Early warning method, system and equipment during industrial product production, medium and program product

Abstract

The invention discloses an early warning method, system and device during industrial product production, a medium and a program product. The early warning method comprises the steps of obtaining initial parameter data corresponding to a plurality of scrapped products; the initial parameter data comprises a plurality of pieces of production factor information; based on the initial parameter data, obtaining target scrap influence factor information of the scrap product in different production environments; the target scrap influence factor information represents environmental parameter information of different scrap products in corresponding production environments; and generating early warning rule information when the industrial product is produced in different production environments based on the multiple pieces of target scrapping influence factor information. Through the early warning method provided by the invention, the early warning information of the industrial product during production can be obtained and updated at any time, so that the production efficiency of the product is improved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the field of data processing, and in particular, to a pre-warning method, system, device, medium and program product for industrial product production. BACKGROUND

[0002] In the manufacturing process, in order to ensure product quality, it is usually necessary to monitor the operating parameters (such as temperature, pressure, speed, etc.) of the production equipment in real time, and to judge whether the product produced is qualified according to the preset threshold rules.

[0003] However, the threshold monitoring method widely used at present mainly relies on manual experience to set static rules, which has the following obvious limitations: first, the formulation of threshold rules often needs experienced engineers to manually set according to limited experience or samples, lacks driving and optimization based on large-scale historical data, resulting in strong subjectivity of the rules and difficulty in continuous evolution; second, the traditional rules are usually simple in form, mostly single-point thresholds for a single parameter (for example, temperature ≥ 90°C is judged as abnormal), which ignores the objective requirement that the parameter should be within a reasonable range, and cannot accurately capture the complex situation that both too low and too high can lead to defects; third, when a large amount of historical data containing qualified and scrapped records is accumulated, the existing method lacks effective algorithms to automatically analyze these data, so as to inversely reason out scientific and reasonable parameter threshold range intervals.

[0004] Since different product defects or scrap modes often correspond to different parameter abnormal combinations, and the existing method cannot automatically identify and distinguish these various scrap modes, it is also impossible to generate a corresponding set of fine-grained multi-parameter threshold range rules for different modes. SUMMARY

[0005] The technical problem to be solved by the present disclosure is to overcome the defect in the prior art that a corresponding set of fine-grained multi-parameter threshold range rules cannot be generated for different modes, and to provide a pre-warning method, system, device, medium and program product for industrial product production.

[0006] The present disclosure solves the above technical problems by the following technical solutions:

[0007] According to a first aspect of the present disclosure, a pre-warning method for industrial product production is provided, the pre-warning method comprising:

[0008] obtaining initial parameter data corresponding to a plurality of scrapped products; the initial parameter data comprising a plurality of production factor information;

[0009] Based on the initial parameter data, target scrap influencing factor information of the scrap product in different production environments is obtained; the target scrap influencing factor information represents environmental parameter information of different scrap products in corresponding production environments;

[0010] Based on the target scrap influencing factor information, pre-warning rule information of the industrial product produced in different production environments is generated.

[0011] Optionally, the step of obtaining the target scrap influencing factor information of the scrap product in different production environments based on the initial parameter data comprises:

[0012] Based on the initial parameter data, initial purity information is obtained;

[0013] Based on the initial parameter data, actual scrap influencing factor information under a set production environment condition is obtained;

[0014] Based on the initial purity information and the actual scrap influencing factor information, the target scrap influencing factor information is determined.

[0015] Optionally, the step of obtaining the initial purity information based on the initial parameter data comprises:

[0016] Based on a preset classification condition, the scrap products are divided into several scrap groups; wherein the preset classification condition at least includes one of mechanical failure condition, process failure condition and electrical failure condition;

[0017] First quantity information of the scrap products in different scrap groups is obtained; the first quantity information represents the quantity of the scrap products in different scrap groups;

[0018] Based on the first quantity information, the initial purity information is determined; the initial purity information represents failure degree information of the industrial product in different production environments.

[0019] Optionally, the step of obtaining the actual scrap influencing factor information under the set production environment condition based on the initial parameter data comprises:

[0020] Second quantity information of the scrap products belonging to different scrap groups under the set production environment condition is obtained; the second quantity information represents the quantity of the scrap products in different scrap groups under the set production environment condition;

[0021] Based on the second quantity information, the actual scrap influencing factor information under the set production environment condition is obtained.

[0022] Optionally, the step of determining the target scrap impact factor information based on the initial purity information and the actual scrap impact factor information comprises:

[0023] determining the target scrap impact factor information corresponding to the set production environment condition based on a difference between the initial purity information and the actual scrap impact factor information.

[0024] Optionally, the step of generating the early warning rule information of the industrial product produced in different production environments based on the target scrap impact factor information comprises:

[0025] sequentially sorting the target scrap impact factor information in different production environments from high to low, and generating the early warning rule corresponding to the production environment according to the sorting result.

[0026] According to a second aspect of the present disclosure, an early warning system for industrial product production is provided, which comprises:

[0027] a data acquisition module configured to acquire initial parameter data corresponding to a plurality of scrap products; the initial parameter data comprises a plurality of production factor information;

[0028] a scrap impact factor acquisition module configured to acquire target scrap impact factor information of the scrap products in different production environments based on a plurality of the initial parameter data; the target scrap impact factor information represents environmental parameter information of different scrap products in corresponding production environments;

[0029] an early warning classification module configured to generate early warning rule information of the industrial product produced in different production environments based on a plurality of the target scrap impact factor information.

[0030] Optionally, the scrap impact factor acquisition module comprises an initial purity acquisition unit, an actual impact factor acquisition unit, and a target impact factor acquisition unit.

[0031] The initial purity acquisition unit is configured to acquire initial purity information based on a plurality of the initial parameter data.

[0032] The actual impact factor acquisition unit is configured to acquire actual scrap impact factor information in a set production environment condition based on a plurality of the initial parameter data.

[0033] The target impact factor acquisition unit is configured to determine the target scrap impact factor information based on the initial purity information and the actual scrap impact factor information.

[0034] Optionally, the initial purity acquisition unit is configured to:

[0035] a plurality of scrap products are divided into a plurality of scrap groups based on preset classification conditions; wherein the preset classification conditions at least include one of mechanical failure conditions, process failure conditions and electrical failure conditions;

[0036] a first quantity information of the scrap products in different scrap groups is obtained; the first quantity information represents the quantity of the scrap products in different scrap groups;

[0037] the initial purity information is determined based on a plurality of the first quantity information; the initial purity information represents the failure degree information of the industrial products under different production environments.

[0038] Optionally, the actual impact factor obtaining unit is configured to:

[0039] a second quantity information of the scrap products in different scrap groups under the set production environment condition is obtained; the second quantity information represents the quantity of the scrap products in different scrap groups under the set production environment condition;

[0040] the actual scrap impact factor information under the set production environment condition is obtained based on a plurality of the second quantity information.

[0041] Optionally, the target impact factor obtaining unit is configured to:

[0042] the target scrap impact factor information corresponding to the set production environment condition is determined based on the difference between the initial purity information and the actual scrap impact factor information.

[0043] Optionally, the early warning classification module is configured to:

[0044] the target scrap impact factor information under different production environments is sequentially sorted from high to low, and early warning rules corresponding to the production environments are generated according to the sorting results.

[0045] According to a third aspect of the present disclosure, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and configured to run on the processor, and the processor executes the computer program to implement the early warning method for industrial product production according to the first aspect of the present disclosure.

[0046] According to a fourth aspect of the present disclosure, a computer readable storage medium is provided, which stores a computer program, and the computer program is executed by a processor to implement the early warning method for industrial product production according to the first aspect of the present disclosure.

[0047] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the early warning method for industrial product production according to the first aspect of the present disclosure.

[0048] On the basis of common knowledge in the art, the above-mentioned preferred conditions can be combined arbitrarily, i.e. to obtain each preferred example of the present disclosure.

[0049] The positive progress effect of the present disclosure is that:

[0050] In the early warning method for industrial product production provided by the present disclosure, the reasonable threshold range of the key production factor is automatically inferred by directly analyzing the initial parameter data of the historical scrapped products, which changes the mode of relying on manual experience to set static threshold and improves the accuracy of product production. Further, by converting the historical scrapped data into executable early warning rules and applying them to real-time monitoring and future production, the stability of product production is further improved. BRIEF DESCRIPTION OF DRAWINGS

[0051] Figure 1 A flowchart of the early warning method for industrial product production provided in Embodiment 1 of the present disclosure;

[0052] Figure 2 A flowchart of determining target scrapped impact factor information provided in Embodiment 1 of the present disclosure;

[0053] Figure 3 A flowchart of determining initial purity information provided in Embodiment 1 of the present disclosure;

[0054] Figure 4 A flowchart of determining actual scrapped impact factor information provided in Embodiment 1 of the present disclosure;

[0055] Figure 5 A structural diagram of the early warning system for industrial product production provided in Embodiment 2 of the present disclosure;

[0056] Figure 6 A structural diagram of the scrapped impact factor acquisition module provided in Embodiment 2 of the present disclosure;

[0057] Figure 7 A structural diagram of the electronic device provided in Embodiment 3 of the present disclosure. DETAILED DESCRIPTION

[0058] The present disclosure will be further illustrated by way of examples below, but the present disclosure is not limited in the scope of the examples.

[0059] The prefix words such as "first", "second" in the embodiments of the present disclosure are only used to distinguish different description objects, and have no limiting effect on the position, order, priority, quantity or content of the described objects. The use of ordinal words and other prefix words in the embodiments of the present disclosure to distinguish the described objects does not constitute a limitation on the described objects, and the description of the described objects should be referred to the description of the context in the claims or embodiments, and should not constitute an unnecessary limitation because of the use of such prefix words. In addition, in the description of the embodiments, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0060] Embodiment 1

[0061] As Figure 1 indicated, the present embodiment provides a pre-warning method during production of an industrial product, which comprises:

[0062] S11: obtaining initial parameter data corresponding to a plurality of scrapped products; the initial parameter data comprises a plurality of production factor information;

[0063] S12: based on the plurality of initial parameter data, obtaining target scrapped influence factor information of the scrapped products under different production environments; the target scrapped influence factor information represents environmental parameter information of different scrapped products under corresponding production environments;

[0064] S13: based on the plurality of target scrapped influence factor information, generating pre-warning rule information of the industrial product during production under different production environments.

[0065] In an implementation manner, the data collected based on IOT (Internet of Things) in the present embodiment is saved into a database in combination with corresponding business dimension data, and the product identification is scrapped during the reporting process; for example, each product has 5 IOT collection parameters (temperature, pressure, speed, vibration, and current).

[0066] In the pre-warning method during production of an industrial product provided by the present disclosure, the initial parameter data of the historical scrapped products is directly analyzed to automatically infer the reasonable threshold range of the key production factor, which changes the mode of relying on manual experience to set a static threshold, and improves the accuracy during production of the product; further, the historical scrapped data is converted into executable pre-warning rules, which are applied to real-time monitoring and future production, and further improve the stability during production of the product.

[0067] As Figure 2 indicated, step S12 comprises:

[0068] S121: based on the plurality of initial parameter data, obtaining initial purity information;

[0069] S122: based on the plurality of initial parameter data, obtaining actual scrapped influence factor information under a set production environment condition;

[0070] S123: determining the target scrap impact factor information based on the initial purity information and the actual scrap impact factor information.

[0071] As shown in Figure 3 , step S121 includes:

[0072] S1211: dividing a plurality of scrap products into a plurality of scrap groups based on preset classification conditions; wherein the preset classification conditions at least include one of a mechanical failure condition, a process failure condition, and an electrical failure condition;

[0073] According to the preset classification conditions related to the root cause (such as mechanical failure, process parameter deviation, electrical element failure, raw material defect, etc.), all scrap product records in the historical database are automatically classified to form a plurality of scrap groups; the corresponding classification conditions can be defined and matched based on product failure codes, maintenance work order records or expert knowledge base, thereby ensuring the business relevance of the classification.

[0074] S1212: obtaining first quantity information of corresponding scrap products in different scrap groups; the first quantity information represents the number of scrap products in different scrap groups;

[0075] S1213: determining initial purity information based on a plurality of first quantity information; the initial purity information represents the fault degree information of the industrial product under different production environments.

[0076] Wherein, the formula for calculating purity is: initial purity = 1 -∑ᵢ(Nᵢ / N_total)², wherein N_total is the total number of scrap products. This value represents the concentration or dispersion degree of the fault reason category in all scrap samples.

[0077] By converting the qualitative scrap phenomenon into a calculable purity index, an objective and quantitative benchmark is provided for subsequent analysis; the lower the purity, the more concentrated the historical scrap problem is on a certain two categories of reasons.

[0078] As shown in Figure 4 , step S122 includes:

[0079] S1221: obtaining second quantity information of scrap products belonging to different scrap groups under a set production environment condition; the second quantity information represents the number of scrap products in different scrap groups under the set production environment condition;

[0080] First, a set production environment condition (for example: a specific workshop A, a specific model B equipment, a batch with an environment humidity > 70%) is received. Then, under the constraint of this environment condition, the number of scrap products belonging to each scrap group is re-screened and counted to obtain the second quantity information.

[0081] S1222: Based on the second quantity information, the actual scrap impact factor information under the set production environment condition is obtained.

[0082] Based on the second quantity information, the actual scrap impact factor under the specific environment is calculated. The calculation method is similar to the initial purity, but the data range is different: actual scrap impact factor = 1-∑ᵢ(Mᵢ / M_total)², where M_total is the total number of scrapped products under this environment, and the factor represents the concentration degree of the scrap reason distribution under this specific environment.

[0083] By calculating the independent factors under different environments, the difference in the occurrence frequency of the same failure mode under different environments is identified, or the unique failure mode combination under a specific environment is found, which provides a basis for precise control.

[0084] Among them, step S123 includes:

[0085] Based on the difference between the initial purity information and the actual scrap impact factor information, the target scrap impact factor information corresponding to the set production environment condition is determined.

[0086] Step S13 in this embodiment includes:

[0087] The target scrap impact factor information under different production environments is sequentially sorted from high to low, and the pre-warning rules under the corresponding production environment are generated according to the sorting result.

[0088] For example: sorting: all analyzed production environments are sorted from low to high according to their corresponding target scrap impact factor values (i.e. risk from high to low); the higher the ranking, the higher the specific risk, and the more strict monitoring rules need to be developed in priority.

[0089] Rule generation: for each high-risk environment in the front of the ranking:

[0090] Identify the dominant failure mode: backtrack to analyze the one or several scrap groups with the highest proportion in the second quantity information under this environment.

[0091] Extract the key parameter threshold: for these dominant failure modes, from all related scrap product data under this environment, analyze the initial parameter data (such as temperature, pressure, etc.), and automatically fit the parameter danger range interval that leads to the failure (for example: temperature <85°C or >115°C).

[0092] Rule combination: finally generate one or more pre-warning rules. For example: IF production environment = workshop A & device model = B & environment humidity > 70% THEN pre-warning trigger condition: temperature sensor X reading <85°C OR >115°C.

[0093] The implementation principle of the early warning method in the production of industrial products in the embodiment is described below with a specific implementation.

[0094] First, prepare the data: IOT (Internet of Things) collected data and business dimension data combined data will be saved to the database, and the product identification will be scrapped during the reporting process. The data we need is the scrapped product corresponding to the collected + business dimension data.

[0095] Suppose there are 150 scrapped products, each product has 5 IOT collection parameters, and the specific product attributes are as follows:

[0096] Product 1: temperature 85°C, pressure 4.8MPa, speed 1200rpm, vibration 12mm / s, current 40A.

[0097] Product 2: temperature 95°C, pressure 5.2MPa, speed 1100rpm, vibration 8mm / s, current 45A.

[0098] Product 3: temperature 84°C, pressure 4.7MPa, speed 1210rpm, vibration 13mm / s, current 39A.

[0099] (There are 150 products in total)

[0100] Second, select the cluster center according to different fault problems. In this embodiment, mechanical failure problems, process failure problems, and electrical failure problems are selected as initial cluster centers:

[0101] Mechanical failure problem as center 1: temperature 90°C, pressure 5.0MPa, speed 1150rpm, vibration 10mm / s, current 42A.

[0102] Process failure problem as center 2: temperature 80°C, pressure 4.5MPa, speed 1250rpm, vibration 15mm / s, current 38A.

[0103] Electrical failure problem as center 3: temperature 100°C, pressure 5.5MPa, speed 1050rpm, vibration 5mm / s, current 48A.

[0104] Then calculate the distance. For each product, calculate the distance to the three centers:

[0105] Distance formula:

[0106] Distance = √[(temperature difference)² + (pressure difference)² + (speed difference)² + (vibration difference)² + (current difference)²]

[0107] Example: the distance of product 1 to centers 1-3.

[0108] Distance 1 (Product 1 to Center 1):

[0109] Distance 1 = √[(85-90)² + (4.8-5.0)² + (1200-1150)² + (12-10)² + (40-42)²]= √[(-5)² + (-0.2)² + (50)² + (2)² + (-2)²]= √[25 + 0.04 + 2500 + 4 + 4]=√2533.04= 50.33

[0110] Conclusion: Product 1 is closest to Center 1 (50.33), assigned to Group 1.

[0111] Finally, group different products according to the closest center:

[0112] For example: Product 1 is closest to Center 1, assigned to Group 1.

[0113] Product 2 is closest to Center 3, assigned to Group 3.

[0114] Product 3 is closest to Center 1, assigned to Group 1.

[0115] (All products are grouped).

[0116] After grouping:

[0117] Mechanical failure problem - Group 1: 60 products.

[0118] Process failure problem - Group 2: 30 products.

[0119] Electrical failure problem - Group 3: 60 products.

[0120] After grouping, analyze the causes of product scrapping through decision tree analysis algorithm:

[0121] Mechanical failure problem group (60 products):

[0122] Product 1: Temperature 85°C, pressure 4.8MPa, speed 1200rpm, vibration 12mm / s, current 40A, temperature fluctuation 2.5°C, pressure fluctuation 0.3Mpa.

[0123] Product 2: Temperature 86°C, pressure 4.9MPa, speed 1180rpm, vibration 11mm / s, current 41A, temperature fluctuation 2.8°C, pressure fluctuation 0.2Mpa.

[0124] Product 3: Temperature 84°C, Pressure 4.7 MPa, Speed 1220 rpm, Vibration 13 mm / s, Current 39 A, Temperature Fluctuation 2.7°C, Pressure Fluctuation 0.2 MPa.

[0125] (60 products in total)

[0126] Process failure problem group (30 products):

[0127] Product 1: Temperature 95°C, Pressure 5.2 MPa, Speed 1100 rpm, Vibration 8 mm / s, Current 45 A, Temperature Fluctuation 3.5°C, Pressure Fluctuation 0.4 MPa.

[0128] Product 2: Temperature 96°C, Pressure 5.3 MPa, Speed 1080 rpm, Vibration 7 mm / s, Current 46 A, Temperature Fluctuation 3.8°C, Pressure Fluctuation 0.4 MPa.

[0129] Product 3: Temperature 94°C, Pressure 5.1 MPa, Speed 1120 rpm, Vibration 9 mm / s, Current 44 A, Temperature Fluctuation 3.2°C, Pressure Fluctuation 0.3 MPa.

[0130] (30 products in total)

[0131] Electrical failure problem group (60 products):

[0132] Product 1: Temperature 100°C, Pressure 5.5 MPa, Speed 1050 rpm, Vibration 5 mm / s, Current 48 A, Temperature Fluctuation 4.0°C, Pressure Fluctuation 0.5 MPa.

[0133] Product 2: Temperature 101°C, Pressure 5.6 MPa, Speed 1030 rpm, Vibration 4 mm / s, Current 49 A, Temperature Fluctuation 4.2°C, Pressure Fluctuation 0.5 MPa.

[0134] Product 3: Temperature 99°C, Pressure 5.4 MPa, Speed 1070 rpm, Vibration 6 mm / s, Current 47 A, Temperature Fluctuation 3.8°C, Pressure Fluctuation 0.4 MPa.

[0135] (60 products in total)

[0136] Calculate initial purity based on different failure problems:

[0137] Specifically: Initial purity = 1 - (number of mechanical failures / total number of products)² - (number of process problems / total number of products)² - (number of electrical problems / total number of products)²

[0138] = 1 - (60 / 150)² - (30 / 150)² - (60 / 150)²

[0139] =1-(0.4)²-(0.2)²-(0.4)²=1-0.16-0.04-0.16=0.64

[0140] Then, the optimal split point is determined for different production environments:

[0141] Try the temperature cutoff point:

[0142] Dividing point 1: Temperature = 90°C

[0143] Temperature > 90°C: 90 products, of which 0 had mechanical failures, 30 had process problems, and 60 had electrical problems.

[0144] Temperature ≤ 90°C: 60 products, of which 60 had mechanical failures, 0 had process problems, and 0 had electrical problems.

[0145] Purity at temperatures >90°C:

[0146] The purity at temperatures > 90°C = 1 - (0 / 90)² - (30 / 90)² - (60 / 90)² = 1 - 0 - (1 / 3)² - (2 / 3)² = 1 - 0 - 0.111 - 0.444 = 0.445

[0147] Purity at temperatures ≤90°C:

[0148] Purity at temperatures ≤ 90°C = 1 - (60 / 60)² - (0 / 60)² - (0 / 60)² = 1 - 1.0 - 0 - 0 = 0

[0149] Information gain:

[0150] Information gain = Initial purity - Weighted average purity = 0.64 - (90 / 150 × 0.445 + 60 / 150 × 0) = 0.64 - (0.6 × 0.445 + 0.4 × 0) = 0.64 - (0.267 + 0) = 0.64 - 0.267 = 0.373

[0151] Division point 2: Temperature = 95°C

[0152] Temperature > 95°C: 60 products, of which 0 had mechanical failures, 30 had process problems, and 30 had electrical problems.

[0153] Temperature ≤ 95°C: 90 products, of which 60 had mechanical failures, 0 had process problems, and 30 had electrical problems.

[0154] Purity at temperatures >95°C:

[0155] Purity at temperatures > 95°C = 1 - (0 / 60)² - (30 / 60)² - (30 / 60)² = 1 - 0 - 0.25 - 0.25 = 0.50

[0156] Purity at temperatures ≤95°C:

[0157] The purity at temperatures ≤ 95°C = 1 - (60 / 90)² - (0 / 90)² - (30 / 90)² = 1 - (2 / 3)² - 0 - (1 / 3)² = 1 - 0.444 - 0 - 0.111 = 0.445

[0158] Information gain:

[0159] Information gain = 0.64 - (60 / 150 × 0.50 + 90 / 150 × 0.445) = 0.64 - (0.4 × 0.50 + 0.6 × 0.445) = 0.64 - (0.20 + 0.267) = 0.64 - 0.467 = 0.173

[0160] Division point 3: Temperature = 93°C

[0161] Temperature > 93°C: 60 products, of which 0 had mechanical failures, 0 had process problems, and 60 had electrical problems.

[0162] Temperature ≤ 93°C: 90 products, of which 60 had mechanical failures, 30 had process problems, and 0 had electrical problems.

[0163] Purity at temperatures >93°C:

[0164] Purity at temperatures > 93°C = 1 - (0 / 60)² - (0 / 60)² - (60 / 60)² = 1 - 0 - 0 - 1.0 = 0

[0165] Purity at temperatures ≤93°C:

[0166] The purity at temperatures ≤ 93°C = 1 - (60 / 90)² - (30 / 90)² - (0 / 90)² = 1 - (2 / 3)² - (1 / 3)² - 0 = 1 - 0.444 - 0.111 - 0 = 0.445

[0167] Information gain:

[0168] Information gain = 0.64 - (60 / 150 × 0 + 90 / 150 × 0.445) = 0.64 - (0.4 × 0 + 0.6 × 0.445) = 0.64 - (0 + 0.267) = 0.64 - 0.267 = 0.373

[0169] The optimal information gain (maximum) for temperature characteristics is 0.373 (temperature = 90°C or 93°C).

[0170] Try other feature segmentation methods:

[0171] Try using a vibration of 6.5 mm / s for segmentation:

[0172] Vibration > 6.5 mm / s: 90 products, of which 60 were mechanical failures, 30 were process problems, and 0 were electrical problems.

[0173] Vibration ≤ 6.5 mm / s: 60 products, of which 0 had mechanical failures, 0 had process problems, and 60 had electrical problems.

[0174] Purity with vibrations > 6.5 mm / s:

[0175] The purity of vibrations > 6.5 mm / s = 1 - (60 / 90)² - (30 / 90)² - (0 / 90)² = 1 - (2 / 3)² - (1 / 3)² - 0 = 1 - 0.444 - 0.111 - 0 = 0.445

[0176] Purity with vibration ≤ 6.5 mm / s:

[0177] The purity of vibration ≤ 6.5 mm / s = 1 - (0 / 60)² - (0 / 60)² - (60 / 60)² = 1 - 0 - 0 - 1.0 = 0

[0178] Information gain:

[0179] Information gain = 0.64 - (90 / 150 × 0.445 + 60 / 150 × 0) = 0.64 - (0.6 × 0.445 + 0.4 × 0) = 0.64 - (0.267 + 0) = 0.373

[0180] The optimal information gain for vibration characteristics is 0.373.

[0181] Compare the information gain of all features:

[0182] Temperature: 0.373

[0183] Vibration: 0.372

[0184] Pressure: 0.285

[0185] Temperature fluctuation: 0.198

[0186] Speed: 0.156

[0187] Current: 0.134

[0188] Pressure fluctuation: 0.089

[0189] Temperature has the highest information gain (0.373), so temperature is chosen as the first split point.

[0190] Constructing the first layer of the decision tree:

[0191] First layer segmentation: Temperature > 90°C

[0192] Left branch (temperature ≤ 90°C): 60 products, of which 60 had mechanical failures, 0 had process problems, and 0 had electrical problems.

[0193] Right branch (temperature > 90°C): 90 products, of which 0 were mechanical failures, 30 were process problems, and 60 were electrical problems.

[0194] Second layer (taking the right branch of the first layer as an example):

[0195] First-level right branch data: 90 products, including 30 process issues and 60 electrical issues.

[0196] Try different feature segmentation:

[0197] Experiment with vibration breakpoint: Breakpoint: Vibration = 6.5 mm / s

[0198] Vibration > 6.5 mm / s: 30 products, 30 of which had process issues, and 0 had electrical issues.

[0199] Vibration ≤ 6.5 mm / s: 60 products, of which 0 had process issues and 60 had electrical issues.

[0200] Calculate purity (using vibration as an example):

[0201] Purity with vibrations > 6.5 mm / s:

[0202] The purity of vibrations > 6.5 mm / s = 1 - (30 / 30)² - (0 / 30)² = 1 - 1.0 - 0 = 0

[0203] Purity with vibration ≤ 6.5 mm / s:

[0204] The purity of vibration ≤ 6.5 mm / s = 1 - (0 / 60)² - (60 / 60)² = 1 - 0 - 1.0 = 0

[0205] Information gain:

[0206] Information gain = 0.445 - (30 / 90 × 0 + 60 / 90 × 0) = 0.445 - (0 + 0) = 0.445

[0207] Other segments

[0208] Finally, it was calculated that the information gain of vibration was the largest (0.445), and vibration was chosen as the split point of the right branch.

[0209] Continue recursively splitting until the stopping condition is met.

[0210] Stop condition:

[0211] All branches have a purity of 0 (completely pure).

[0212] Reach the maximum depth (e.g., 3 layers).

[0213] Current status check:

[0214] Left branch (temperature ≤ 90°C): 60 products, all with mechanical failures, purity = 0.

[0215] Right branch (temperature > 90°C): 90 products, need to be further divided.

[0216] Sub-branches of the right branch:

[0217] Vibration > 6.5 mm / s: 30 products, all with process issues, purity = 0.

[0218] Vibration ≤ 6.5 mm / s: 60 products, all with electrical problems, purity = 0.

[0219] The purity of all branches is 0, which satisfies the stopping condition, so the splitting stops.

[0220] The importance of each feature is equal to the sum of the information gains of that feature across all segmentations.

[0221] Temperature: 0.373 (first division).

[0222] Vibration: 0.445 (right branch split).

[0223] Pressure: 0.285 (right branch split).

[0224] Temperature fluctuation: 0.198 (right branch split).

[0225] Speed: 0.156 (right branch split).

[0226] Current: 0.134 (right branch split).

[0227] Pressure fluctuation: 0.089 (right branch split).

[0228] Normalization process:

[0229] The total sum is 0.373 + 0.445 + 0.285 + 0.198 + 0.156 + 0.134 + 0.089 = 1.680

[0230] The importance of normalization:

[0231] Temperature: 0.373 / 1.680 = 0.222 (22.2%)

[0232] Vibration: 0.445 / 1.680 = 0.265 (26.5%)

[0233] Pressure: 0.285 / 1.680 = 0.170 (17.0%)

[0234] Temperature fluctuation: 0.198 / 1.680 = 0.118 (11.8%)

[0235] Speed: 0.156 / 1.680 = 0.093 (9.3%)

[0236] Current: 0.134 / 1.680 = 0.080 (8.0%)

[0237] Pressure fluctuation: 0.089 / 1.680 = 0.053 (5.3%)

[0238] Practical significance:

[0239] By combining decision tree analysis with K-means clustering results, we found that:

[0240] Vibration is the most important factor (26.5%), and it can effectively distinguish between process problems and electrical problems.

[0241] Temperature was the second most important factor (22.2%), and it was able to effectively distinguish mechanical failures from other problems.

[0242] Stress is the third most important factor (17.0%), and it also has a significant impact on problem classification.

[0243] Decision-making rules:

[0244] If the temperature is ≤90°C, it is determined to be a mechanical failure.

[0245] If the temperature is >90°C and the vibration is >6.5 mm / s, it is determined to be a process problem.

[0246] If the temperature is >90°C and the vibration is ≤6.5mm / s, it is considered an electrical problem.

[0247] This analysis result is completely consistent with the findings of K-means clustering, forming a complete analytical chain: first, three scrapping patterns are discovered through clustering, and then the key features and rules that distinguish these three patterns are identified through decision tree analysis.

[0248] New rules are generated based on the above data:

[0249] Group 1 Mechanical Failure Problems (Low Temperature and High Vibration Mode): 60 products, temperature 85°C, pressure 4.8MPa, speed 1200rpm, vibration 12mm / s, current 40A.

[0250] Group 2 Process Issues (Medium Temperature and Medium Vibration Mode): 30 products, temperature 95°C, pressure 5.2MPa, speed 1100rpm, vibration 8mm / s, current 45A.

[0251] Group 3 Electrical Issues (High Temperature and Low Vibration Mode): 60 products, temperature 100°C, pressure 5.5MPa, speed 1050rpm, vibration 5mm / s, current 48A.

[0252] Results of the decision tree:

[0253] Decision tree structure: Temperature > 90°C → Vibration > 6.5 mm / s → Process problem / Electrical problem.

[0254] Importance of characteristics: temperature 22.2%, vibration 26.5%, pressure 17.0%, temperature fluctuation 11.8%, velocity 9.3%, current 8.0%, pressure fluctuation 5.3%.

[0255] Furthermore, rules are generated based on clustering results.

[0256] Cluster 1 Mechanical Failure Problems (Low Temperature High Vibration Mode):

[0257] Sample size: 60 products

[0258] Average temperature: 85°C

[0259] Average pressure: 4.8 MPa

[0260] Vibration mean: 12 mm / s

[0261] Average speed: 1200 rpm

[0262] Average current: 40A

[0263] Rule parameter calculation:

[0264] Temperature threshold calculation:

[0265] Mean = (83 + 84 + 85 + 86 + 87 + ... + 86) / 60 = 5100 / 60 = 85°C

[0266] Variance = Σ(each value - mean)² / sample size = [(83-85)² + (84-85)² + (85-85)² + (86-85)² + (87-85)² + ...] / 60 = 240 / 60 = 4 Standard deviation = √variance = √4 = 2°C

[0267] The lower limit of the temperature threshold = 85°C - 2°C = 83°C (the average value minus 2°C is used as the warning threshold).

[0268] Upper limit of temperature threshold = 85°C + 2°C = 87°C

[0269] Pressure threshold calculation:

[0270] The lower limit of the pressure threshold = 4.8MPa - 0.2MPa = 4.6MPa (the average value minus 0.2MPa is used as the warning threshold).

[0271] Upper limit of pressure threshold = 4.8 MPa + 0.2 MPa = 5.0 MPa

[0272] Vibration threshold calculation:

[0273] Vibration threshold lower limit = 12mm / s - 1mm / s = 11mm / s (mean value minus 1mm / s is used as the warning threshold)

[0274] Vibration threshold upper limit = 12mm / s + 1mm / s = 13mm / s

[0275] Rule 1: Mechanical Fault Early Warning

[0276] Rule 1: Mechanical fault warning IF temperature ∈ [83°C, 87°C] AND pressure ∈ [4.6MPa, 5.0MPa] AND vibration ∈ [11mm / s, 13mm / s] THEN trigger alarm (Confidence: High, Sample size: 60).

[0277] Cluster 2 (Medium-temperature, medium-vibration mode):

[0278] Sample size: 30 products

[0279] Average temperature: 95°C

[0280] Average pressure: 5.2 MPa

[0281] Vibration mean: 8 mm / s

[0282] Average speed: 1100 rpm

[0283] Average current: 45A

[0284] Rule parameter calculation:

[0285] Lower limit of temperature threshold = 95°C - 2°C = 93°C

[0286] Upper limit of temperature threshold = 95°C + 2°C = 97°C

[0287] Lower limit of pressure threshold = 5.2 MPa - 0.2 MPa = 5.0 MPa

[0288] Upper limit of pressure threshold = 5.2MPa + 0.2MPa = 5.4MPa Lower limit of vibration threshold = 8mm / s - 1mm / s = 7mm / s

[0289] Vibration threshold upper limit = 8mm / s + 1mm / s = 9mm / s

[0290] Generation Rule 2: Early Warning of Process Issues

[0291] Rule 2: Process Problem Warning IF Temperature ∈ [93°C, 97°C] AND Pressure ∈ [5.0MPa, 5.4MPa] AND Vibration ∈ [7mm / s, 9mm / s] THEN Trigger Alarm (Confidence: Medium, Sample Size: 30)

[0292] Cluster 3 (High Temperature Low Vibration Mode):

[0293] Sample size: 60 products

[0294] Average temperature: 100°C

[0295] Average pressure: 5.5 MPa

[0296] Vibration mean: 5 mm / s

[0297] Average speed: 1050 rpm

[0298] Average current: 48A

[0299] Rule parameter calculation:

[0300] Lower limit of temperature threshold = 100°C - 2°C = 98°C

[0301] Upper limit of temperature threshold = 100°C + 2°C = 102°C

[0302] Lower limit of pressure threshold = 5.5MPa - 0.2MPa = 5.3MPa

[0303] Upper limit of pressure threshold = 5.5MPa + 0.2MPa = 5.7MPa

[0304] Vibration threshold lower limit = 5mm / s - 1mm / s = 4mm / s

[0305] Vibration threshold upper limit = 5mm / s + 1mm / s = 6mm / s

[0306] Rule 3: Electrical Problem Warning

[0307] Rule 3: Electrical problem warning IF temperature ∈ [98°C, 102°C] AND pressure ∈ [5.3MPa, 5.7MPa] AND vibration ∈ [4mm / s, 6mm / s] THEN trigger alarm (Confidence: High, Sample size: 60).

[0308] Rule generation is based on feature importance.

[0309] Importance of temperature: 22.2%

[0310] Rule 4: Temperature Anomaly Warning IF Temperature < 87°C OR Temperature > 93°C THEN Trigger Alarm (Confidence: High, Importance: 22.2%)

[0311] Vibration importance: 26.5%

[0312] Rule 5: Vibration Anomaly Warning IF Vibration <6mm / s OR Vibration >10mm / s THEN Trigger Alarm (Confidence: High, Importance: 26.5%)

[0313] Importance of stress: 17.0%

[0314] Rule 6: Pressure Anomaly Warning IF Pressure < 4.9 MPa OR Pressure > 5.3 MPa THEN Trigger Alarm (Confidence: Medium, Importance: 17.0%)

[0315] Path generation rules based on decision tree

[0316] Path 1: Temperature ≤ 90°C → Mechanical failure.

[0317] Rule 7: Mechanical Fault Warning IF Temperature ≤ 90°C THEN Trigger Alarm (Confidence: High, Problem Type: Mechanical Fault).

[0318] Path 2: Temperature > 90°C, vibration > 6.5 mm / s → process problem.

[0319] Rule 8: Process Problem Warning IF Temperature > 90°C AND Vibration > 6.5 mm / s THEN Trigger Alarm (Confidence: High, Problem Type: Process Problem).

[0320] Path 3: Temperature > 90°C, vibration ≤ 6.5 mm / s → electrical problem.

[0321] Rule 9: Electrical Problem Warning IF Temperature > 90°C AND Vibration ≤ 6.5 mm / s THEN Trigger Alarm (Confidence: High, Problem Type: Electrical Problem).

[0322] Sorting based on rule priority

[0323] High-priority rules (based on clustering results):

[0324] Rule 1: Early warning of mechanical failure (sample size: 60, confidence level: high).

[0325] Rule 3: Electrical problem warning (sample size: 60, confidence level: high).

[0326] Rule 2: Early warning of process problems (sample size: 30, confidence level: medium).

[0327] Medium priority rules (based on feature importance):

[0328] Rule 5: Vibration Anomaly Warning (Importance: 26.5%).

[0329] Rule 4: Temperature Anomaly Warning (Importance: 22.2%).

[0330] Rule 6: Warning of abnormal stress (Importance: 17.0%).

[0331] Low-priority rules (based on decision tree paths):

[0332] Rule 8: Early warning of process problems (decision tree path).

[0333] Rule 9: Electrical Problem Warning (Decision Tree Path).

[0334] Rule 7: Mechanical Failure Warning (Decision Tree Path).

[0335] The above rules are deduplicated and merged.

[0336] Deduplication analysis:

[0337] Both Rule 1 and Rule 7 apply to mechanical failures, but the conditions differ.

[0338] Both Rule 2 and Rule 8 address process issues, but the conditions differ.

[0339] Both Rule 3 and Rule 9 address electrical issues, but the conditions differ.

[0340] Merge strategy:

[0341] Retain rules based on clustering results (more precise).

[0342] The decision tree path rule is used as an auxiliary rule.

[0343] Final rule set:

[0344] High-priority rules:

[0345] Rule 1: Mechanical fault warning IF temperature ∈ [83°C, 87°C] AND pressure ∈ [4.6MPa, 5.0MPa] AND vibration ∈ [11mm / s, 13mm / s] THEN trigger alarm (Confidence: High, Sample size: 60).

[0346] Rule 2: Electrical problem warning IF temperature ∈ [98°C, 102°C] AND pressure ∈ [5.3MPa, 5.7MPa] AND vibration ∈ [4mm / s, 6mm / s] THEN trigger alarm (Confidence: High, Sample size: 60).

[0347] Rule 3: Process problem warning IF temperature ∈ [93°C, 97°C] AND pressure ∈ [5.0MPa, 5.4MPa] AND vibration ∈ [7mm / s, 9mm / s] THEN trigger alarm (confidence level: medium, sample size: 30).

[0348] Medium priority rules:

[0349] Rule 4: Vibration Anomaly Warning IF Vibration <6mm / s OR Vibration >10mm / s THEN Trigger Alarm (Confidence: High, Importance: 26.5%).

[0350] Rule 5: Temperature Anomaly Warning IF Temperature < 87°C OR Temperature > 93°C THEN Trigger Alarm (Confidence: High, Importance: 22.2%).

[0351] Rule 6: Pressure Anomaly Warning IF Pressure < 4.9 MPa OR Pressure > 5.3 MPa THEN Trigger Alarm (Confidence: Medium, Importance: 17.0%).

[0352] Low priority rules:

[0353] Rule 7: Comprehensive early warning IF Temperature > 90°C AND Vibration > 6.5mm / s THEN Trigger Alarm (Confidence: Medium, Problem Type: Process Problem).

[0354] Rule 8: Comprehensive warning IF Temperature > 90°C AND Vibration ≤ 6.5mm / s THEN Trigger Alarm (Confidence: Medium, Problem Type: Electrical Problem).

[0355] Rule 9: Comprehensive early warning IF temperature ≤ 90°C THEN trigger alarm (Confidence level: Medium, Problem type: Mechanical failure).

[0356] Optimize rule parameters based on historical data to improve the accuracy and effectiveness of the rules.

[0357] Detailed calculation process:

[0358] Step 1: Calculate rule coverage

[0359] Coverage rate = Number of scrapped products triggered by the rule / Total number of scrapped products

[0360] Step 2: Calculate the rule accuracy

[0361] Accuracy = Number of scrapped products triggered by the rule / Total number of products triggered by the rule

[0362] Step 3: Calculate rule validity

[0363] Effectiveness = Coverage × Accuracy

[0364] Step 4: Optimize rule parameters

[0365] Adjust rule parameters based on validity:

[0366] Rule 1 optimization:

[0367] Original: Temperature ∈ [83°C, 87°C] AND Pressure ∈ [4.6MPa, 5.0MPa] AND Vibration ∈ [11mm / s, 13mm / s].

[0368] After optimization: Temperature ∈ [84°C, 86°C] AND Pressure ∈ [4.7MPa, 4.9MPa] AND Vibration ∈ [11.5mm / s, 12.5mm / s].

[0369] Reason: To improve accuracy and reduce false alarms.

[0370] Rule 2 optimization:

[0371] Original: Temperature ∈ [93°C, 97°C] AND Pressure ∈ [5.0MPa, 5.4MPa] AND Vibration ∈ [7mm / s, 9mm / s].

[0372] After optimization: Temperature ∈ [94°C, 96°C] AND Pressure ∈ [5.1MPa, 5.3MPa] AND Vibration ∈ [7.5mm / s, 8.5mm / s].

[0373] Reason: To improve accuracy and reduce false alarms.

[0374] Rule 3 optimization:

[0375] Original: Temperature ∈ [98°C, 102°C] AND Pressure ∈ [5.3MPa, 5.7MPa] AND Vibration ∈ [4mm / s, 6mm / s].

[0376] After optimization: Temperature ∈ [99°C, 101°C] AND Pressure ∈ [5.4MPa, 5.6MPa] AND Vibration ∈ [4.5mm / s, 5.5mm / s].

[0377] Reason: To improve accuracy and reduce false alarms.

[0378] Rule Validation Algorithm - Validating Rule Effectiveness

[0379] Verify the effectiveness of the generated rules in practical applications to ensure their reliability.

[0380] Detailed calculation process:

[0381] Step 1: Cross-validation

[0382] The effect of using K-fold cross-validation to verify the rules:

[0383] 50% cross-validation:

[0384] The 150 scrapped products were divided into 5 groups of 30 each.

[0385] Use 4 sets of training rules and 1 set of validation rules.

[0386] Repeat 5 times and calculate the average effect.

[0387] Step 2: Calculate the validation index

[0388] Accuracy = Number of correctly predicted scrapped products / Total number of scrapped products Recall = Number of correctly predicted scrapped products / Actual number of scrapped products F1 score = 2 × (accuracy × recall) / (accuracy + recall).

[0389] Step 3: Rule Effectiveness Evaluation

[0390] Rule 1: Precision = 95%, Recall = 90%, F1 score = 92%.

[0391] Rule 2: Precision = 88%, Recall = 85%, F1 score = 86%.

[0392] Rule 3: Precision = 92%, Recall = 88%, F1 score = 90%.

[0393] Step 4: Sorting by Rules

[0394] Sort the rules according to their F1 scores:

[0395] Rule 1: F1 score = 92%.

[0396] Rule 3: F1 score = 90%.

[0397] Rule 2: F1 score = 86%.

[0398] The final generated rule set

[0399] High-priority rules:

[0400] Rule 1: Mechanical Failure Warning IF Temperature ∈ [84°C, 86°C] AND Pressure ∈ [4.7MPa, 4.9MPa] AND Vibration ∈ [11.5mm / s, 12.5mm / s] THEN Trigger Alarm (Confidence: High, F1 Score: 92%).

[0401] Rule 2: Electrical problem warning IF temperature ∈ [99°C, 101°C] AND pressure ∈ [5.4MPa, 5.6MPa] AND vibration ∈ [4.5mm / s, 5.5mm / s] THEN trigger alarm (Confidence: High, F1 score: 90%).

[0402] Rule 3: Process Problem Warning IF Temperature ∈ [94°C, 96°C] AND Pressure ∈ [5.1MPa, 5.3MPa] AND Vibration ∈ [7.5mm / s, 8.5mm / s] THEN Trigger Alarm (Confidence: Medium, F1 Score: 86%).

[0403] Medium priority rules:

[0404] Rule 4: Temperature Anomaly Warning IF Temperature < 87°C OR Temperature > 93°C THEN Trigger Alarm (Confidence: High, Importance: 22.2%).

[0405] Rule 5: Vibration Anomaly Warning IF Vibration <6mm / s OR Vibration >10mm / s THEN Trigger Alarm (Confidence: High, Importance: 26.5%).

[0406] Rule 6: Pressure Anomaly Warning IF Pressure < 4.9 MPa OR Pressure > 5.3 MPa THEN Trigger Alarm (Confidence: Medium, Importance: 17.0%).

[0407] Low priority rules:

[0408] Rule 7: Comprehensive early warning IF Temperature > 90°C AND Vibration > 6.5mm / s THEN Trigger Alarm (Confidence: Medium, Problem Type: Process Problem).

[0409] Rule 8: Comprehensive warning IF Temperature > 90°C AND Vibration ≤ 6.5mm / s THEN Trigger Alarm (Confidence: Medium, Problem Type: Electrical Problem).

[0410] Rule 9: Comprehensive early warning IF temperature ≤ 90°C THEN trigger alarm (Confidence level: Medium, Problem type: Mechanical failure).

[0411] Practical application effect

[0412] Before optimization:

[0413] Accuracy rate of early warning for scrapped products: 75%.

[0414] False alarm rate: 15%.

[0415] Early warning time: 5 minutes.

[0416] After optimization:

[0417] Accuracy rate of early warning for obsolete products: 92%.

[0418] False alarm rate: 8%.

[0419] Early warning time: 15 minutes.

[0420] The early warning method for industrial product production provided in this disclosure automatically infers the reasonable threshold range of key production factors by directly analyzing the initial parameter data of historical scrapped products. This changes the model of relying on manual experience to set static thresholds and improves the accuracy of product production. Furthermore, by converting historical scrapped data into executable early warning rules and applying them to real-time monitoring and future production, the stability of product production is further improved.

[0421] Example 2

[0422] like Figure 5 As shown, this embodiment provides an early warning system for industrial product production. The early warning system includes:

[0423] The data acquisition module 100 is used to acquire initial parameter data corresponding to several scrapped products; the initial parameter data includes several production factor information.

[0424] The scrap impact factor acquisition module 200 is used to acquire target scrap impact factor information of scrapped products under different production environments based on several initial parameter data; the target scrap impact factor information is represented as environmental parameter information of different scrapped products under corresponding production environments;

[0425] The early warning classification module 300 is used to generate early warning rule information for industrial products in different production environments based on information on several target scrapping impact factors.

[0426] like Figure 6 As shown, the scrap impact factor acquisition module 200 in this embodiment includes an initial purity acquisition unit 201, an actual impact factor acquisition unit 202, and a target impact factor acquisition unit 203.

[0427] The initial purity acquisition unit 201 is used to acquire initial purity information based on several initial parameter data;

[0428] The actual impact factor acquisition unit 202 is used to acquire actual scrapping impact factor information under set production environment conditions based on several initial parameter data;

[0429] The target impact factor acquisition unit 203 is used to determine the target scrap impact factor information based on the initial purity information and the actual scrap impact factor information.

[0430] The initial purity acquisition unit 201 in this embodiment is used for:

[0431] Based on preset classification conditions, several scrapped products are divided into several scrap groups; wherein the preset classification conditions include at least one of mechanical failure conditions, process failure conditions, and electrical failure conditions;

[0432] Obtain the first quantity information of the corresponding scrapped products in different scrapping groups; the first quantity information represents the quantity of scrapped products in different scrapping groups;

[0433] Based on several primary quantitative information, the initial purity information is determined; the initial purity information is characterized as the degree of failure information of industrial products under different production environments.

[0434] In this embodiment, the actual impact factor acquisition unit 202 is used for:

[0435] Obtain second quantity information of scrapped products belonging to different scrapping groups under set production environment conditions; the second quantity information represents the quantity of scrapped products in different scrapping groups under set production environment conditions;

[0436] Based on several second-order quantitative information, information on actual scrapping impact factors under given production environment conditions is obtained.

[0437] In this embodiment, the target influence factor acquisition unit 203 is used for:

[0438] Based on the difference between the initial purity information and the actual scrap impact factor information, the target scrap impact factor information under the corresponding set production environment conditions is determined.

[0439] In this embodiment, the early warning classification module 300 is used for:

[0440] The impact factors of target scrapping under different production environments are sorted from high to low, and early warning rules for the corresponding production environments are generated based on the sorting results.

[0441] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The system embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this disclosure according to actual needs.

[0442] The early warning system for industrial product production provided in this disclosure automatically infers reasonable threshold ranges for key production factors by directly analyzing the initial parameter data of historical scrapped products. This changes the model of relying on manual experience to set static thresholds and improves the accuracy of product production. Furthermore, by converting historical scrapped data into executable early warning rules and applying them to real-time monitoring and future production, the stability of product production is further improved.

[0443] Example 3

[0444] like Figure 7 As shown, Figure 7 This is a schematic diagram of the structure of an electronic device provided in Embodiment 4 of this disclosure. The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the methods described in the above embodiments. Figure 7 The electronic device 30 shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments disclosed herein.

[0445] like Figure 7 As shown, the electronic device 30 can be represented in the form of a general computing device, such as a server device. The components of the electronic device 30 may include, but are not limited to: at least one processor 31, at least one memory 32, and a bus 33 connecting different system components (including memory 32 and processor 31).

[0446] Bus 33 includes a data bus, an address bus, and a control bus.

[0447] The memory 32 may include volatile memory, such as random access memory (RAM) 321 and / or cache memory 322, and may further include read-only memory (ROM) 323.

[0448] The memory 32 may also include a program tool 325 having a set (at least one) of program modules 324, including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0449] The processor 31 performs various functional applications and data processing, such as the methods described in the above embodiments of this disclosure, by running computer programs stored in the memory 32.

[0450] Electronic device 30 can also communicate with one or more external devices 34 (e.g., keyboard, pointing device, etc.). This communication can be performed via input / output (I / O) interface 35. Furthermore, the model-generated electronic device 30 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 36. ​ As shown, network adapter 36 communicates with other modules of the model-generated electronic device 30 via bus 33. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with the model-generated electronic device 30, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (disk array) systems, tape drives, and data backup storage systems.

[0451] It should be noted that although several units / modules or sub-units / modules of the electronic device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided and embodied by multiple units / modules.

[0452] Example 4

[0453] This disclosure also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the early warning method for industrial product production provided in any of the above embodiments.

[0454] The readable storage medium may be more specifically adopted, including but not limited to: portable disk, hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0455] Example 5

[0456] This disclosure also provides a computer program product, including a computer program that, when executed by a processor, implements the early warning method for the production of industrial products as described above.

[0457] The program code for executing the computer program product of this disclosure can be written in any combination of one or more programming languages, and the program code can be executed entirely on a user device, partially on a user device, as a stand-alone software package, partially on a user device and partially on a remote device, or entirely on a remote device.

[0458] While specific embodiments of this disclosure have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of this disclosure is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of this disclosure, but all such changes and modifications fall within the scope of protection of this disclosure.

Claims

1. A method for early warning during the production of industrial products, characterized in that, The early warning method includes: Acquire initial parameter data corresponding to several scrapped products; the initial parameter data includes several production factor information. Based on several initial parameter data, target scrapping impact factor information of the scrapped products under different production environments is obtained; the target scrapping impact factor information is characterized as environmental parameter information of different scrapped products under corresponding production environments; Based on the information of several target scrapping influencing factors, early warning rule information is generated for the industrial products when they are produced in different production environments.

2. The early warning method for industrial product production according to claim 1, characterized in that, The step of obtaining the target scrapping impact factor information of the scrapped product under different production environments based on several initial parameter data includes: Based on the aforementioned initial parameter data, obtain the initial purity information; Based on several initial parameter data, information on actual scrapping impact factors under set production environment conditions is obtained; Based on the initial purity information and the actual scrap impact factor information, the target scrap impact factor information is determined.

3. The early warning method for industrial product production according to claim 2, characterized in that, The step of obtaining initial purity information based on several of the initial parameter data includes: The scrapped products are divided into several scrap groups based on preset classification conditions; wherein the preset classification conditions include at least one of mechanical failure conditions, process failure conditions, and electrical failure conditions. Obtain first quantity information of the corresponding scrapped products in different scrapping groups; the first quantity information represents the quantity of scrapped products in different scrapping groups; Based on several of the first quantity information, the initial purity information is determined; the initial purity information is characterized as the degree of failure information of the industrial product under different production environments.

4. The early warning method for industrial product production according to claim 3, characterized in that, The step of obtaining actual scrapping impact factor information under set production environment conditions based on several initial parameter data includes: Obtain second quantity information of the scrapped products belonging to different scrapping groups under the set production environment conditions; the second quantity information represents the quantity of the scrapped products in different scrapping groups under the set production environment conditions; Based on several of the second quantitative information, the actual scrap impact factor information under the set production environment conditions is obtained.

5. The early warning method for industrial product production according to claim 4, characterized in that, The step of determining the target scrapping impact factor information based on the initial purity information and the actual scrapping impact factor information includes: Based on the difference between the initial purity information and the actual scrap impact factor information, the target scrap impact factor information corresponding to the set production environment conditions is determined.

6. The early warning method for industrial product production according to claim 1, characterized in that, The step of generating early warning rule information for the industrial product during production in different production environments based on several target scrapping impact factors includes: The target scrapping impact factors under different production environments are sorted from high to low, and early warning rules corresponding to the production environment are generated based on the sorting results.

7. An early warning system for industrial product manufacturing, characterized in that, The early warning system includes: The data acquisition module is used to acquire initial parameter data corresponding to several scrapped products; the initial parameter data includes several production factor information. The scrap impact factor acquisition module is used to acquire target scrap impact factor information of the scrapped product under different production environments based on several initial parameter data; the target scrap impact factor information is characterized as environmental parameter information of different scrapped products under corresponding production environments; The early warning classification module is used to generate early warning rule information for the industrial products produced in different production environments based on several target scrapping impact factors.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and for running on the processor, characterized in that, When the processor executes the computer program, it implements the early warning method for the production of industrial products as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the early warning method for the production of industrial products as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the early warning method for the production of industrial products as described in any one of claims 1 to 6.

Images