CN116402574A - Flexible intelligent manufacturing method and system based on live broadcast information feedback - Google Patents
Flexible intelligent manufacturing method and system based on live broadcast information feedback Download PDFInfo
- Publication number
- CN116402574A CN116402574A CN202310323741.2A CN202310323741A CN116402574A CN 116402574 A CN116402574 A CN 116402574A CN 202310323741 A CN202310323741 A CN 202310323741A CN 116402574 A CN116402574 A CN 116402574A
- Authority
- CN
- China
- Prior art keywords
- data
- order
- real
- information
- representing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 claims abstract description 20
- 239000000463 material Substances 0.000 claims description 24
- 241000255581 Drosophila <fruit fly, genus> Species 0.000 claims description 18
- 239000003086 colorant Substances 0.000 claims description 14
- 230000006870 function Effects 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000003062 neural network model Methods 0.000 claims description 6
- 241000255588 Tephritidae Species 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 230000006872 improvement Effects 0.000 claims description 3
- 230000006386 memory function Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims 1
- 230000008901 benefit Effects 0.000 abstract description 5
- 238000010586 diagram Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004064 recycling Methods 0.000 description 1
- 238000005507 spraying Methods 0.000 description 1
- 238000004148 unit process Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0633—Lists, e.g. purchase orders, compilation or processing
- G06Q30/0635—Processing of requisition or of purchase orders
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- General Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Development Economics (AREA)
- Manufacturing & Machinery (AREA)
- Human Resources & Organizations (AREA)
- Primary Health Care (AREA)
- Tourism & Hospitality (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a flexible intelligent manufacturing method and a system according to live information feedback, wherein the method comprises the following steps: collecting real-time data, and reading and recording real-time order data; classifying real-time order data by means of hierarchical retrieval; an effective order prediction model based on the IFOA-LSTM is constructed, and the effective order quantity is predicted according to the real-time order data and the historical order data which are processed in a classified mode; intelligent production is carried out according to the prediction result of the effective amount of orders, and product information of finished production is recorded; and marking the finished product information as historical data, and inputting the historical data as an effective order prediction model again, so that the prediction accuracy of the effective order prediction model is improved. The invention can avoid blind production of products or economic loss caused by unsubscribing of users, and can transmit products with different requirements to respective production lines, thereby improving the order and the efficiency of the production process and effectively improving the economic benefit.
Description
Technical Field
The invention belongs to the technical field of intelligent manufacturing, and particularly relates to a flexible intelligent manufacturing method and system based on live broadcast information feedback.
Background
With the rise of a new technological revolution, the traditional manufacturing industry faces a huge impact, the problems of weak innovation capability, low technological content of products, few high-end equipment and the like are gradually exposed, the traditional manufacturing industry is urgently required to be adjusted and upgraded, and the intelligent manufacturing industry is gradually a new direction of industry development. At present, the traditional advantages of the manufacturing industry in China are gradually weakened. For the current manufacturing industry, the problems of excessive production and complicated post-production treatment process of products are unavoidable, and a large amount of manpower and material resources are consumed.
Disclosure of Invention
The invention aims to: the invention provides a flexible intelligent manufacturing method and a system for feeding back according to live broadcast information, which directly acquire order information through live broadcast, process the order information and predict the effective order quantity; the intelligent processing is carried out on the product according to the effective order data, so that not only can the economic loss caused by blind production and user unsubscribing be avoided, but also the production process becomes fast and advanced, the time cost and the cost are effectively solved, and the net income is improved.
The technical scheme is as follows: the invention provides a flexible intelligent manufacturing method based on live information feedback, which specifically comprises the following steps:
(1) Collecting real-time data, and reading and recording real-time order data;
(2) Classifying real-time order data by means of hierarchical retrieval;
(3) An effective order prediction model based on the IFOA-LSTM is constructed, and the effective order quantity is predicted according to the real-time order data and the historical order data which are processed in a classified mode;
(4) Intelligent production is carried out according to the prediction result of the effective amount of orders, and product information of finished production is recorded; and marking the finished product information as historical data, and inputting the historical data as an effective order prediction model again, so that the prediction accuracy of the effective order prediction model is improved.
Further, the real-time data in the step (1) is acquired through a live broadcast platform.
Further, the implementation process of the step (2) is as follows:
classifying the product data of various different materials, specifications and colors, and classifying the product data by taking the materials of orders as key information for the first time, wherein the product data are divided into a plurality of groups of order math of different materials; carrying out intra-group classification on orders of different materials by taking the specifications of products as key information to obtain a plurality of groups of order data with different specifications; and then adopting the same method for orders of various different specifications, and carrying out intra-group classification by taking the color of the product as key information to obtain product order data of different colors.
Further, the historical order data in the step (3) refers to the amount of orders and the amount of orders of products with different materials, different specifications and different colors.
Further, the implementation process of constructing the effective order prediction model based on the IFOA-LSTM in the step (3) is as follows:
s1: constructing an LSTM neural network model, and constructing three modules with memory functions, namely an input gate, an output gate and a forget gate of the LSTM in the neural network model;
s2: forgetting door f t Is responsible for deciding which information to discard from the memory unit, and updates the formula as follows:
f t =σ(w fx x t +w fh h t-1 +w fc C t-1 +b f )
wherein, sigma (·) -sigmoid: x is x t Information representing a t-th order; h is a t Representing the t < th > forecast valid order information; h is a t-1 Representing t-1 forecast valid order information, C t A candidate vector representing time t; w (w) fx 、w fh 、w fc Weight coefficients representing forgetting gates; b f Representing a forget gate bias;
s3: output gate i t Is responsible for deciding which information can be stored in the memory unit, and updating the formula as follows:
i t =σ(w ix x i +w ih h t-1 +w ic C t-1 +b i )
wherein: w (w) ix 、w ih 、w ic A weight coefficient representing an input gate; b i Representing input gate bias; w (w) cx 、w ch A weight coefficient representing the candidate vector; b 0 Representing candidate vector bias; tanh () represents a hyperbolic tangent activation function;an updated value representing the candidate vector;
s4: the output gate decides which information to output, expressed in the following way:
o t =σ(w ox x t +w oh h t-1 +w oc C t-1 +b o )
h t =o t tanh(C t )
wherein: o (o) t Representing an output gate; w (w) ox 、w oh 、w oc A weight coefficient representing an output gate; the real-time order data and the historical order data after intelligent classification are used as the input of a prediction model and are output as effective order information;
s5: the method comprises the steps of optimizing the number of initial hidden layer nodes and the learning rate of an LSTM prediction model by adopting an improved drosophila optimization algorithm IFOA, and finding out the optimal values of the number of the initial hidden layer nodes and the learning rate, wherein the specific implementation process is as follows:
1) Initializing Drosophila population scale Sizepop, maximum iteration number Maxgen and Drosophila population position X axis And Y axis ;
2) The random search direction and distance are given to the drosophila individuals, and the calculation formula is as follows:
3) Optimizing and improving the search step length:
R=α×e -(β×g)/Margen
wherein alpha is a step control factor, beta is an exponential regulation factor, g is the current iteration number, and Maxge is the maximum iteration number;
4) After improvement, the fruit fly individual updates the position:
5) Calculating the distance Dist from the drosophila individual to the origin:
6) Calculating taste concentration determination value S i The calculation formula is as follows:
S i =1/Dist i
7) Determining the concentration of taste by sign function i And (3) optimizing and improving:
S i =sign(2×rand-1)/Dist
wherein, rand is a random number uniformly distributed in the range of 0, 1;
8) Inputting the concentration determination value into an objective function, and calculating a taste concentration value Smell i The calculation formula is as follows:
Smell i =Fitness(S i )
wherein Fitness represents an objective function for calculating a taste intensity value;
9) Obtaining the Drosophila individual with the optimal taste concentration value, and recording the position information and the response taste concentration value, wherein the formula is as follows:
[bestSmell,bestindex]=min(Smell)
10 Preserving the optimal taste concentration value bestshell, performing a location update to form a new population center:
Smellbest=bestSmell
11 And (3) iteratively optimizing until the maximum iteration times are met, and outputting the optimal initial hidden layer node number and the learning rate.
Based on the same inventive concept, the invention also provides a flexible intelligent manufacturing system based on live broadcast information feedback, which comprises a real-time data acquisition unit, an intelligent classification unit, an effective order number prediction unit and an intelligent manufacturing unit; the real-time data participation unit collects real-time data through the live broadcast platform, reads and records order data, and transmits the recorded data to the intelligent classification unit; the intelligent classification unit classifies the order data by a hierarchical search mode, classifies the product data of various different materials, specifications and colors, and provides a data source for the later-stage effective order prediction unit; the effective order predicting unit predicts the effective order quantity according to the real-time order data processed by the intelligent classifying unit and the historical order data; and the intelligent manufacturing unit performs intelligent production according to the prediction result of the effective order prediction unit.
The beneficial effects are that: compared with the prior art, the invention has the beneficial effects that:
1. according to the method, the order information is directly obtained through live broadcast, intelligent classification processing is carried out on the order data, the quantity of the effective orders is predicted by combining the historical order data, economic loss caused by excessive production or unsubscribe of users is avoided, and economic benefits are effectively improved;
2. the intelligent manufacturing unit can process products according to materials, specifications and colors, and can output the products with the same materials, specifications and colors at the same delivery port, so that the products can be classified rapidly and efficiently, the complex manual classification process in the later stage of the products is avoided, and the economic cost and the time cost are saved;
3. the invention can realize the recycling of data, record the product information which is finished in production, mark the finished product information as the history data, and take the history data as the input of the effective order prediction model based on the IFOA-LSTM constructed by the invention in the transmitted history order database, thereby improving the accuracy of the effective order prediction.
Drawings
FIG. 1 is a flow chart of a flexible intelligent manufacturing method based on live information feedback;
FIG. 2 is a schematic diagram of the intelligent classification unit process according to the present invention;
FIG. 3 is a schematic diagram of an intelligent manufacturing process according to the present invention;
FIG. 4 is a schematic diagram of a flexible intelligent manufacturing system according to live information feedback;
FIG. 5 is a graph of predicted results using the present invention;
FIG. 6 is a graph of net benefit versus the present invention.
Detailed Description
The invention is described in further detail below with reference to the accompanying drawings.
The invention provides a flexible intelligent manufacturing method according to live information feedback, which is shown in fig. 1 and specifically comprises the following steps:
step 1: collecting real-time data through a live broadcast platform, and reading and recording real-time order data; and classifying the real-time order data by means of hierarchical retrieval.
As shown in fig. 2, sorting is performed on order data by means of hierarchical search, and the first search is performed by using the materials of the order as key information, so that order math of different materials is divided into a plurality of groups; carrying out intra-group classification on orders of different materials by taking the specifications of products as key information to obtain a plurality of groups of order data with different specifications; then adopting the same method for orders of various different specifications, and carrying out intra-group classification by taking the color of the product as key information to obtain product order data of different colors; by the processing, the product data of various different materials, specifications and colors can be classified, and a data source is provided for a later-period effective order number prediction unit.
Step 2: and constructing an effective order prediction model based on the IFOA-LSTM, and predicting the effective order quantity according to the real-time order data and the historical order data which are processed in a classified mode.
The prediction of the effective order quantity is carried out according to the processed real-time order data and the history order data, so that economic loss caused by excessive production of products and unsubscribe of users can be effectively avoided. The historical order data refers to the amount of orders and the amount of orders of products with different materials, different specifications and different colors.
The implementation process for establishing the IFOA-LSTM effective order number prediction model is as follows:
s1: and constructing an LSTM neural network model, and constructing three modules with memory functions, namely an input gate, an output gate and a forgetting gate of the LSTM in the neural network model.
S2: forgetting door f t Is responsible for deciding which information to discard from the memory unit, and updates the formula as follows:
f t =σ(w fx x t +w fh h t-1 +w fc C t-1 +b f )
wherein, sigma (·) -sigmoid: x is x t Information representing a t-th order; h is a t Representing the t < th > forecast valid order information; h is a t-1 Representing t-1 forecast valid order information, C t A candidate vector representing time t; w (w) fx 、w fh 、w fc Weight coefficients representing forgetting gates; b f Indicating a forgetting gate bias.
S3: output gate i t Is responsible for deciding which information can be stored in the memory unit, and updating the formula as follows:
i t =σ(w ix x i +w ih h t-1 +w ic C t-1 +b i )
wherein: w (w) ix 、w ih 、w ic A weight coefficient representing an input gate; b i Representing input gate bias; w (w) cx 、w ch A weight coefficient representing the candidate vector; b 0 Representing candidate vector bias; tanh () represents a hyperbolic tangent activation function;representing the updated value of the candidate vector.
S4: the output gate decides which information to output, expressed in the following way:
o t =σ(w ox x t +w oh h t-1 +w oc C t-1 +b o )
h t =o t tanh(C t )
wherein: o (o) t Representing an output gate; w (w) ox 、w oh 、w oc A weight coefficient representing an output gate; and taking the real-time order data and the historical order data after intelligent classification as inputs of a prediction model, and outputting the inputs as effective order information.
S5: the method comprises the steps of optimizing the number of initial hidden layer nodes and the learning rate of an LSTM prediction model by adopting an improved drosophila optimization algorithm IFOA, and finding out the optimal values of the number of the initial hidden layer nodes and the learning rate, wherein the specific implementation process is as follows:
1) Initializing Drosophila population scale Sizepop, maximum iteration number Maxgen and Drosophila population position X axis And Y axis 。
2) The random search direction and distance are given to the drosophila individuals, and the calculation formula is as follows:
3) Optimizing and improving the search step length:
R=α×e -(β×g)/Margen
wherein alpha is a step control factor, beta is an exponential regulation factor, g is the current iteration number, and Maxge is the maximum iteration number.
4) After improvement, the fruit fly individual updates the position:
5) Calculating the distance Dist from the drosophila individual to the origin:
6) Calculating taste concentration determination value S i The calculation formula is as follows:
S i =1/Dist i
7) Determining the concentration of taste by sign function i And (3) optimizing and improving:
S i =sign(2×rand-1)/Dist
wherein rand is a random number ranging between [0,1] uniformly distributed.
8) Inputting the concentration determination value into an objective function, and calculating a taste concentration value Smell i The calculation formula is as follows:
Smell i =Fitness(S i )
wherein Fitness represents the objective function of calculating the taste intensity value.
9) Obtaining the Drosophila individual with the optimal taste concentration value, and recording the position information and the response taste concentration value, wherein the formula is as follows:
[bestSmell,bestindex]=min(Smell)
10 Preserving the optimal taste concentration value bestshell, performing a location update to form a new population center:
Smellbest=bestSmell
11 And (3) iteratively optimizing until the maximum iteration times are met, and outputting the optimal initial hidden layer node number and the learning rate.
Step 3: intelligent production is carried out according to the prediction result of the effective amount of orders, and product information of finished production is recorded; and marking the finished product information as historical data, and inputting the historical data as an effective order prediction model again, so that the prediction accuracy of the effective order prediction model is improved.
As shown in fig. 3, the intelligent production process mainly comprises material selection, specification price and color spraying, and is respectively carried out in a material classification section, a product specification processing section and a product color processing section of the assembly line; the production mode adopts a mode of batch division and classification to carry out intelligent production, the production track is provided with a plurality of gears which can move, products with different requirements are transmitted to respective production lines according to received order orders, and the same product outlets of the same batch, the same material, the same specification and the same color are ensured.
The intelligent production can record the finished product information, mark the finished product information as historical data, and take the historical data as input of an effective order prediction model in a transmitted historical order database, so that the prediction accuracy is effectively improved.
Based on the same inventive concept, the invention also provides a flexible intelligent manufacturing system according to live information feedback, as shown in fig. 4, comprising a real-time data acquisition unit, an intelligent classification unit, an effective order number prediction unit and an intelligent manufacturing unit; the real-time data participation unit collects real-time data through the live broadcast platform, reads and records order data, and transmits the recorded data to the intelligent classification unit; the intelligent classification unit classifies the order data in a hierarchical search mode, classifies the product data of various different materials, specifications and colors, and provides data sources for the later-stage effective order prediction unit; the effective order predicting unit predicts the effective order quantity according to the real-time order data processed by the intelligent classifying unit and the historical order data; and the intelligent manufacturing unit performs intelligent production according to the prediction result of the effective order prediction unit.
As shown in fig. 5, taking data of the month of the year 2023 as an example, the average daily effective order quantity of the month of the year is about 595, and the average daily effective order quantity of the predicted result is 604, the error is only 1.47%, and the predicted result is accurate and can be used as a reliable basis for the post-production.
As shown in FIG. 6, the average net gain per quarter is about 6.625 ten thousand yuan in the traditional production mode, while the average net gain per quarter is about 8.4 ten thousand yuan in the intelligent manufacturing mode, so that the economic benefit is effectively improved.
Claims (6)
1. The flexible intelligent manufacturing method based on live information feedback is characterized by comprising the following steps of:
(1) Collecting real-time data, and reading and recording real-time order data;
(2) Classifying real-time order data by means of hierarchical retrieval;
(3) An effective order prediction model based on the IFOA-LSTM is constructed, and the effective order quantity is predicted according to the real-time order data and the historical order data which are processed in a classified mode;
(4) Intelligent production is carried out according to the prediction result of the effective amount of orders, and product information of finished production is recorded; and marking the finished product information as historical data, and inputting the historical data as an effective order prediction model again, so that the prediction accuracy of the effective order prediction model is improved.
2. The flexible intelligent manufacturing method according to the live information feedback of claim 1, wherein the real-time data of step (1) is acquired through a live platform.
3. The flexible intelligent manufacturing method according to the live information feedback of claim 1, wherein the implementation process of the step (2) is as follows:
classifying the product data of various different materials, specifications and colors, and classifying the product data by taking the materials of orders as key information for the first time, wherein the product data are divided into a plurality of groups of order math of different materials; carrying out intra-group classification on orders of different materials by taking the specifications of products as key information to obtain a plurality of groups of order data with different specifications; and then adopting the same method for orders of various different specifications, and carrying out intra-group classification by taking the color of the product as key information to obtain product order data of different colors.
4. The flexible intelligent manufacturing method according to live information feedback of claim 1, wherein the historical order data in the step (3) refers to orders and orders of products with different materials, different specifications and different colors.
5. The flexible intelligent manufacturing method according to the live information feedback of claim 1, wherein the construction of the effective order prediction model based on the IFOA-LSTM in the step (3) is implemented as follows:
s1: constructing an LSTM neural network model, and constructing three modules with memory functions, namely an input gate, an output gate and a forget gate of the LSTM in the neural network model;
s2: forgetting door f t Is responsible for deciding which information to discard from the memory unit, and updates the formula as follows:
f t =σ(w fx x t +w fh h t-1 +w fc C t-1 +b f )
wherein, sigma (·) -sigmoid: x is x t Information representing a t-th order; h is a t Representing the t < th > forecast valid order information; h is a t-1 Representing t-1 forecast valid order information, C t A candidate vector representing time t; w (w) fx 、w fh 、w fc Weight coefficients representing forgetting gates; b f Representing a forget gate bias;
s3: output gate i t Is responsible for deciding which information can be stored in the memory unit, and updating the formula as follows:
i t =σ(w ix x i +w ih h t-1 +w ic C t-1 +b i )
wherein: w (w) ix 、w ih 、w ic A weight coefficient representing an input gate; b i Representing input gate bias; w (w) cx 、w ch A weight coefficient representing the candidate vector; b 0 Representing candidate vector bias; tanh () represents a hyperbolic tangent activation function;an updated value representing the candidate vector;
s4: the output gate decides which information to output, expressed in the following way:
o t =σ(w ox x t +w oh h t-1 +w oc C t-1 +b o )
h t =o t tanh(C t )
wherein: o (o) t Representing an output gate; w (w) ox 、w oh 、w oc A weight coefficient representing an output gate; the real-time order data and the historical order data after intelligent classification are used as the input of a prediction model and are output as effective order information;
s5: the method comprises the steps of optimizing the number of initial hidden layer nodes and the learning rate of an LSTM prediction model by adopting an improved drosophila optimization algorithm IFOA, and finding out the optimal values of the number of the initial hidden layer nodes and the learning rate, wherein the specific implementation process is as follows:
1)initializing Drosophila population scale Sizepop, maximum iteration number Maxgen and Drosophila population position X axis And Y axis ;
2) The random search direction and distance are given to the drosophila individuals, and the calculation formula is as follows:
3) Optimizing and improving the search step length:
R=α×e -(β×g)/Margen
wherein alpha is a step control factor, beta is an exponential regulation factor, g is the current iteration number, and Maxge is the maximum iteration number;
4) After improvement, the fruit fly individual updates the position:
5) Calculating the distance Dist from the drosophila individual to the origin:
6) Calculating taste concentration determination value S i The calculation formula is as follows:
S i =1/Dist i
7) Determining the concentration of taste by sign function i And (3) optimizing and improving:
S i =sign(2×rand-1)/Dist
wherein, rand is a random number uniformly distributed in the range of 0, 1;
8) Inputting the concentration determination value into an objective function to calculate the tasteLane concentration value Smell i The calculation formula is as follows:
Smell i =Fitness(S i )
wherein Fitness represents an objective function for calculating a taste intensity value;
9) Obtaining the Drosophila individual with the optimal taste concentration value, and recording the position information and the response taste concentration value, wherein the formula is as follows:
[bestSmell,bestindex]=min(Smell)
10 Preserving the optimal taste concentration value bestshell, performing a location update to form a new population center:
Smellbest=bestSmell
11 And (3) iteratively optimizing until the maximum iteration times are met, and outputting the optimal initial hidden layer node number and the learning rate.
6. A flexible intelligent manufacturing system according to live information feedback employing the method of any one of claims 1 to 5, comprising a real-time data acquisition unit, an intelligent classification unit, an effective order number prediction unit, and an intelligent manufacturing unit; the real-time data participation unit collects real-time data through the live broadcast platform, reads and records order data, and transmits the recorded data to the intelligent classification unit; the intelligent classification unit classifies the order data by a hierarchical search mode, classifies the product data of various different materials, specifications and colors, and provides a data source for the later-stage effective order prediction unit; the effective order predicting unit predicts the effective order quantity according to the real-time order data processed by the intelligent classifying unit and the historical order data; and the intelligent manufacturing unit performs intelligent production according to the prediction result of the effective order prediction unit.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310323741.2A CN116402574A (en) | 2023-03-29 | 2023-03-29 | Flexible intelligent manufacturing method and system based on live broadcast information feedback |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310323741.2A CN116402574A (en) | 2023-03-29 | 2023-03-29 | Flexible intelligent manufacturing method and system based on live broadcast information feedback |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116402574A true CN116402574A (en) | 2023-07-07 |
Family
ID=87013627
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310323741.2A Pending CN116402574A (en) | 2023-03-29 | 2023-03-29 | Flexible intelligent manufacturing method and system based on live broadcast information feedback |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116402574A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117455208A (en) * | 2023-12-25 | 2024-01-26 | 苏州特铭精密科技有限公司 | Injection molding production scheduling optimization method and system based on artificial intelligence |
-
2023
- 2023-03-29 CN CN202310323741.2A patent/CN116402574A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117455208A (en) * | 2023-12-25 | 2024-01-26 | 苏州特铭精密科技有限公司 | Injection molding production scheduling optimization method and system based on artificial intelligence |
CN117455208B (en) * | 2023-12-25 | 2024-03-12 | 苏州特铭精密科技有限公司 | Injection molding production scheduling optimization method and system based on artificial intelligence |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Hassan et al. | A hybrid of multiobjective Evolutionary Algorithm and HMM-Fuzzy model for time series prediction | |
CN116596044B (en) | Power generation load prediction model training method and device based on multi-source data | |
CN114678080B (en) | Converter end point phosphorus content prediction model, construction method and phosphorus content prediction method | |
CN111008726B (en) | Class picture conversion method in power load prediction | |
CN113449919B (en) | Power consumption prediction method and system based on feature and trend perception | |
CN113705877A (en) | Real-time monthly runoff forecasting method based on deep learning model | |
CN114386324A (en) | Ultra-short-term wind power segmented prediction method based on turning period identification | |
CN111950195B (en) | Project progress prediction method based on portrait system and depth regression model | |
CN116402574A (en) | Flexible intelligent manufacturing method and system based on live broadcast information feedback | |
CN112434848A (en) | Nonlinear weighted combination wind power prediction method based on deep belief network | |
CN111476677A (en) | Big data-based electricity consumption type electricity sales quantity analysis and prediction method and system | |
CN111461404A (en) | Short-term load and hydropower prediction method based on neural network prediction interval | |
CN115907611B (en) | Fitting inventory control method based on fitting market value | |
CN115423145A (en) | Photovoltaic power prediction method based on Kmeans-VMD-WT-LSTM method | |
CN116187835A (en) | Data-driven-based method and system for estimating theoretical line loss interval of transformer area | |
CN113205223A (en) | Electric quantity prediction system and prediction method thereof | |
CN115204035A (en) | Generator set operation parameter prediction method and device based on multi-scale time sequence data fusion model and storage medium | |
CN115907195A (en) | Photovoltaic power generation power prediction method, system, electronic device and medium | |
CN116415177A (en) | Classifier parameter identification method based on extreme learning machine | |
CN115034430A (en) | Carbon emission prediction method, device, terminal and storage medium | |
CN116739118A (en) | Power load prediction method for realizing error correction mechanism based on LSTM-XGBoost | |
CN112819256A (en) | Convolution time sequence room price prediction method based on attention mechanism | |
CN112508286A (en) | Short-term load prediction method based on Kmeans-BilSTM-DMD model | |
CN113537469A (en) | Urban water demand prediction method based on LSTM network and Attention mechanism | |
CN110837929A (en) | Least square support vector machine electricity utilization prediction method based on adaptive genetic algorithm |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |