Procedure parameter-based result feedback method and device
A technology of process parameters and feedback requests, applied in the field of process knowledge systems, can solve problems affecting manufacturing efficiency and quality, low system early warning efficiency, and inaccurate analysis results, and achieve the effect of improving classification accuracy
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0059] In one embodiment of the present invention, a kind of result feedback method based on process parameter is provided, see figure 1 , the method includes the following processes:
[0060] S1. Receive a result feedback request, where the feedback request includes an intermediate result type.
[0061] Specifically, the intermediate result type is the intermediate result that needs to be controlled in a certain section of the process manufacturing process, and the specific type should be determined according to the actual section, such as: in the extraction section, the component content is the main CQA; in the concentration section, The specific gravity of solid content and extract is CQA; in the alcohol precipitation and extraction process, the transfer rate of index components and the removal rate of impurities are CQA, etc.
[0062] S2. Obtain a process parameter corresponding to the feedback request, where the process parameter is a multidimensional parameter and forms...
Embodiment 2
[0069] In one embodiment of the present invention, a kind of training method of result feedback neural network model is provided, see figure 2 , the method process includes:
[0070] S21. Obtain process parameter sample data to be trained, where the process parameter sample includes multiple process parameter sets and corresponding known target values;
[0071] S22. Establish an initial network model, the initial network model including an input layer, a hidden layer, an output layer, an initial weight, and an initial bias;
[0072] S23. Using the backpropagation method, update the initial weight and the initial bias;
[0073] S24. Determine whether weight convergence is achieved, and if yes, obtain the result and feed it back to the neural network model; otherwise, repeat S23-S24.
[0074] The process of error backpropagation is to apportion the error to all units in each layer, so as to obtain the error signal of each layer unit, and then correct the weight of each unit, ...
Embodiment 3
[0103] In one embodiment of the present invention, a kind of result feedback device based on process parameter is provided, see Figure 5 , the device includes the following modules:
[0104] A request module 310, configured to receive a result feedback request, where the feedback request includes an intermediate result type;
[0105] A parameter module 320, configured to acquire a process parameter corresponding to the feedback request, where the process parameter is a multidimensional parameter and forms a process parameter set;
[0106] The input module 330 is used to input the process parameter set into the result feedback neural network model, and the result feedback neural network model is obtained by training the process parameter samples;
[0107] The output module 340 is configured to obtain the output result of the result feedback neural network model.
[0108] Further, the device also includes a training module 350, and the training module 350 includes:
[0109] ...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com