, the recommended value is used as the calculation initial value of the adjustable process parameter that is not currently undergoing prediction value calculation; if there is no recommended value for the remaining adjustable process parameter that is not currently undergoing prediction value calculation, the target value is used as the calculation initial value of the adjustable process parameter that is not currently undergoing prediction value calculation; the process parameters under normal conditions use the target value in the benchmark operating condition scorecard as the calculation initial value; the process parameters that cannot be adjusted under abnormal conditions use the abnormal prediction value as the calculation initial value. This method takes into account that the product performance index prediction value is affected by all the process parameters of the product. When calculating the prediction value of the recommended value of the process parameter, the values ​​of the remaining process parameters are also pre-set, which can make the prediction value more accurate.

[0173] Optionally, the processor 401 is further configured to:

[0174] If the index evaluation value obtained in the current round of production is less than the preset index evaluation reference value, the current product performance parameter prediction deviation value is recalculated based on the current product performance parameter actual value and the historical product performance parameter actual value;

[0175] Determining the compensated current product performance parameter prediction value according to the current product performance parameter prediction deviation value and the current product performance parameter prediction value;

[0176] The abnormal operating condition recommendation algorithm is optimized according to the compensated current product performance parameter prediction value until the index evaluation value is controlled within the preset index evaluation reference value range during multiple consecutive productions, and the optimization is terminated.

[0177] In an embodiment of the present application, if the determined index evaluation value is less than the preset index evaluation reference value, it means that the product performance indicators of multiple rounds of historical production do not meet the standards (for example, one or more of the performance parameters such as the primary amine content, iodine value, and primary amide of the amine are less than the standard range interval of the corresponding performance parameters). The abnormal working condition recommendation algorithm can be optimized and trained based on the actual value of each product performance parameter currently produced and its corresponding product performance parameter prediction value, and then based on the current predicted value of each product performance parameter after compensation, a new mathematical model is generated, and the previous recommendation algorithm optimization is repeated until the index evaluation value after multiple trainings is stable within the preset threshold range (for example, the index evaluation value remains between 75 and 80 points, and the preset index evaluation reference value range is 60 to 80 points, so the optimized current product performance index meets the reference standard), and the algorithm is no longer optimized. This solution evaluates the comprehensive closeness of each product performance parameter prediction value in the multiple product performance parameter prediction values ​​corresponding to the process parameters to the corresponding product performance parameter target value, obtains a process parameter recommendation value, and uses the process parameter recommendation value for production, which can effectively improve the quality of the produced products and the accuracy of the prediction algorithm.

[0178] The present application also provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed on at least one processor, the aforementioned method for recommending production process parameters based on abnormal working conditions is implemented, for example Figure 2 The method described.

[0179] The present application also provides a computer program product, which includes computer instructions, and when executed by the recommended device, implements the aforementioned method for recommending production process parameters based on abnormal working conditions, for example Figure 2 The method described.

[0180] In the embodiments of this application, words such as "for example" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described in this application as "for example" or "for example" should not be interpreted as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "for example" or "for example" is intended to present the relevant concepts in a concrete way.

[0181] The “at least one” mentioned in the embodiments of this application refers to one or more, and “plurality” refers to two or more. “At least one of the following items” or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, (a and b), (a and c), (b and c), or (a and b and c), where a, b, c can be single or multiple. “And / or” describes the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character “ / ” generally indicates that the previous and next associated objects are in an “or” relationship.

[0182] Furthermore, unless otherwise specified, ordinal numbers such as "first" and "second" in the embodiments of this application are used to distinguish multiple objects and are not used to limit the order, timing, priority, or importance of multiple objects. For example, the first device and the second device are only for ease of description and do not indicate differences in structure, importance, etc. between the first and second devices. In some embodiments, the first device and the second device can also be the same device.

[0183] In the above embodiments, the term "when" can be interpreted to mean "if...", "after...", "in response to determining...", or "in response to detecting...", depending on the context. The above are merely optional embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the concepts and principles of the present application shall be included in the scope of protection of the present application.

[0184] Those skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware, or by a program to instruct the relevant hardware, and the program may be stored in a computer-readable storage medium, which may be a read-only memory, a disk, or an optical disk, etc.

[0185] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present application, and such modifications or substitutions should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for recommending production process parameters based on abnormal working conditions, characterized in that: The method comprises: Determine whether an abnormal operating condition occurs in the current production process, wherein the abnormal operating condition is an abnormal operating condition that occurs in the production process when production is under normal standard conditions; If the abnormal working condition occurs, and there is no available process parameter version in the production process under the abnormal working condition or the index evaluation value of the process parameter version does not meet the standard, the process parameter recommended value is determined according to the abnormal working condition recommendation algorithm; Obtaining actual values ​​of historical product performance parameters and predicted values ​​of historical product performance parameters for multiple rounds under the abnormal operating conditions stored in a database; Calculating a product performance parameter prediction deviation value based on the actual value of the historical product performance parameter and the predicted value of the historical product performance parameter, wherein the predicted value of the historical product performance parameter is obtained by prediction using a multi-objective optimization model; Inputting the recommended process parameter values ​​into the multi-objective optimization model to obtain predicted values ​​of current product performance parameters; Obtaining a compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, wherein the compensated current product performance parameter prediction value is used to evaluate whether to use the process parameter recommended value for production; If the abnormal operating condition has not appeared in the database, determining a first reference value of the process parameter under normal operating conditions, wherein the first reference value is a target value of the process parameter in the benchmark operating condition scorecard under the normal operating conditions; determining a second reference value of the adjustable process parameter; If the adjustable process parameter has the process parameter recommended value, the second reference value is the process parameter recommended value of the adjustable process parameter; If the process parameter recommended value does not exist for the adjustable process parameter, the second reference value is the target value of the adjustable process parameter, wherein the target value of the adjustable process parameter is the target value of the adjustable process parameter in the benchmark operating condition scorecard; Determine the abnormal prediction value of non-adjustable process parameters; The process parameter recommended value, the first reference value, the second reference value and the abnormal prediction value of the non-adjustable process parameter are respectively input into the multi-objective optimization model to obtain multiple current product performance index prediction values.

2. The method according to claim 1, characterized in that If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard, before determining the recommended process parameter value according to the abnormal working condition recommendation algorithm, the following steps are also included: If the abnormal working condition occurs, obtaining the working condition code of the abnormal working condition; querying in a database whether there is an available process parameter version or whether the index evaluation value of the process parameter version meets the standard according to the working condition code; if there is no available process parameter version in the production process under the abnormal working condition or the index evaluation value of the process parameter version does not meet the standard, executing the step of determining the recommended process parameter value according to the abnormal working condition recommendation algorithm; in, Process parameters include equipment control parameters, production steps, production time, and feed amount.

3. The method according to claim 1, characterized in that After obtaining the compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, the method further includes: Determining an index evaluation value based on the compensated current product performance parameter prediction value, wherein the index evaluation value is used to evaluate the comprehensive closeness between each of the multiple product performance parameter prediction values ​​corresponding to the process parameter and the corresponding product performance parameter target value; Determining an optimal version of abnormal operating conditions among the process parameter versions, wherein the optimal version of abnormal operating conditions is the process parameter version with the largest indicator evaluation value; If the indicator evaluation value is greater than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, production is carried out according to the process parameter value corresponding to the current indicator evaluation value; If the indicator evaluation value is less than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, production is performed according to the optimal version of the abnormal working condition.

4. The method according to claim 3, characterized in that After determining the index evaluation value according to the compensated current product performance parameter prediction value, the method further includes: If the index evaluation value obtained in the current round of production is less than the preset index evaluation reference value, the current product performance parameter prediction deviation value is recalculated based on the current product performance parameter actual value and the historical product performance parameter actual value; Determining the compensated current product performance parameter prediction value according to the current product performance parameter prediction deviation value and the current product performance parameter prediction value; The abnormal operating condition recommendation algorithm is optimized according to the compensated current product performance parameter prediction value until the index evaluation value is controlled within the preset index evaluation reference value range during multiple consecutive productions, and the optimization is terminated.

5. A device for recommending production process parameters based on abnormal working conditions, characterized in that: It includes a determination unit, an acquisition unit, a calculation unit, an input unit and a compensation unit, wherein: The determining unit is configured to determine whether an abnormal operating condition occurs in the current production process, wherein the abnormal operating condition is an abnormal operating condition that occurs in the production process during production under normal standard conditions; if the abnormal operating condition occurs, and no process parameter version is available in the production process under the abnormal operating condition or the indicator evaluation value of the process parameter version does not meet the standard, determining a recommended value for the process parameter according to an abnormal operating condition recommendation algorithm; The acquisition unit is used to acquire the actual values ​​of historical product performance parameters and the predicted values ​​of historical product performance parameters for multiple rounds under the abnormal working conditions stored in the database; The calculation unit is used to calculate the product performance parameter prediction deviation value based on the historical product performance parameter actual value and the historical product performance parameter prediction value, wherein the historical product performance parameter prediction value is obtained by prediction using a multi-objective optimization model; The input unit is used to input the recommended process parameter values ​​into the multi-objective optimization model to obtain the predicted values ​​of the current product performance parameters; The compensation unit is configured to obtain a compensated current product performance parameter prediction value based on the product performance parameter prediction deviation value and the current product performance parameter prediction value, wherein the compensated current product performance parameter prediction value is used to evaluate whether to use the process parameter recommended value for production; The determining unit is further configured to determine a first reference value of the process parameter under normal operating conditions if the abnormal operating condition has not occurred in the database, wherein the first reference value is a target value of the process parameter in the benchmark operating condition scorecard under the normal operating conditions; The determining unit is further configured to determine a second reference value of the adjustable process parameter; If the adjustable process parameter has the process parameter recommended value, the second reference value is the process parameter recommended value of the adjustable process parameter; If the process parameter recommended value does not exist for the adjustable process parameter, the second reference value is the target value of the adjustable process parameter, wherein the target value of the adjustable process parameter is the target value of the adjustable process parameter in the benchmark operating condition scorecard; The determining unit is further configured to determine an abnormality prediction value of an unadjustable process parameter; The input unit is also used to input the recommended value of the process parameter, the first reference value, the second reference value and the abnormal prediction value of the non-adjustable process parameter into the multi-objective optimization model respectively to obtain multiple current product performance index prediction values.

6. The device according to claim 5, characterized in that Also includes: The acquisition unit is further configured to acquire a working condition code of the abnormal working condition if the abnormal working condition occurs; A query unit, configured to query in a database whether there is an available process parameter version or whether an indicator evaluation value of the process parameter version meets the standard according to the working condition code; If there is no available process parameter version in the production process under abnormal working conditions or the index evaluation value of the process parameter version does not meet the standard, the execution unit is used to execute the step of determining the recommended process parameter value according to the abnormal working condition recommendation algorithm; in, Process parameters include equipment control parameters, production steps, production time, and feed amount.

7. The device according to claim 5 or 6, characterized in that Also includes: The determining unit is further configured to determine an index evaluation value based on the compensated current product performance parameter prediction value, wherein the index evaluation value is used to evaluate the comprehensive closeness between each of the multiple product performance parameter prediction values ​​corresponding to the process parameter and the corresponding product performance parameter target value; The determining unit is further configured to determine an optimal version for abnormal working conditions among the process parameter versions, wherein the optimal version for abnormal working conditions is the process parameter version with the largest indicator evaluation value; If the indicator evaluation value is greater than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, the production unit is configured to perform production according to the process parameter value corresponding to the current indicator evaluation value; If the indicator evaluation value is less than the indicator evaluation value corresponding to the optimal version of the abnormal working condition, the production unit is further used to produce according to the optimal version of the abnormal working condition.

8. A device for recommending production process parameters based on abnormal working conditions, characterized in that: The device includes a processor and a memory, wherein the memory is used to store computer instructions, and the processor is used to call the computer instructions to implement the method according to any one of claims 1 to 4.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed on at least one processor, the method according to any one of claims 1 to 4 is implemented.

10. A computer program product, characterized in that The computer program product is used to implement the method according to any one of claims 1 to 4.

Images