CN108649112B - LED product yield optimization method - Google Patents

Abstract

The present invention relates to a method for optimizing the yield of LED products, comprising: obtaining a preset requirement of a first product; selecting raw materials through a material recommendation model and the preset requirement of the first product; wherein the material recommendation model is used to characterize the corresponding relationship between the material and the first product parameter; and obtaining the LED product according to the raw materials. The material processing method provided in this embodiment can quickly and effectively find the most suitable raw materials through the material recommendation model, and realizes the prediction of the color rendering of LED white light under different phosphor ratios through the material ratio recommendation model, thereby eliminating the influence of human experience, reducing losses, and improving timeliness.

Description

A kind of LED product yield optimization method

technical field

The invention belongs to the technical field of artificial intelligence, and particularly relates to a method for optimizing the yield rate of LED products.

Background technique

As the fastest-growing light source product in recent years, LED is bound to replace most traditional lighting in the near future and become the mainstream product in the lighting market. Among them, white LEDs have the advantages of energy saving, environmental protection, small size and long lighting time. They are widely used in many fields such as automotive lighting and indoor lighting, and are the most promising next-generation solid-state light sources. This not only brings a broad market space to white LED packaging factories, but also puts forward higher requirements for the quality level of white LED packaging.

At present, the white light LED packaging process needs to go through multiple links such as R&D design, raw material selection, solid welding assembly, phosphor powder ratio, dispensing packaging, spectroscopic inspection, tape packaging, etc. Among them, raw material selection and phosphor powder ratio will be affected. Product luminous performance and production yield have an important impact, and the existing raw material selection and phosphor ratio are generally determined by R&D personnel based on their own experience, supplemented by a small amount of trial production experimental data, and only need to meet acceptable standards. It can be judged to pass and put into production.

The existing white LED packaging process can basically meet the current production needs of most enterprises, but there are still some problems, such as the selection of raw materials, the determination of formulas, and the setting of process parameters, etc., which are affected by R&D personnel. It is difficult to effectively guarantee the accuracy and optimality of the results. On the other hand, the determination of the packaging process requires repeated experiments, which has a large cost loss and poor timeliness.

SUMMARY OF THE INVENTION

In order to solve the above problems existing in the prior art, the present invention provides a method for optimizing the yield rate of LED products. The technical problem to be solved by the present invention is realized by the following technical solutions:

A method for optimizing LED product yield, comprising:

Obtain the preset requirements of the first product;

The original material is selected through the material recommendation model and the preset requirements of the first product; wherein, the material recommendation model is used to characterize the corresponding relationship between the material and the parameters of the first product;

LED products are obtained according to the original materials.

In an embodiment of the present invention, the first product parameter includes one or more of color temperature, wavelength, and yield parameters.

In an embodiment of the present invention, establishing the material recommendation model includes:

Collect parameter data of multiple groups of materials;

obtaining the first product parameter data corresponding to each group of materials;

Matching the multiple sets of material parameter data with the corresponding multiple sets of the first product parameter data to obtain the material recommendation model.

In an embodiment of the present invention, the obtaining the first product parameter data corresponding to each group of materials includes:

establishing a material yield relationship curve for different types of products; wherein, the different types of products are produced by each group of materials;

The first product parameter corresponding to each group of materials is acquired through the material yield relationship curve.

In an embodiment of the present invention, the material yield relationship curve is a curve between the different types of products and corresponding yields.

The present invention also provides another method for optimizing the yield rate of LED products, including:

Obtain the preset requirements of the second product;

Obtain the ratio of raw materials through the ratio recommendation model and the preset requirements of the second product; wherein, the ratio recommendation model is used to characterize the corresponding relationship between the material ratio and the second product parameter;

LED products are obtained according to the ratio of the original materials.

In an embodiment of the present invention, the second product preset requirement includes the original material and the first product parameter;

Wherein, the first product parameter includes one or more of color temperature, wavelength, and yield parameters.

In an embodiment of the present invention, establishing the proportioning recommendation model includes:

Collect multiple sets of material ratio data;

Obtain the second product parameter data corresponding to each group of material ratios;

Matching the multiple sets of material proportioning data with multiple sets of the second product parameter data to obtain the proportioning recommendation model.

In an embodiment of the present invention, the acquiring the second product parameter data corresponding to each group of material ratios includes:

Establishing a ratio yield relationship curve of products with different packaging modes; wherein, the products in different packaging modes are produced by the ratio of each group of materials;

The second product parameter is obtained through the proportioning yield relationship curve.

In an embodiment of the present invention, after the LED product is obtained according to the ratio of the original materials, it further includes:

Collect product parameter data of the LED product;

comparing the LED product parameter data with the second product parameter data;

If the LED product yield is higher than the yield corresponding to the second product parameter, the LED product parameter is used as the second product parameter.

Compared with the prior art, the beneficial effects of the present invention:

1) Through the material recommendation model, the most suitable raw materials can be found quickly and effectively, eliminating the influence of human experience and ensuring the accuracy and optimality of the results;

2) Through the material ratio recommendation model, the prediction of the color rendering of LED white light under different phosphor powder ratios is realized, which reduces the process of trial and error, reduces the loss, and improves the timeliness.

Description of drawings

FIG. 1 is a schematic diagram of an optimization flow through a material recommendation model of a method for optimizing the yield of an LED product provided by an embodiment of the present invention;

FIG. 2 is a schematic diagram of an optimization flow through a ratio recommendation model of an LED product yield optimization method provided by an embodiment of the present invention;

FIG. 3 is a schematic flow chart of establishing a material recommendation model of a method for optimizing LED product yield according to an embodiment of the present invention;

FIG. 4 is a schematic flow chart of establishing a ratio recommendation model of an LED product yield optimization method according to an embodiment of the present invention;

FIG. 5 is a schematic flowchart of a method for optimizing the yield of an LED product according to an embodiment of the present invention.

Detailed ways

The present invention will be described in further detail below with reference to specific embodiments, but the embodiments of the present invention are not limited thereto.

Example 1

Please refer to FIG. 1 and FIG. 3 . FIG. 1 is a schematic diagram of an optimization flow through a material recommendation model of an LED product yield optimization method provided by an embodiment of the present invention; FIG. 3 is an LED product yield optimization process provided by an embodiment of the present invention. Schematic diagram of the process of establishing a material recommendation model for the method. The yield rate is the so-called pass rate, which is obtained by dividing the number of finished products shipped by the total number of shipments. The RGB color mode is a color standard in the industry. It obtains various colors by changing the three color channels of red (R), green (G), and blue (B) and superimposing them on each other. Yes, RGB is the color representing the three channels of red, green, and blue. When their lights are superimposed on each other, the colors are mixed, and the more mixed, the higher the brightness. For the superposition of red, green, and blue lights, the brightest superposition area of the central three colors is white, and the more superimposed, the brighter. The existing white LED packaging process mainly relies on the principle of RGB color rendering. First, the mixture of yellow and green phosphors is mixed to form a mixture that displays yellow after stress, and then is packaged with a blue chip. In this process, the color rendering of the final white light can be controlled by controlling the selection of the light source material.

As shown in FIG. 1 , a method for optimizing the yield rate of LED products includes obtaining preset requirements of a first product; selecting original materials through a material recommendation model and the preset requirements of the first product; wherein, the material recommendation model is used for Characterize the corresponding relationship between the material and the first product parameter; obtain the LED product according to the original material.

Preferably, before producing a new white LED product, parameters such as the color temperature and wavelength of the white light that need to be produced will generally be determined. The first product preset requirements include the white light color temperature, wavelength, etc. of the new white LED product to be produced. parameter.

Preferably, the material recommendation model is called, and the parameters and constraints of the product to be produced are input, that is, the preset requirements and constraints of the first product are input into the material recommendation model, and the model will output the original materials of the product to be produced. Wherein, the product parameter corresponding to the raw material in the material recommendation model, that is, the first product parameter, is the product parameter of the product with the highest yield among the products of the same type produced by the material, and the first product parameter data and its corresponding Material data is historical data, obtained through big data processing.

Preferably, as shown in Figure 4, establishing the material recommendation model includes the following steps:

Step 1. Obtain historical material data, and preprocess historical recipe data;

In the production of white LED light source products, the raw materials required include red phosphors, green phosphors, A/B glue, anti-precipitation powder and other raw materials. Among them, these raw materials have many different specifications and models, such as phosphors produced by different manufacturers. phosphors of different specifications. These raw material combinations of different specifications and models can produce different white LED light source products, and one group of raw material combinations that can produce white light LED light source products is called a group of raw materials, wherein the historical materials include multiple groups of raw materials. Different types of white light LED products can be produced through a set of raw materials, wherein the product parameters of these different types of white light LED products are also different.

Preferably, obtaining historical material data includes obtaining multiple sets of raw material data, that is, obtaining multiple sets of material parameter data, and then obtaining product parameter data of different products produced by each set of raw materials, and storing these data, wherein the raw material data includes specifications , model, name and other data, product parameter data including product color, wavelength, service life and other data. Preprocessing the historical formula data, that is, extracting characteristic attributes from the raw material combination data and product parameter data, wherein the characteristic attributes of raw materials include specifications and models, and the characteristic attributes of product parameters include color temperature and wavelength.

Step 2. Analyze data to obtain product parameters corresponding to raw materials;

Integrate the product parameter data of different products produced from the same set of raw materials, analyze the proportion of the same color product in the total number of products, analyze the proportion of the number of products in each wavelength band in the total number of products, and Analyze data such as yield rate, find out that the group of raw materials is better at making products through data such as color rendering, wavelength, yield, etc., and use the product parameters of the group of raw materials that are better at making products as the first product parameter corresponding to the group of raw materials, That is, the material yield relationship curve of different products is established; the first product parameter corresponding to each group of materials is obtained through the material yield relationship curve. The material yield relationship curve is a curve between the different types of products and corresponding yields. Feature attributes of raw materials and product parameters are extracted and stored.

Step 3. Establish a material recommendation model;

Repeat step 2, analyze and find out the products produced by multiple groups of raw materials, and find out the product parameters that each group of raw materials is good at making products, and match each group of raw materials with the product parameters that are good at making products, establish a matching relationship, and establish Material recommendation model. Wherein, each group of raw materials corresponds to a group of product parameters, or may be multiple groups of product parameters arranged in a preferred order.

The material processing method provided in this embodiment can quickly and effectively find the most suitable raw material through the material recommendation model, eliminates the influence of human experience, and ensures the accuracy and optimality of the results.

Embodiment 2

Please refer to FIG. 2 and FIG. 4 . FIG. 2 is a schematic diagram of an optimization flow of a method for optimizing the yield rate of an LED product provided by an embodiment of the present invention through a ratio recommendation model; FIG. 4 is a schematic diagram of an LED product yield rate provided by an embodiment of the present invention. Schematic diagram of the flow chart of establishing a proportioning recommendation model for the optimization method. The RGB color mode is a color standard in the industry. It obtains various colors by changing the three color channels of red (R), green (G), and blue (B) and superimposing them on each other. Yes, RGB is the color representing the three channels of red, green, and blue. When their lights are superimposed on each other, the colors are mixed, and the more mixed, the higher the brightness. For the superposition of red, green, and blue lights, the brightest superposition area of the central three colors is white, and the more superimposed, the brighter. The existing white LED packaging process mainly relies on the principle of RGB color rendering. First, the mixture of yellow and green phosphors is mixed to form a mixture that displays yellow after stress, and then it is packaged with a blue base. In this process, the color rendering of the final white light can be controlled by controlling the material ratio of the light source.

As shown in FIG. 2 , a method for optimizing the yield rate of LED products includes: obtaining preset requirements for a second product; obtaining the ratio of original materials through a ratio recommendation model and the preset requirements of the second product; wherein, the The ratio recommendation model is used to characterize the corresponding relationship between the material ratio and the second product parameter, and the LED product is obtained according to the ratio of the original material.

Preferably, before producing new products of white LEDs, parameters such as the color temperature and wavelength of white light that need to be produced for new products of white LEDs are generally determined, and raw materials for manufacturing products are selected. Second, the current product parameters include the white light that needs to be produced new products of white LEDs. Parameters such as color temperature, wavelength, and material parameters of raw materials.

Preferably, as shown in Figure 4, establishing a proportioning recommendation model includes the following steps:

Step 1. Obtain historical proportioning data, and preprocess the historical proportioning data;

Obtain the historical ratio data of raw materials for manufacturing white LED light source products, among which raw materials include red phosphor, green phosphor, A/B glue, anti-precipitation powder and other raw materials. The different configuration ratios of these raw materials in production can determine the product parameters for producing white LED products. The historical ratio data refers to the ratio of raw materials that can produce white LED light source products with a fixed ratio in the reproduction process, which is called a set of material ratio data, that is, the ratio data of the original materials, to obtain the historical ratio. The ratio data includes acquiring multiple sets of matching data, and then storing the matching data.

Step 2, obtain the product parameters corresponding to the raw material ratio;

Different products will be obtained by producing products in different packaging modes through a set of material ratios, and product parameters of white LED light source products produced by a set of material ratios in different packaging modes will be obtained. Through different packaging yields, color temperatures, wavelengths Wait for records, analyze the factors that cause the difference, analyze the relationship between the material ratio and product parameters according to these factors, and find out the product parameter with the highest yield as the second product parameter, that is, establish the ratio and yield relationship of products with different packaging modes curve; wherein, products of different packaging modes are produced by the ratio of each group of materials; the second product parameter is obtained through the ratio relationship curve. In the same way, the second product parameters corresponding to the ratios of multiple sets of characterization materials are obtained.

Step 3. Establish a ratio recommendation model;

A mapping relationship is established between the multiple sets of characterizing material proportioning data and their corresponding second product parameter data, and a proportioning recommendation model is established, wherein each group of material proportioning corresponds to a set of product parameters.

Preferably, after obtaining the preferred material ratio through the material ratio recommendation model, the actual consumption of each material is calculated respectively, and production is carried out according to the actual consumption of each material to obtain a white LED light source product.

Preferably, after obtaining the white light LED light source product, the product parameters of the white light LED light source product are collected and compared with the corresponding second product parameters. If the yield of the white light LED light source product is higher than the yield corresponding to the second product parameter If it is high, the product parameters of the white LED light source product are used as the second product parameters.

The material processing method provided in this embodiment realizes the prediction of the color rendering of LED white light under different phosphor powder ratios through the material ratio recommendation model, which reduces the process of repeated tests, reduces the loss, and improves the performance. Timeliness.

Embodiment 3

Please refer to FIG. 3 , FIG. 4 and FIG. 5 . FIG. 3 is a schematic flowchart of a material recommendation model establishment of a method for optimizing the yield of an LED product according to an embodiment of the present invention; FIG. 4 is a good LED product according to an embodiment of the present invention. Figure 5 is a schematic flowchart of an LED product yield optimization method according to an embodiment of the present invention. The RGB color mode is a color standard in the industry. It obtains various colors by changing the three color channels of red (R), green (G), and blue (B) and superimposing them on each other. Yes, RGB is the color representing the three channels of red, green, and blue. When their lights are superimposed on each other, the colors are mixed, and the more mixed, the higher the brightness. For the superposition of red, green, and blue lights, the brightest superposition area of the central three colors is white, and the more superimposed, the brighter. The existing white LED packaging process mainly relies on the principle of RGB color rendering. First, the mixture of yellow and green phosphors is mixed to form a mixture that displays yellow after stress, and then it is packaged with a blue base. In this process, the final white light color rendering can be controlled by controlling the selection of light source materials and the ratio of materials.

As shown in FIG. 5 , a method for optimizing the yield rate of LED products includes: obtaining preset requirements of a first product; selecting original materials through a material recommendation model and the preset requirements of the first product; wherein, the material recommendation model uses to characterize the corresponding relationship between the material and the parameters of the first product; to obtain the preset demand of the second product; to obtain the proportion of the original material through the proportioning recommendation model and the preset demand of the second product; wherein, the proportioning recommendation The model is used to characterize the corresponding relationship between the material ratio and the second product parameter. According to the current material ratio, the actual amount of the current material is obtained to obtain an LED product.

Preferably, before producing a new white LED product, parameters such as the color temperature and wavelength of the white light that need to be produced for the new white LED are generally determined, that is, the first product preset requirement, which includes the white light color temperature of the new white LED to be produced. , wavelength and other parameters, the second current product parameters include white light color temperature, wavelength and other parameters and original material parameters that need to produce new white LED products.

Preferably, the material recommendation model is called, and the parameters and constraints of the product to be produced are input, that is, the preset requirements and constraints of the first product are input into the material recommendation model, and the model will output the raw material combination of the product to be produced. Among them, it is the product parameter of the product with the highest yield among the same type of products produced by the material, and the first product parameter data and its corresponding material data are historical data, which are obtained through big data processing.

Preferably, as shown in Figure 3, establishing the material recommendation model includes the following steps:

Step 1. Obtain historical material data and preprocess historical recipe data

In the production of white LED light source products, the raw materials required include red phosphors, green phosphors, A/B glue, anti-precipitation powder and other raw materials. Among them, these raw materials have many different specifications and models, such as phosphors produced by different manufacturers. phosphors of different specifications. These raw material combinations of different specifications and models can produce different white LED light source products, and one group of raw material combinations that can produce white light LED light source products is called a group of raw materials, wherein the historical materials include multiple groups of raw materials. Different types of white light LED products can be produced through a set of raw materials, wherein the product parameters of these different types of white light LED products are also different.

Preferably, obtaining historical material data includes obtaining multiple sets of raw material data, that is, obtaining multiple sets of material parameter data, and then obtaining product parameter data of different products produced by each set of raw materials, and storing these data, wherein the raw material data includes specifications , model, name and other data, product parameter data including product color, wavelength, service life and other data. Preprocessing the historical formula data, that is, extracting characteristic attributes from the raw material combination data and product parameter data, wherein the characteristic attributes of raw materials include specifications and models, and the characteristic attributes of product parameters include color temperature and wavelength.

Step 2. Analyze data to obtain product parameters corresponding to raw materials

Integrate the product parameter data of different products produced from the same set of raw materials, analyze the proportion of the same color product in the total number of products, analyze the proportion of the number of products in each wavelength band in the total number of products, and Analyze data such as yield rate, find out that the group of raw materials is better at making products through data such as color rendering, wavelength, yield, etc., and use the product parameters of the group of raw materials that are better at making products as the first product parameter corresponding to the group of raw materials, That is, the material yield relationship curve of different products is established; the first product parameter corresponding to each group of materials is obtained through the material yield relationship curve. The material yield relationship curve is a curve between the different types of products and corresponding yields. Feature attributes of raw materials and product parameters are extracted and stored.

Step 3. Establish a material recommendation model

Repeat step 2, analyze and find out the products produced by multiple groups of raw materials, and find out the product parameters that each group of raw materials is good at making products, and match each group of raw materials with the product parameters that are good at making products, establish a matching relationship, and establish Material recommendation model. Wherein, each group of raw materials corresponds to a group of product parameters, or may be multiple groups of product parameters arranged in a preferred order.

Preferably, as shown in Figure 4, establishing a proportioning recommendation model includes the following steps:

Step 1. Obtain historical proportioning data and preprocess the historical proportioning data

Obtain the historical ratio data of raw materials for manufacturing white LED light source products, among which raw materials include red phosphor, green phosphor, A/B glue, anti-precipitation powder and other raw materials. The different configuration ratios of these raw materials in production can determine the product parameters for producing white LED products. The historical ratio data refers to the ratio of raw materials that can produce white LED light source products with a fixed ratio in the reproduction process, which is called a set of material ratio data, that is, the ratio data of the original materials, to obtain the historical ratio. Ratio data includes obtaining multiple sets of matching data. Then the proportioning data is stored.

Step 2. Obtain the product parameters corresponding to the raw material ratio

Different products will be obtained by producing products in different packaging modes through a set of material ratios, and product parameters of white LED light source products produced by a set of material ratios in different packaging modes will be obtained. Through different packaging yields, color temperatures, wavelengths Wait for records, analyze the factors that cause the difference, analyze the relationship between the material ratio and product parameters according to these factors, and find out the product parameter with the highest yield as the second product parameter, that is, establish the ratio and yield relationship of products with different packaging modes curve; wherein, products of different packaging modes are produced by the ratio of each group of materials; the second product parameter is obtained through the ratio relationship curve. In the same way, the second product parameters corresponding to the ratios of multiple sets of characterization materials are obtained.

Step 3. Establish a proportioning recommendation model

A mapping relationship is established between the multiple sets of characterizing material proportioning data and their corresponding second product parameter data, and a proportioning recommendation model is established, wherein each group of material proportioning corresponds to a set of product parameters.

Preferably, after obtaining the preferred material ratio through the material ratio recommendation model, the actual consumption of each material is calculated respectively, and production is carried out according to the actual consumption of each material to obtain a white LED light source product.

Preferably, after obtaining the white light LED light source product, the product parameters of the white light LED light source product are collected and compared with the corresponding second product parameters. If the yield of the white light LED light source product is higher than the yield corresponding to the second product parameter If it is high, the product parameters of the white LED light source product are used as the second product parameters.

The material processing method provided in this embodiment can quickly and effectively find the most suitable raw materials through the material recommendation model, and realize the prediction of the color rendering of LED white light under different phosphor powder ratios through the material ratio recommendation model. It eliminates the influence of human experience, ensures the accuracy and optimality of the results, reduces the process of trial and error, reduces the loss, and improves the timeliness.

The above content is a further detailed description of the present invention in combination with specific preferred embodiments, and it cannot be considered that the specific implementation of the present invention is limited to these descriptions. For those of ordinary skill in the technical field of the present invention, without departing from the concept of the present invention, some simple deductions or substitutions can be made, which should be regarded as belonging to the protection scope of the present invention.

Claims (3)

1. a method for optimizing the yield of LED products, characterized in that, comprising: Obtain the preset requirements of the second product; Obtain the ratio of raw materials through the ratio recommendation model and the preset requirements of the second product; wherein, the ratio recommendation model is used to characterize the corresponding relationship between the material ratio and the second product parameter; LED products are obtained according to the ratio of the original materials; Establishing the proportioning recommendation model includes: Collect multiple sets of material ratio data; Obtain the second product parameter data corresponding to each group of material ratios; Matching the multiple sets of material proportioning data with multiple sets of the second product parameter data to obtain the proportioning recommendation model; The obtaining of the second product parameter data corresponding to each group of material ratios includes: Establishing a ratio yield relationship curve of products with different packaging modes; wherein, the products in different packaging modes are produced by the ratio of each group of materials; The second product parameter is obtained through the proportioning yield relationship curve. 2. The LED product yield optimization method according to claim 1, wherein the second product preset requirement includes the original material and the first product parameter; Wherein, the first product parameter includes one or more of color temperature, wavelength, and yield parameters. 3. The LED product yield optimization method according to claim 1, wherein after obtaining the LED product according to the ratio of the original materials, the method further comprises: Collect product parameter data of the LED product; comparing the LED product parameter data with the second product parameter data; If the LED product yield is higher than the yield corresponding to the second product parameter, the LED product parameter is used as the second product parameter.

Images