Manufacturing process optimization method and system of high vacuum furnace

Abstract

The invention provides a manufacturing process optimization method and system for a high-vacuum furnace, and relates to the technical field of data processing, and the method comprises the steps: obtaining the part performance demand feature data of a single part of the high-vacuum furnace according to the analysis of application scene information, carrying out the importance recognition of the single part, and constructing a part process optimization constraint operator, and generating processing scene source data according to the design scheme and the performance demand characteristics, performing process optimization analysis in combination with a multi-source heterogeneous manufacturing process optimization algorithm and a component process optimization constraint operator, and generating a component processing process scheme of the high vacuum furnace to perform component processing. The technical problems that in the prior art, manufacturing process optimization of a high-vacuum furnace depends on multi-round pilot production and assembly operation of single parts of the high-vacuum furnace, time is consumed for manufacturing process optimization, and cost is high are solved. The technical effects that the optimization difficulty of the high-vacuum furnace manufacturing process is reduced, the part machining scheme and production requirement adaptation degree is improved, and the manufacturing process optimization scheme generation efficiency and the scheme scientization degree are improved are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a manufacturing process optimization method and system for a high vacuum furnace. Background Art

[0002] At present, the manufacturing process optimization of high vacuum furnaces still relies on multiple rounds of trial production and assembly operations of high vacuum furnace components. However, each round of trial production and assembly operations in this practice takes a lot of time and money, and may lead to an extension of the production cycle and an increase in production costs.

[0003] In this case, it becomes particularly urgent to seek optimization of the manufacturing process in order to reduce manufacturing costs, improve production efficiency, and shorten the time to bring products to market.

[0004] In summary, the manufacturing process optimization of the high vacuum furnace in the prior art relies on multiple rounds of trial production and assembly operations of high vacuum furnace components, which results in the technical problem that the manufacturing process optimization is time-consuming and costly. Summary of the invention

[0005] The present application provides a method and system for optimizing the manufacturing process of a high vacuum furnace, which is used to solve the technical problem that the manufacturing process optimization of the high vacuum furnace in the prior art relies on multiple rounds of trial production and assembly operation of high vacuum furnace components, and the manufacturing process optimization is time-consuming and costly.

[0006] In view of the above problems, the present application provides a method and system for optimizing the manufacturing process of a high vacuum furnace.

[0007] In a first aspect of the present application, a method for optimizing the manufacturing process of a high vacuum furnace is provided, the method comprising: collecting design parameters and application scenario parameters of the high vacuum furnace to be produced, and obtaining a target design scheme and target application scenario information of the high vacuum furnace to be produced; performing disassembly based on the target design scheme, and obtaining Y component design schemes corresponding to Y components of the high vacuum furnace to be produced, wherein Y is the total number of components of the high vacuum furnace to be produced; performing performance requirement analysis on the Y components based on the target application scenario information, and obtaining performance requirement feature data of the Y components; performing importance identification on the Y components, and constructing Y component importance identification results based on the component importance identification results; A component process optimization constraint operator, wherein each component process optimization constraint operator includes a component process optimization cost constraint, a component process optimization step constraint, and a component process optimization iteration constraint; data fusion is performed based on the Y component design schemes and the Y component performance requirement characteristic data to obtain Y component processing scene source data; based on a multi-source heterogeneous manufacturing process optimization algorithm, component processing process optimization analysis is performed on the Y components respectively according to the Y component processing scene source data and the Y component process optimization constraint operators to determine the component processing process scheme of the high vacuum furnace to be produced; and component processing of the high vacuum furnace to be produced is performed according to the component processing process scheme.

[0008] According to a second aspect of the present application, a manufacturing process optimization system for a high vacuum furnace is provided, the system comprising: a solution output unit, for collecting design parameters and application scenario parameters of the high vacuum furnace to be produced, and obtaining a target design scheme and target application scenario information of the high vacuum furnace to be produced; a solution disassembly execution unit, for disassembling based on the target design scheme, and obtaining Y component design schemes corresponding to Y components of the high vacuum furnace to be produced, wherein Y is the total number of components of the high vacuum furnace to be produced; a demand analysis execution unit, for performing performance demand analysis on the Y components based on the target application scenario information, and obtaining performance demand feature data of the Y components; an optimization constraint construction unit, for identifying the importance of the Y components, and determining the optimal constraints based on the component importance identification results; The invention relates to a method for manufacturing a high vacuum furnace of the present invention, wherein Y component process optimization constraint operators are constructed, wherein each component process optimization constraint operator includes a component process optimization cost constraint, a component process optimization step constraint and a component process optimization iteration constraint; a data fusion execution unit is used to perform data fusion based on the Y component design schemes and the Y component performance requirement characteristic data to obtain Y component processing scene source data; an optimization analysis execution unit is used to perform component processing process optimization analysis on the Y components respectively according to the Y component processing scene source data and the Y component process optimization constraint operators based on a multi-source heterogeneous manufacturing process optimization algorithm to determine the component processing process scheme of the high vacuum furnace to be produced; and a component processing execution unit is used to perform component processing of the high vacuum furnace to be produced according to the component processing process scheme.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] The method provided in the embodiment of the present application obtains the target design scheme and target application scenario information of the high vacuum furnace to be produced by collecting the design parameters and application scenario parameters of the high vacuum furnace to be produced; performs disassembly based on the target design scheme to obtain Y component design schemes corresponding to Y components of the high vacuum furnace to be produced, wherein Y is the total number of components of the high vacuum furnace to be produced; performs performance requirement analysis on the Y components based on the target application scenario information to obtain performance requirement feature data of the Y components; performs importance identification according to the Y components, and constructs Y component process optimization constraint operators based on the component importance identification results, wherein In the method, each component process optimization constraint operator includes component process optimization cost constraint, component process optimization step constraint and component process optimization iteration constraint; data fusion is performed based on the Y component design schemes and the Y component performance requirement characteristic data to obtain Y component processing scene source data; based on the multi-source heterogeneous manufacturing process optimization algorithm, component processing process optimization analysis is performed on the Y components according to the Y component processing scene source data and the Y component process optimization constraint operators to determine the component processing process scheme of the high vacuum furnace to be produced; and component processing of the high vacuum furnace to be produced is performed according to the component processing process scheme. The technical effect of reducing the difficulty of optimizing the manufacturing process of the high vacuum furnace, improving the adaptability of the component processing scheme and production requirements, and improving the efficiency of generating the manufacturing process optimization scheme and the scientific degree of the scheme is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic diagram of a manufacturing process optimization method for a high vacuum furnace provided in this application;

[0012] Figure 2 A schematic diagram of a process for obtaining component performance requirement characteristic data in a manufacturing process optimization method for a high vacuum furnace provided in the present application;

[0013] Figure 3 A schematic diagram of the structure of a high vacuum furnace manufacturing process optimization system provided in this application.

[0014] Explanation of the accompanying drawings: solution output unit 1, solution disassembly execution unit 2, requirement analysis execution unit 3, optimization constraint construction unit 4, data fusion execution unit 5, optimization analysis execution unit 6, component processing execution unit 7. DETAILED DESCRIPTION

[0015] The present application provides a method and system for optimizing the manufacturing process of a high vacuum furnace, which is used to solve the technical problem that the manufacturing process optimization of a high vacuum furnace in the prior art depends on multiple rounds of trial production and assembly of high vacuum furnace components, and the manufacturing process optimization is time-consuming and costly. The method achieves the technical effect of reducing the difficulty of optimizing the manufacturing process of a high vacuum furnace, improving the adaptability of component processing plans and production requirements, and improving the efficiency of generating manufacturing process optimization plans and the scientificity of the plans.

[0016] The acquisition, storage, use, and processing of data in the technical solution of the present invention are in compliance with relevant regulations.

[0017] Below, the technical solutions in the present invention will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments described herein. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. It should also be noted that, for the convenience of description, only the parts related to the present invention are shown in the accompanying drawings, rather than all of them.

[0018] Embodiment 1

[0019] like Figure 1 As shown, the present application provides a method for optimizing the manufacturing process of a high vacuum furnace, the method comprising:

[0020] A100: collecting design parameters and application scenario parameters of the high vacuum furnace to be produced, and obtaining target design scheme and target application scenario information of the high vacuum furnace to be produced;

[0021] Specifically, it should be understood that the high vacuum furnace is a device used for material processing or experiments in a high vacuum environment, which is composed of multiple components and these components are connected together through connectors to form a complete device.

[0022] At the same time, in the modern production of high vacuum furnaces, in order to reduce component loss and maintenance costs, components are usually produced based on standardized design parameters and standard design schemes to facilitate low-cost replacement of damaged components.

[0023] Based on this, this embodiment interactively obtains multiple sets of standard design parameters-standard design schemes for multiple historical standardized components to construct a standard scheme retrieval map in which the design parameter is the first attribute, the design scheme is the second attribute, and the corresponding standard design parameters and standard design schemes are used as the first attribute value and the second attribute value, respectively.

[0024] In this embodiment, the high vacuum furnace to be produced is an unspecified high vacuum furnace including Y components, and Y design parameters of the Y components of the high vacuum furnace to be produced are collected and obtained, and then the Y design parameters are traversed through the standard scheme retrieval map one by one to compare the same standard design parameters and call the corresponding standard design schemes, and finally Y (standard) component design schemes corresponding to the Y design parameters are obtained.

[0025] The Y (standard) component design solutions combined with the connection relationship of the Y components in the vacuum blast furnace to be produced constitute the target design solution.

[0026] Furthermore, the application scenario parameters of the usable environmental conditions of the high vacuum furnace to be produced set in the design phase are used as the target application scenario information to participate in the subsequent performance requirement analysis of the components of the high vacuum furnace to be produced.

[0027] A200: Disassembling the target design scheme to obtain Y component design schemes corresponding to the Y components of the high vacuum furnace to be produced, where Y is the total number of components of the high vacuum furnace to be produced;

[0028] Specifically, in this embodiment, the spatial connection relationship of Y components of the target design scheme is disassembled to obtain Y component design schemes corresponding to the Y components in the high vacuum furnace to be produced. The component design scheme of the component monomer is used as a reference for optimizing the manufacturing process of the high vacuum furnace to be produced in the component dimension.

[0029] A300: Analyze the performance requirements of the Y components based on the target application scenario information to obtain performance requirement characteristic data of the Y components;

[0030] In one embodiment, Figure 2 As shown, based on the target application scenario information, the performance requirements of the Y components are analyzed to obtain the performance requirement characteristic data of the Y components. The method step A300 provided in the present application also includes:

[0031] A310: performing performance requirement analysis record collection based on the Y components to obtain Y component performance requirement analysis record sets, wherein each component performance requirement analysis record set includes a sample application scenario information record set and a sample component performance requirement feature data record set;

[0032] A320: Based on the BP neural network, respectively train and test the Y component performance requirement analysis record sets to construct Y component performance requirement analysis subnetworks that meet the preset loss constraint conditions;

[0033] A330: Connect the Y component performance requirement analysis sub-networks to generate a component performance requirement analysis integrated channel;

[0034] A340: Based on the component performance requirement analysis integrated channel, the performance requirement analysis of the Y components is performed according to the target application scenario information to generate performance requirement characteristic data of the Y components.

[0035] Specifically, in this embodiment, a component is randomly selected from the Y components as the first component, and then the performance requirement analysis record is collected interactively with the high vacuum furnace manufacturer to obtain multiple sample high vacuum furnaces containing the first component that have been historically produced, and then multiple sample application scenario information records of the first component in the multiple sample vacuum blast furnaces and multiple sample component performance requirement characteristic data are interactively obtained, wherein the sample component performance requirement characteristic data characterizes the performance characteristic parameters that the first component needs to have to be able to stably operate in the sample high vacuum furnace in the actual application scenario corresponding to the sample application scenario information record.

[0036] Multiple sample application scenario information records of the first component constitute a first sample application scenario information record set, multiple sample component performance requirement characteristic data of the first component constitute a first sample component performance requirement characteristic data record set, and at the same time, the first sample application scenario information record set and the first sample component performance requirement characteristic data record set constitute a first component performance requirement analysis record set of the first component.

[0037] Based on the BP (back propagation) neural network, a first component performance requirement analysis subnetwork is constructed to scientifically analyze and determine the performance characteristic parameters required for the first component to stably operate in the current high vacuum furnace to be produced. The input data of the first component performance requirement analysis subnetwork is the component application scenario information, and the output result is the component performance requirement characteristic data.

[0038] The first component performance requirement analysis record set of the first component is used as model training data, and is divided into a training set, a test set and a validation set using a preset partitioning threshold (eg, training set: test set: validation set = 8:1:1).

[0039] Then, conventional BP neural network training methods are adopted to train, test and verify the first component performance requirement analysis subnetwork based on the training set, test set and verification set, respectively, until the component performance characteristic analysis output accuracy of the first component performance requirement analysis subnetwork is stably higher than the preset loss constraint condition (for example, stably higher than 97%).

[0040] The same method as that for obtaining the first component performance requirement analysis record set of the first component is adopted to collect performance requirement analysis records of the Y components to obtain Y component performance requirement analysis record sets.

[0041] Then, the same method as that of obtaining the first component performance requirement analysis subnetwork of the first component is adopted, and based on the BP neural network, the Y component performance requirement analysis record sets are trained and tested respectively to construct Y component performance requirement analysis subnetworks that meet the preset loss constraints.

[0042] The Y component performance requirement analysis sub-networks are arranged in parallel and synchronized to a pre-built component performance requirement analysis integrated channel. After the target application scenario information is synchronized to the component performance requirement analysis integrated channel, the Y component performance requirement analysis sub-networks that enter the component performance requirement analysis integrated channel are synchronized, and performance requirement analysis is performed on the Y components based on the mapping of the Y component performance requirement analysis sub-networks to generate performance requirement characteristic data of the Y components that characterize the performance characteristic parameters required for the Y components to operate stably in the current high vacuum furnace to be produced.

[0043] This embodiment achieves the technical effect of simultaneously performing performance requirement analysis on multiple components by constructing a multi-channel component performance requirement analysis integrated channel, thereby improving the efficiency of component performance requirement analysis and the accuracy of component performance requirement analysis.

[0044] A400: performing importance identification according to the Y components, and constructing Y component process optimization constraint operators based on the component importance identification results, wherein each component process optimization constraint operator includes a component process optimization cost constraint, a component process optimization step constraint, and a component process optimization iteration constraint;

[0045] In one embodiment, importance identification is performed according to the Y components, and Y component process optimization constraint operators are constructed based on the component importance identification results. The method step A400 provided in the present application also includes:

[0046] A410: Obtaining Y component efficacy characteristic data corresponding to the Y components;

[0047] A420: performing importance evaluation on the Y components based on the Y component efficacy characteristic data to obtain the component importance identification result, wherein the component importance identification result includes Y component importance evaluation coefficients;

[0048] A430: performing a proportion calculation based on the importance evaluation coefficients of the Y components to obtain individual value coefficients of the Y components;

[0049] A440: loading the global component processing cost constraints of the high vacuum furnace to be produced, and performing processing cost allocation calculation in combination with the individual value coefficients of the Y components to generate the process optimization cost constraints of the Y components;

[0050] A450: inputting the individual value coefficients of the Y components into a pre-constructed optimization step constraint table to obtain the process optimization step constraints of the Y components;

[0051] A460: Inputting the individual value coefficients of the Y components into a pre-constructed optimization iteration constraint table to obtain the process optimization iteration constraints of the Y components;

[0052] A470: Generate the Y component process optimization constraint operators according to the Y component process optimization cost constraints, the Y component process optimization step constraints and the Y component process optimization iteration constraints.

[0053] Specifically, in this embodiment, Y component efficiency characteristic data corresponding to the Y components are obtained, and the component efficiency characteristic data reflect the function and role of the component in the high vacuum furnace to be produced, specifically including the component heat transfer coefficient, component pressure range, component mass parameter, and component volume parameter.

[0054] Pre-weighting is performed on the various constituent data of the component efficacy characteristic data to obtain a plurality of weight parameters, and mapping and weighting are performed on the Y component efficacy characteristic data corresponding to the Y components after normalization preprocessing to obtain the Y component importance evaluation coefficients, which constitute the component importance identification result.

[0055] The sum of the importance evaluation coefficients of Y components is used as the denominator, and the importance evaluation coefficients of Y components are used as the numerator. The importance evaluation coefficients of the Y components are calculated as a percentage to obtain the individual value coefficients of the Y components.

[0056] The global component processing cost constraint preset for characterizing the overall production economic cost constraint of the high vacuum furnace to be produced is loaded, and the processing cost allocation calculation is performed in combination with the individual value coefficients of the Y components to generate Y component process optimization cost constraints, and the component process optimization cost constraint characterizes the monomer processing cost constraint of the corresponding component.

[0057] In this embodiment, the optimization step constraint table consisting of several value coefficient thresholds-optimization step scales is pre-constructed, and the individual value coefficients of the Y components are input into the pre-constructed optimization step constraint table one by one to compare the value coefficient thresholds they fall into, so as to obtain Y component process optimization step constraints. Each component process optimization step constraint represents the adjustment step length when optimizing the process parameters of the corresponding component.

[0058] In this embodiment, the optimization iteration constraint table consisting of several value coefficient thresholds-optimization iteration scales is pre-constructed, and the individual value coefficients of the Y components are input into the pre-constructed optimization iteration constraint table one by one to compare the value coefficient thresholds they fall into, so as to obtain Y component process optimization iteration constraints. Each component process optimization iteration constraint represents the maximum number of iterations (adjustments) of the corresponding component when performing process parameter optimization adjustment.

[0059] According to the Y components, data integration is performed on the process optimization cost constraints of the Y components, the process optimization step constraints of the Y components and the process optimization iteration constraints of the Y components to generate the process optimization constraint operators of the Y components that characterize the optimization constraints when the Y components are subjected to process optimization.

[0060] This embodiment obtains the component unit processing cost constraints, component processing parameter adjustment constraints and processing parameter adjustment frequency constraints through scientific calculation, thereby achieving the technical effect of providing an optimization direction reference for the subsequent execution of component unit processing technology optimization.

[0061] A500: Perform data fusion based on the Y component design solutions and the Y component performance requirement characteristic data to obtain Y component processing scene source data;

[0062] According to the Y components, the Y component design schemes and the Y component performance requirement characteristic data are mapped and stored to complete data fusion, and obtain Y component processing scene source data. The component processing source data of the component monomer is used to subsequently guide and support the selection and optimization of component processing technology.

[0063] A600: Based on a multi-source heterogeneous manufacturing process optimization algorithm, perform component processing process optimization analysis on the Y components according to the Y component processing scene source data and the Y component process optimization constraint operators, and determine the component processing process plan for the high vacuum furnace to be produced;

[0064] In one embodiment, based on a multi-source heterogeneous manufacturing process optimization algorithm, component processing process optimization analysis is performed on the Y components according to the Y component processing scene source data and the Y component process optimization constraint operators, and a component processing process scheme for the high vacuum furnace to be produced is determined. The method step A600 provided in the present application also includes:

[0065] A610: extracting the yth component processing scenario source data and the yth component process optimization constraint operator corresponding to the yth component according to the Y component processing scenario source data and the Y component process optimization constraint operator, wherein the yth component processing scenario source data includes the yth component design scheme and the yth component performance requirement characteristic data, and the yth component process optimization constraint operator includes the yth component process optimization cost constraint, the yth component process optimization step constraint and the yth component process optimization iteration constraint;

[0066] A620: Based on the processing scene registration constraint, mining processing technology features according to the processing scene source data of the yth component, and constructing the yth component processing technology optimization space;

[0067] A630: Based on the process optimization cost constraint of the yth component, perform optimization analysis on the processing technology optimization space of the yth component according to the process fitting optimization analysis model and the process fitting fitness constraint to obtain the processing technology optimization domain of the yth component;

[0068] A640: adjusting the yth component processing technology optimization domain based on the yth component process optimization stepping constraint to obtain the yth component processing technology optimization neighborhood;

[0069] A650: Based on the iterative constraints of the yth component process optimization, the yth component processing process optimization domain and the yth component processing process optimization neighborhood are jointly optimized according to the process fitting optimization analysis model to obtain the yth component processing process plan, and the yth component processing process plan is added to the component processing process plans.

[0070] In one embodiment, based on the processing scene registration constraint, processing technology feature mining is performed according to the processing scene source data of the yth component to construct the yth component processing technology optimization space. The method step A620 provided in the present application also includes:

[0071] A621: Based on the yth component, processing record data is collected to obtain multiple homologous component processing records, wherein each homologous component processing record includes historical component processing scene source data and historical component processing process plan;

[0072] A622: Based on the yth component processing scene source data, respectively perform processing scene feature similarity analysis on each historical component processing scene source data in the multiple homologous component processing records to determine multiple processing scene registration coefficients;

[0073] A623: Compare the processing scene registration constraint with the multiple processing scene registration coefficients to obtain a preferred processing scene registration coefficient set that satisfies the processing scene registration constraint, and match the processing schemes of the multiple homologous component processing records based on the preferred processing scene registration coefficient set to obtain a preferred registration processing scheme set;

[0074] A624: Classifying based on the preferred registration processing scheme set, obtaining a multi-level component processing index characteristic parameter set, and performing interval analysis on the multi-level component processing index characteristic parameter set to construct a multi-level component processing index characteristic interval;

[0075] A625: Traverse the multi-level component processing index characteristic interval and perform multiple random value sampling to generate the y-th component processing technology optimization space, wherein the y-th component processing technology optimization space includes multiple component processing technology decisions.

[0076] In one embodiment, based on the process optimization cost constraint of the yth component, the process optimization space of the yth component is optimized according to the process fitting optimization analysis model and the process fitting fitness constraint to obtain the yth component process optimization domain. The method step A630 provided in the present application also includes:

[0077] A631: extracting a first component processing technology decision according to the yth component processing technology optimization space;

[0078] A632: Based on the process optimization cost constraint of the yth component and the process fitting optimization analysis model, the fitness of the first component processing process decision is evaluated to obtain a first decision fitness;

[0079] A633: If the first decision fitness satisfies the process fitting fitness constraint, adding the first component processing process decision to the yth component processing process optimization domain;

[0080] A634: Based on the process optimization cost constraint of the yth component, the process fitting optimization analysis model and the process fitting fitness constraint, continue to perform optimization analysis on the processing technology optimization space of the yth component to generate the processing technology optimization domain of the yth component that meets the preset decision capacity constraint.

[0081] In one embodiment, based on the process optimization cost constraint of the yth component and the process fitting optimization analysis model, the fitness of the first component processing process decision is evaluated to obtain the first decision fitness. The method step A632 provided in the present application also includes:

[0082] A6321: fitting the first component processing process decision according to the process fitting module in the process fitting optimization analysis model to obtain a first decision process fitting result, wherein the first decision process fitting result includes the first decision fitting component structure information, the first decision fitting component performance information and the first decision fitting component production cost;

[0083] A6322: Based on the component structure fitness evaluation network in the process fitting optimization analysis model, perform fitness evaluation on the first decision fitting component structure information according to the yth component design scheme to obtain a first decision structure fitness evaluation coefficient;

[0084] A6323: Based on the component performance fitness evaluation network in the process fitting optimization analysis model, the first decision fitting component performance information is evaluated for fitness according to the yth component performance requirement characteristic data to obtain a first decision performance fitness evaluation coefficient;

[0085] A6324: Based on the component cost fitness evaluation network in the process fitting optimization analysis model, the fitness of the first decision fitting component production cost is evaluated according to the process optimization cost constraint of the yth component to obtain the first decision cost fitness evaluation coefficient;

[0086] A6325: Based on the fitness evaluation weight constraint within the process fitting optimization analysis model, the first decision structure fitness evaluation coefficient, the first decision performance fitness evaluation coefficient and the first decision cost fitness evaluation coefficient are weightedly calculated to generate the first decision fitness.

[0087] In one embodiment, based on the iterative constraint of the process optimization of the yth component, the yth component processing process optimization domain and the yth component processing process optimization neighborhood are jointly optimized according to the process fitting optimization analysis model to obtain the yth component processing process solution. The method step A650 provided in the present application also includes:

[0088] A651: extracting the optimal processing technology decision in the domain and the optimal decision fitness in the domain corresponding to the optimal processing technology decision in the domain according to the processing technology optimization domain of the yth component;

[0089] A652: extracting the first processing technology neighborhood decision according to the yth component processing technology optimization neighborhood, and calculating the fitness of the first neighborhood decision according to the process fitting optimization analysis model;

[0090] A653: Screening and analyzing the optimal decision fitness within the domain, the first neighborhood decision fitness, the optimal processing technology decision within the domain, and the first processing technology neighborhood decision to determine the current optimal decision fitness and the current optimal processing technology decision corresponding to the current optimal decision fitness;

[0091] A654: Based on the process fitting optimization analysis model, continue to perform neighborhood iterative optimization on the y-th component processing technology optimization neighborhood according to the current optimal decision fitness and the current optimal processing technology decision, until the number of neighborhood iterative optimization times meets the y-th component process optimization iteration constraint, and generate the y-th component processing technology plan.

[0092] Specifically, this embodiment randomly selects the yth component from Y components, and then extracts the yth component processing scene source data and the yth component process optimization constraint operator corresponding to the yth component from the Y component processing scene source data and the Y component process optimization constraint operators.

[0093] Based on step A500, it can be known that the yth component processing scene source data includes the yth component design plan and the yth component performance requirement characteristic data; based on step A400, it can be known that the yth component process optimization constraint operator includes the yth component process optimization cost constraint, the yth component process optimization step constraint and the yth component process optimization iteration constraint.

[0094] It should be understood that homologous components refer to components with the same design scheme and are used in different products or systems. This embodiment collects processing record data of homologous components for the yth component to obtain multiple homologous component processing records, each homologous component processing record includes historical component processing scene source data and historical component processing process scheme, and the data indicator composition of the historical component processing scene source data is consistent with the component processing scene source data of this embodiment.

[0095] The performance requirement characteristic data of the yth component is called based on the yth component processing scene source data, and multiple homologous component performance requirement characteristic data are obtained based on the historical component processing scene source data in the multiple homologous component processing records.

[0096] Then, the mean values ​​of the performance requirement characteristic deviations between the performance requirement characteristic data of the yth component and the performance requirement characteristic data of multiple homologous components are calculated one by one as the multiple processing scene alignment coefficients of the multiple homologous components. The processing scene alignment coefficients reflect the deviation between the homologous component and the yth component from the dimension of component performance deviation.

[0097] The preset standard allows the processing scene registration constraint within the performance requirement deviation range between the homologous component and the yth component, and compares the processing scene registration constraint with the multiple processing scene registration coefficients to obtain the preferred processing scene registration coefficient set consisting of K processing scene registration coefficients that satisfy the processing scene registration constraint, and based on the K processing scene registration coefficients, obtain the corresponding K historical component processing scene source data, and then match the corresponding processing process plans for the multiple homologous component processing records based on the K historical component processing scene source data to obtain the preferred registration processing process plan set including the K historical component processing process plans.

[0098] After obtaining the preferred registration processing technology solution set, this embodiment extracts processing technology parameters from the preferred registration processing technology solution set to obtain multiple types of processing control indicators.

[0099] Then, the preferred registration processing scheme set is classified based on multiple processing control index types to obtain the multi-level component processing index characteristic parameter set including multiple processing control index types corresponding to a single type of control index data.

[0100] Then, the multi-level component processing index characteristic parameter set is subjected to data serialization processing within the set to extract the maximum value and the minimum value to form a multi-level component processing index characteristic interval. Each level of component processing index characteristic interval represents the parameter value range of a control index.

[0101] A multidimensional space is constructed based on multiple types of processing control indicators, each dimension in the multidimensional space corresponds to a type of processing control indicator, and then the multidimensional space is constrained by the multi-level component processing indicator feature interval to generate the yth component processing process optimization space.

[0102] The multi-level component processing index characteristic interval is traversed to perform multiple random value taking to obtain multiple groups of processing index characteristic values, and the data of each group of processing index characteristic values ​​includes multiple random values ​​of the multi-level component processing index characteristic interval.

[0103] A plurality of sets of processing index characteristic values ​​are spatially positioned as particle points in the y-th component processing technology optimization space to obtain a plurality of component processing technology decisions in the y-th component processing technology optimization space.

[0104] In this embodiment, a process fitting optimization analysis model is pre-constructed, and the process fitting optimization analysis model includes a process fitting module, a component structure fitness evaluation network, a component performance fitness evaluation network, a component cost fitness evaluation network, and a fitness evaluation module, wherein the component structure fitness evaluation network, the component performance fitness evaluation network, and the component cost fitness evaluation network are arranged in parallel, and the input ends of the three network structures are all connected to the output end of the process fitting module, and the output ends of the three network structures are merged into the input end of the fitness evaluation module, and the fitness evaluation weight constraint is pre-configured in the fitness evaluation module, and the fitness evaluation weight constraint is a weight configuration for weighted calculation of the output results of the four network structures.

[0105] In the subsequent description of this embodiment, taking the random extraction of the first component processing process decision from the y-th component processing process optimization space, performing fitness evaluation on the first component processing process decision based on the y-th component process optimization cost constraint and the process fitting optimization analysis model, and obtaining the first decision fitness as an example, the construction method of each module and network in the process fitting optimization analysis model and the data processing logic are elaborated in detail.

[0106] In this embodiment, the function of the process fitting module is to perform component structure fitting, component performance fitting and component production cost prediction based on processing control parameter information, and the process fitting module is constructed based on a BP neural network.

[0107] A plurality of sample process fitting data sets are interactively obtained, and the data structure of each sample process fitting data set includes a sample processing index characteristic value combination, sample production component structure information, sample production component performance information and sample production component cost record.

[0108] The process fitting analysis module is trained by using multiple sample process fitting data sets with the same logic as the component performance requirement analysis sub-network construction and training, so as to obtain the process fitting analysis module which can use the processing control parameters as input and analyze the structural information, performance information and production cost of the components produced under the processing control parameters.

[0109] The first component processing process decision is input as a processing control parameter into the process fitting module within the process fitting optimization analysis model for fitting, so as to obtain a first decision process fitting result, wherein the first decision process fitting result includes the first decision fitting component structure information, the first decision fitting component performance information and the first decision fitting component production cost.

[0110] In this embodiment, the function of the component structure fitness evaluation network is to perform automatic three-dimensional modeling of the component according to the component design plan, perform automatic three-dimensional modeling of the component according to the component structure, and then perform intersection calculation on the two three-dimensional models, and obtain the fitness evaluation coefficient based on the intersection of the three-dimensional model divided by the union of the three-dimensional models.

[0111] The component structure fitness evaluation network is constructed based on a deep learning model, and is based on multiple groups of sample component design solutions-sample component models, multiple groups of sample component structure information-sample component models as training data, and is trained and generated using a conventional deep learning model training method.

[0112] The y-th component design scheme and the first decision fitting component structure information are input as input data into the component structure fitness evaluation network within the process fitting optimization analysis model. First, the two component three-dimensional models are automatically modeled, and then the two component three-dimensional models are overlapped, the intersection volume and the union volume are calculated, and then the intersection volume is divided by the union volume to evaluate the fitness of the first decision fitting component structure information, and the first decision structure fitness evaluation coefficient is obtained.

[0113] In this embodiment, the function of the component performance fitness evaluation network is to calculate the performance deviation of each component performance and the average of multiple component performance deviation calculation results, and use the average calculation result as the component performance fitness evaluation result.

[0114] After synchronizing the first decision fitting component performance information of the first decision process fitting result and the yth component performance requirement characteristic data to the component performance fitness evaluation network, firstly group the data with the same performance requirement characteristics, and then calculate the performance deviation within the group to obtain multiple performance requirement deviations, and then calculate the mean of the multiple performance requirement deviations to complete the fitness evaluation of the first decision fitting component performance information with the yth component performance requirement characteristic data to obtain the first decision performance fitness evaluation coefficient.

[0115] In this embodiment, the function of the component cost fitness evaluation network is as follows: the first decision fitting component production cost of the first decision process fitting result and the y-th component process optimization cost constraint are synchronized to the component cost fitness evaluation network, and the deviation percentage of the first decision fitting component production cost from the y-th component process optimization cost constraint is calculated in the component cost fitness evaluation network, and the deviation percentage is output as the fitness evaluation result to obtain the first decision cost fitness evaluation coefficient.

[0116] It should be understood that the first decision fitting component structure information, the first decision fitting component performance information and the first decision fitting component production cost in the first decision process fitting result output by the process fitting module are simultaneously input into the component structure fitness evaluation network, the component performance fitness evaluation network and the component cost fitness evaluation network, and adaptability evaluation is performed synchronously based on the three networks, and the first decision structure fitness evaluation coefficient, the first decision performance fitness evaluation coefficient and the first decision cost fitness evaluation coefficient are synchronously output.

[0117] Then, the first decision structure fitness evaluation coefficient, the first decision performance fitness evaluation coefficient and the first decision cost fitness evaluation coefficient are synchronously input into the fitness evaluation module, and the first decision structure fitness evaluation coefficient, the first decision performance fitness evaluation coefficient and the first decision cost fitness evaluation coefficient are mapped and weighted calculated based on the fitness evaluation weight constraint in the fitness evaluation module to generate the first decision fitness.

[0118] The first decision fitness is compared with the process fitting fitness constraint based on the manufacturer's customized characterization of the production accuracy and minimum production cost requirements of the high vacuum furnace to be produced. If the first decision fitness satisfies the process fitting fitness constraint, the first component processing process decision is added to the yth component processing process optimization domain.

[0119] The same method as that for determining whether to add the first component processing technology decision to the yth component processing technology optimization domain is adopted, and based on the yth component processing technology optimization cost constraint, the process fitting optimization analysis model and the process fitting fitness constraint, the multiple component processing technology decisions in the yth component processing technology optimization space are continuously determined whether to be added to the yth component processing technology optimization domain, until the number of component processing technology decisions in the yth component processing technology optimization domain reaches the preset decision capacity constraint.

[0120] Based on the yth component process optimization step constraint, parameters of several component processing technology decisions stored in the yth component processing technology optimization domain are adjusted to obtain the yth component processing technology optimization neighborhood consisting of several groups of expanded processing technology decisions that expand the several component processing technology decisions.

[0121] A plurality of decision fitnesses of a plurality of component processing technology decisions stored in the yth component processing technology optimization domain are serialized to obtain a maximum decision fitness.

[0122] The processing technology decision corresponding to the maximum decision fitness is taken as the optimal processing technology decision in the domain, and the maximum decision fitness is taken as the optimal decision fitness in the domain.

[0123] A first processing technology neighborhood decision is randomly extracted from a plurality of groups of expanded processing technology decisions of the processing technology optimization neighborhood of the yth component, and the fitness of the first neighborhood decision is calculated according to the process fitting optimization analysis model.

[0124] Compare the fitness of the optimal decision in the domain and the fitness of the first neighborhood decision, retain the processing technology decision with the larger decision fitness, screen and analyze the optimal processing technology decision in the domain and the first processing technology neighborhood decision, determine the current optimal decision fitness, and the current optimal processing technology decision corresponding to the current optimal decision fitness.

[0125] Based on the process fitting optimization analysis model, the neighborhood iterative optimization of the y-th component processing technology optimization neighborhood is continued according to the current optimal decision fitness and the current optimal processing technology decision, until the number of neighborhood iterative optimization times meets the y-th component process optimization iteration constraint, and the y-th component processing technology plan is generated.

[0126] By analogy, the same method as that for generating the yth component processing scheme is adopted to generate component processing schemes for Y components. The component processing schemes for Y components and the connection relationship between the Y components in the high vacuum furnace to be produced and the component spatial position relationship constitute the component processing scheme. This embodiment achieves the technical effect of quickly obtaining a component processing scheme that simultaneously meets the multi-dimensional requirements of cost, component structure and performance.

[0127] A700: Execute component processing of the high vacuum furnace to be produced according to the component processing process plan.

[0128] Specifically, this embodiment makes a reference to the production equipment according to the component processing process plan, and starts the production equipment to execute the component processing of the high vacuum furnace to be produced, thereby achieving the technical effect of improving the adaptability of the component processing plan to production requirements, and the efficiency and scientificity of the component processing plan generation.

[0129] Embodiment 2

[0130] Based on the same inventive concept as the method for optimizing the manufacturing process of a high vacuum furnace in the aforementioned embodiment, Figure 3 As shown, the present application provides a manufacturing process optimization system for a high vacuum furnace, wherein the system comprises:

[0131] A solution output unit 1 is used to collect design parameters and application scenario parameters of the high vacuum furnace to be produced, and obtain target design solution and target application scenario information of the high vacuum furnace to be produced;

[0132] A scheme disassembly execution unit 2 is used to disassemble the target design scheme to obtain Y component design schemes corresponding to the Y components of the high vacuum furnace to be produced, wherein Y is the total number of components of the high vacuum furnace to be produced;

[0133] A demand analysis execution unit 3 is used to perform performance demand analysis on the Y components based on the target application scenario information to obtain performance demand characteristic data of the Y components;

[0134] The optimization constraint construction unit 4 is used to identify the importance of the Y components and construct Y component process optimization constraint operators based on the component importance identification results, wherein each component process optimization constraint operator includes a component process optimization cost constraint, a component process optimization step constraint and a component process optimization iteration constraint;

[0135] A data fusion execution unit 5 is used to perform data fusion based on the Y component design solutions and the Y component performance requirement characteristic data to obtain Y component processing scene source data;

[0136] The optimization analysis execution unit 6 is used to perform component processing optimization analysis on the Y components respectively based on the multi-source heterogeneous manufacturing process optimization algorithm according to the Y component processing scene source data and the Y component process optimization constraint operators, and determine the component processing process plan of the high vacuum furnace to be produced;

[0137] The component processing execution unit 7 is used to execute the component processing of the high vacuum furnace to be produced according to the component processing process plan.

[0138] In one embodiment, the demand parsing execution unit 3 further includes:

[0139] Based on the Y components, performance requirement analysis records are collected to obtain Y component performance requirement analysis record sets, wherein each component performance requirement analysis record set includes a sample application scenario information record set and a sample component performance requirement feature data record set;

[0140] Based on the BP neural network, the Y component performance requirement analysis record sets are trained and tested respectively to construct Y component performance requirement analysis subnetworks that meet the preset loss constraint conditions;

[0141] Connecting the Y component performance requirement analysis sub-networks to generate a component performance requirement analysis integrated channel;

[0142] Based on the component performance requirement analysis integrated channel, performance requirement analysis is performed on the Y components according to the target application scenario information to generate performance requirement characteristic data of the Y components.

[0143] In one embodiment, the optimization constraint construction unit 4 further includes:

[0144] Obtaining Y component efficacy characteristic data corresponding to the Y components;

[0145] Performing importance evaluation on the Y components based on the Y component efficacy characteristic data to obtain the component importance identification result, wherein the component importance identification result includes Y component importance evaluation coefficients;

[0146] Performing a proportion calculation based on the importance evaluation coefficients of the Y components to obtain individual value coefficients of the Y components;

[0147] Load the global component processing cost constraints of the high vacuum furnace to be produced, and perform processing cost allocation calculation in combination with the individual value coefficients of the Y components to generate process optimization cost constraints for the Y components;

[0148] Inputting the individual value coefficients of the Y components into a pre-constructed optimization step constraint table to obtain the process optimization step constraints of the Y components;

[0149] Inputting the individual value coefficients of the Y components into a pre-constructed optimization iteration constraint table to obtain the process optimization iteration constraints of the Y components;

[0150] The Y component process optimization constraint operators are generated according to the Y component process optimization cost constraints, the Y component process optimization step constraints and the Y component process optimization iteration constraints.

[0151] In one embodiment, the optimization analysis execution unit 6 further includes:

[0152] Extracting the yth component processing scene source data and the yth component process optimization constraint operator corresponding to the yth component according to the Y component processing scene source data and the Y component process optimization constraint operator, wherein the yth component processing scene source data includes the yth component design scheme and the yth component performance requirement characteristic data, and the yth component process optimization constraint operator includes the yth component process optimization cost constraint, the yth component process optimization step constraint and the yth component process optimization iteration constraint;

[0153] Based on the processing scene registration constraint, mining processing technology features according to the processing scene source data of the yth component, and constructing the processing technology optimization space of the yth component;

[0154] Based on the process optimization cost constraint of the yth component, an optimization analysis is performed on the processing technology optimization space of the yth component according to the process fitting optimization analysis model and the process fitting fitness constraint to obtain the processing technology optimization domain of the yth component;

[0155] Adjusting the yth component processing technology optimization domain based on the yth component process optimization step constraint to obtain the yth component processing technology optimization neighborhood;

[0156] Based on the iterative constraints of the yth component process optimization, the yth component processing process optimization domain and the yth component processing process optimization neighborhood are jointly optimized according to the process fitting optimization analysis model to obtain the yth component processing process plan, and the yth component processing process plan is added to the component processing process plans.

[0157] In one embodiment, the optimization analysis execution unit 6 further includes:

[0158] Based on the yth component, processing record data is collected to obtain multiple homologous component processing records, wherein each homologous component processing record includes historical component processing scene source data and historical component processing process plan;

[0159] Based on the yth component processing scene source data, respectively performing processing scene feature similarity analysis on each historical component processing scene source data in the multiple homologous component processing records to determine multiple processing scene registration coefficients;

[0160] Comparing the processing scene registration constraint with the multiple processing scene registration coefficients to obtain a preferred processing scene registration coefficient set that satisfies the processing scene registration constraint, and matching the processing process plans of the multiple homologous component processing records based on the preferred processing scene registration coefficient set to obtain a preferred registration processing process plan set;

[0161] Classifying based on the preferred registration processing scheme set, obtaining a multi-level component processing index characteristic parameter set, and performing interval analysis on the multi-level component processing index characteristic parameter set to construct a multi-level component processing index characteristic interval;

[0162] The multi-level component processing index characteristic interval is traversed to perform multiple random value selections to generate the y-th component processing technology optimization space, wherein the y-th component processing technology optimization space includes multiple component processing technology decisions.

[0163] In one embodiment, the optimization analysis execution unit 6 further includes:

[0164] Extracting a first component processing technology decision according to the yth component processing technology optimization space;

[0165] Based on the process optimization cost constraint of the yth component and the process fitting optimization analysis model, the fitness evaluation of the first component processing process decision is performed to obtain a first decision fitness;

[0166] If the first decision fitness satisfies the process fitting fitness constraint, adding the first component processing process decision to the yth component processing process optimization domain;

[0167] Based on the process optimization cost constraint of the yth component, the process fitting optimization analysis model and the process fitting fitness constraint, the optimization analysis of the processing technology optimization space of the yth component is continued to be performed to generate the processing technology optimization domain of the yth component that meets the preset decision capacity constraint.

[0168] In one embodiment, the optimization analysis execution unit 6 further includes:

[0169] Fitting the first component processing process decision according to the process fitting module in the process fitting optimization analysis model to obtain a first decision process fitting result, wherein the first decision process fitting result includes first decision fitting component structure information, first decision fitting component performance information and first decision fitting component production cost;

[0170] Based on the component structure fitness evaluation network in the process fitting optimization analysis model, the fitness evaluation of the first decision fitting component structure information is performed according to the yth component design scheme to obtain a first decision structure fitness evaluation coefficient;

[0171] Based on the component performance fitness evaluation network in the process fitting optimization analysis model, the first decision fitting component performance information is evaluated for fitness according to the yth component performance requirement characteristic data to obtain a first decision performance fitness evaluation coefficient;

[0172] Based on the component cost fitness evaluation network in the process fitting optimization analysis model, the fitness evaluation of the production cost of the first decision fitting component is performed according to the process optimization cost constraint of the yth component to obtain a first decision cost fitness evaluation coefficient;

[0173] Based on the fitness evaluation weight constraint in the process fitting optimization analysis model, the first decision structure fitness evaluation coefficient, the first decision performance fitness evaluation coefficient and the first decision cost fitness evaluation coefficient are weightedly calculated to generate the first decision fitness.

[0174] In one embodiment, the optimization analysis execution unit 6 further includes:

[0175] Extracting the optimal processing technology decision in the domain and the optimal decision fitness in the domain corresponding to the optimal processing technology decision in the domain according to the yth component processing technology optimization domain;

[0176] Extracting the first processing technology neighborhood decision according to the yth component processing technology optimization neighborhood, and calculating the fitness of the first neighborhood decision according to the process fitting optimization analysis model;

[0177] Screening and analyzing the optimal decision fitness within the domain, the first neighborhood decision fitness, the optimal processing technology decision within the domain, and the first processing technology neighborhood decision to determine the current optimal decision fitness and the current optimal processing technology decision corresponding to the current optimal decision fitness;

[0178] Based on the process fitting optimization analysis model, the neighborhood iterative optimization of the y-th component processing technology optimization neighborhood is continued according to the current optimal decision fitness and the current optimal processing technology decision, until the number of neighborhood iterative optimization times meets the y-th component process optimization iteration constraint, and the y-th component processing technology plan is generated.

[0179] Any of the methods or steps described above may be stored as computer instructions or programs in various types of computer memories, and the computer instructions or programs may be recognized by various types of computer processors to implement any of the methods or steps described above.

[0180] Based on the above specific embodiments of the present invention, any improvements and modifications made to the present invention by those skilled in the art without departing from the principles of the present invention shall fall within the patent protection scope of the present invention.

Claims

1. A method for optimizing the manufacturing process of a high vacuum furnace, characterized in that: The method comprises: Collecting design parameters and application scenario parameters of the high vacuum furnace to be produced, and obtaining target design scheme and target application scenario information of the high vacuum furnace to be produced; Disassembling the target design scheme to obtain Y component design schemes corresponding to the Y components of the high vacuum furnace to be produced, where Y is the total number of components of the high vacuum furnace to be produced; Based on the target application scenario information, the performance requirements of the Y components are analyzed to obtain performance requirement characteristic data of the Y components; Importance identification is performed according to the Y components, and Y component process optimization constraint operators are constructed based on the component importance identification results, wherein each component process optimization constraint operator includes a component process optimization cost constraint, a component process optimization step constraint, and a component process optimization iteration constraint; Perform data fusion based on the Y component design solutions and the Y component performance requirement characteristic data to obtain Y component processing scene source data; Based on the multi-source heterogeneous manufacturing process optimization algorithm, the component processing process optimization analysis is performed on the Y components according to the Y component processing scene source data and the Y component process optimization constraint operators, and the component processing process plan of the high vacuum furnace to be produced is determined; The component processing of the high vacuum furnace to be produced is performed according to the component processing process plan.

2. The method according to claim 1, characterized in that Based on the target application scenario information, the performance requirements of the Y components are analyzed to obtain performance requirement characteristic data of the Y components, including: Based on the Y components, performance requirement analysis records are collected to obtain Y component performance requirement analysis record sets, wherein each component performance requirement analysis record set includes a sample application scenario information record set and a sample component performance requirement feature data record set; Based on the BP neural network, the Y component performance requirement analysis record sets are trained and tested respectively to construct Y component performance requirement analysis subnetworks that meet the preset loss constraint conditions; Connecting the Y component performance requirement analysis sub-networks to generate a component performance requirement analysis integrated channel; Based on the component performance requirement analysis integrated channel, performance requirement analysis is performed on the Y components according to the target application scenario information to generate performance requirement characteristic data of the Y components.

3. The method according to claim 1, characterized in that Importance identification is performed according to the Y components, and Y component process optimization constraint operators are constructed based on the component importance identification results, including: Obtaining Y component efficacy characteristic data corresponding to the Y components; Performing importance evaluation on the Y components based on the Y component efficacy characteristic data to obtain the component importance identification result, wherein the component importance identification result includes Y component importance evaluation coefficients; Performing a proportion calculation based on the importance evaluation coefficients of the Y components to obtain individual value coefficients of the Y components; Load the global component processing cost constraints of the high vacuum furnace to be produced, and perform processing cost allocation calculation in combination with the individual value coefficients of the Y components to generate process optimization cost constraints for the Y components; Inputting the individual value coefficients of the Y components into a pre-constructed optimization step constraint table to obtain the process optimization step constraints of the Y components; Inputting the individual value coefficients of the Y components into a pre-constructed optimization iteration constraint table to obtain the process optimization iteration constraints of the Y components; The Y component process optimization constraint operators are generated according to the Y component process optimization cost constraints, the Y component process optimization step constraints and the Y component process optimization iteration constraints.

4. The method according to claim 1, characterized in that Based on the multi-source heterogeneous manufacturing process optimization algorithm, according to the Y component processing scene source data and the Y component process optimization constraint operators, the component processing process optimization analysis is performed on the Y components respectively, and the component processing process plan of the high vacuum furnace to be produced is determined, including: Extracting the yth component processing scene source data and the yth component process optimization constraint operator corresponding to the yth component according to the Y component processing scene source data and the Y component process optimization constraint operator, wherein the yth component processing scene source data includes the yth component design scheme and the yth component performance requirement characteristic data, and the yth component process optimization constraint operator includes the yth component process optimization cost constraint, the yth component process optimization step constraint and the yth component process optimization iteration constraint; Based on the processing scene registration constraint, mining processing technology features according to the processing scene source data of the yth component, and constructing the processing technology optimization space of the yth component; Based on the process optimization cost constraint of the yth component, an optimization analysis is performed on the processing technology optimization space of the yth component according to the process fitting optimization analysis model and the process fitting fitness constraint to obtain the processing technology optimization domain of the yth component; Adjusting the yth component processing technology optimization domain based on the yth component process optimization step constraint to obtain the yth component processing technology optimization neighborhood; Based on the iterative constraints of the yth component process optimization, the yth component processing process optimization domain and the yth component processing process optimization neighborhood are jointly optimized according to the process fitting optimization analysis model to obtain the yth component processing process plan, and the yth component processing process plan is added to the component processing process plans.

5. The method according to claim 4, characterized in that Based on the processing scene registration constraint, the processing technology feature mining is performed according to the processing scene source data of the yth component, and the processing technology optimization space of the yth component is constructed, including: Based on the yth component, processing record data is collected to obtain multiple homologous component processing records, wherein each homologous component processing record includes historical component processing scene source data and historical component processing process plan; Based on the yth component processing scene source data, respectively performing processing scene feature similarity analysis on each historical component processing scene source data in the multiple homologous component processing records to determine multiple processing scene registration coefficients; Comparing the processing scene registration constraint with the multiple processing scene registration coefficients to obtain a preferred processing scene registration coefficient set that satisfies the processing scene registration constraint, and matching the processing process plans of the multiple homologous component processing records based on the preferred processing scene registration coefficient set to obtain a preferred registration processing process plan set; Classifying based on the preferred registration processing scheme set, obtaining a multi-level component processing index characteristic parameter set, and performing interval analysis on the multi-level component processing index characteristic parameter set to construct a multi-level component processing index characteristic interval; The multi-level component processing index characteristic interval is traversed to perform multiple random value selections to generate the y-th component processing technology optimization space, wherein the y-th component processing technology optimization space includes multiple component processing technology decisions.

6. The method according to claim 4, characterized in that Based on the process optimization cost constraint of the yth component, the process optimization space of the yth component is optimized according to the process fitting optimization analysis model and the process fitting fitness constraint to obtain the yth component processing process optimization domain, including: Extracting a first component processing technology decision according to the yth component processing technology optimization space; Based on the process optimization cost constraint of the yth component and the process fitting optimization analysis model, the fitness of the first component processing process decision is evaluated to obtain a first decision fitness; If the first decision fitness satisfies the process fitting fitness constraint, adding the first component processing process decision to the yth component processing process optimization domain; Based on the process optimization cost constraint of the yth component, the process fitting optimization analysis model and the process fitting fitness constraint, the optimization analysis of the processing technology optimization space of the yth component is continued to be performed to generate the processing technology optimization domain of the yth component that meets the preset decision capacity constraint.

7. The method according to claim 6, characterized in that The fitness evaluation of the first component processing process decision is performed based on the process optimization cost constraint of the yth component and the process fitting optimization analysis model to obtain the first decision fitness, including: Fitting the first component processing process decision according to the process fitting module in the process fitting optimization analysis model to obtain a first decision process fitting result, wherein the first decision process fitting result includes first decision fitting component structure information, first decision fitting component performance information and first decision fitting component production cost; Based on the component structure fitness evaluation network in the process fitting optimization analysis model, the fitness evaluation of the first decision fitting component structure information is performed according to the yth component design scheme to obtain a first decision structure fitness evaluation coefficient; Based on the component performance fitness evaluation network in the process fitting optimization analysis model, the first decision fitting component performance information is evaluated for fitness according to the yth component performance requirement characteristic data to obtain a first decision performance fitness evaluation coefficient; Based on the component cost fitness evaluation network in the process fitting optimization analysis model, the fitness evaluation of the production cost of the first decision fitting component is performed according to the process optimization cost constraint of the yth component to obtain a first decision cost fitness evaluation coefficient; Based on the fitness evaluation weight constraint in the process fitting optimization analysis model, the first decision structure fitness evaluation coefficient, the first decision performance fitness evaluation coefficient and the first decision cost fitness evaluation coefficient are weightedly calculated to generate the first decision fitness.

8. The method according to claim 4, characterized in that Based on the iterative constraint of the process optimization of the yth component, the process optimization domain of the yth component and the process optimization neighborhood of the yth component are jointly optimized according to the process fitting optimization analysis model to obtain a process solution for the yth component, including: Extracting the optimal processing technology decision in the domain and the optimal decision fitness in the domain corresponding to the optimal processing technology decision in the domain according to the yth component processing technology optimization domain; Extracting the first processing technology neighborhood decision according to the yth component processing technology optimization neighborhood, and calculating the fitness of the first neighborhood decision according to the process fitting optimization analysis model; Screening and analyzing the optimal decision fitness within the domain, the first neighborhood decision fitness, the optimal processing technology decision within the domain, and the first processing technology neighborhood decision to determine the current optimal decision fitness and the current optimal processing technology decision corresponding to the current optimal decision fitness; Based on the process fitting optimization analysis model, the neighborhood iterative optimization of the y-th component processing technology optimization neighborhood is continued according to the current optimal decision fitness and the current optimal processing technology decision, until the number of neighborhood iterative optimization times meets the y-th component process optimization iteration constraint, and the y-th component processing technology plan is generated.

9. A manufacturing process optimization system for a high vacuum furnace, characterized in that: The steps for implementing the method according to any one of claims 1 to 8 include: A solution output unit is used to collect design parameters and application scenario parameters of the high vacuum furnace to be produced, and obtain target design solution and target application scenario information of the high vacuum furnace to be produced; A scheme disassembly execution unit is used to disassemble the target design scheme to obtain Y component design schemes corresponding to the Y components of the high vacuum furnace to be produced, wherein Y is the total number of components of the high vacuum furnace to be produced; A demand analysis execution unit, configured to perform performance demand analysis on the Y components based on the target application scenario information to obtain performance demand characteristic data of the Y components; An optimization constraint construction unit, used for performing importance identification according to the Y components, and constructing Y component process optimization constraint operators based on the component importance identification results, wherein each component process optimization constraint operator includes a component process optimization cost constraint, a component process optimization step constraint, and a component process optimization iteration constraint; A data fusion execution unit, configured to perform data fusion based on the Y component design solutions and the Y component performance requirement characteristic data to obtain Y component processing scene source data; An optimization analysis execution unit is used to perform component processing optimization analysis on the Y components respectively based on the multi-source heterogeneous manufacturing process optimization algorithm according to the Y component processing scene source data and the Y component process optimization constraint operators, and determine the component processing process plan of the high vacuum furnace to be produced; A component processing execution unit is used to execute component processing of the high vacuum furnace to be produced according to the component processing process plan.

Images