A method for technical trade compliance testing and standardization certification of electromechanical products

Abstract

The present invention discloses a method for technical trade compliance detection and standardization certification of electromechanical products. The electromechanical products that have passed the trade compliance detection and standardization certification are taken as target products, and the historical production data of the target products are retrieved. Ideal production data is generated based on the historical production data. The production data at least includes the processing parameter items adjusted by the PID controller in each production process of the product; the production data is monitored in real time, and the deviation from the ideal model is predicted by the fluctuation model. Once the deviation exceeds the standard, an early warning is issued. For the early warning data, the abnormality is first compared, and the PID parameters are fine-tuned if it does not exceed the limit; if it exceeds the limit, the PID parameters are optimized using a genetic algorithm, and the first n groups of parameters are used as the initial population to ensure production compliance and efficiency. This method effectively integrates detection, prediction and intelligent adjustment to ensure the quality and compliance of electromechanical products.

Description

Technical Field

[0001] The present invention belongs to the technical field of standardization management, and specifically relates to a method for technical trade compliance detection and standardization certification of electromechanical products. Background Art

[0002] In the production process of electromechanical products, it is crucial to ensure the technical trade compliance and standardization certification of the products. This is not only related to the market access of the products, but also directly affects the quality and performance of the products. Traditional testing methods usually rely on manual sampling and regular audits, which are inefficient and prone to missing problems. They cannot monitor parameter changes in the production process in real time, resulting in the inability to discover and correct problems in a timely manner. In the production process, automatically adjusting processing parameters according to real-time data to ensure product consistency is the current development direction of technical trade compliance and standardization certification; however, when production data is abnormal, how to quickly identify and take effective measures to adjust it is still a problem faced when integrating technical trade compliance and standardization certification into the entire production process.

[0003] With the development of industrial automation and intelligence, a more efficient and real-time monitoring method is needed to ensure that the production process of electromechanical products always meets the requirements of trade compliance and standardized certification. Summary of the invention

[0004] The purpose of the present invention is to provide a real-time monitoring method to ensure that the electromechanical products produced can pass the technical trade compliance testing and standardization certification of electromechanical products. The specific technical solution is as follows:

[0005] A method for technical trade compliance testing and standardization certification of electromechanical products, the method comprising:

[0006] The electromechanical products that have passed trade compliance testing and standardization certification are taken as target products, the historical production data of the target products are retrieved, and ideal production data are generated based on the historical production data. The production data at least includes the processing parameter items adjusted by the PID controller in each production process of the product.

[0007] Collect the time series production data of electromechanical products in real time, and build a fluctuation model between the time series production data and the ideal production data; predict the standardized distance between the time series production data and the ideal production data at the next moment based on the fluctuation model, and when the standardized distance exceeds the preset distance, trigger the generation of an early warning request signal.

[0008] Abnormal production data is extracted from the early warning request signal, and the abnormal production data is compared with the ideal production data to obtain the degree of abnormality of the abnormal production data and the ideal production data.

[0009] When the abnormality does not exceed the preset threshold, the processing parameter items are adjusted according to the time series production data at the current moment and the time series production data predicted at the next moment until the standardized distance does not exceed the preset distance.

[0010] When the abnormality exceeds a preset threshold, the parameters of the n groups of PID controllers before the current moment are used as the initial population, and the processing parameters of the PID controller at the next moment are optimized based on the genetic algorithm.

[0011] Furthermore, the fluctuation model is used to describe the difference between the time series production data and the ideal production data, and a multivariate nonlinear regression model is used to fit the time-based curve function of the time series production data. The production data at least includes processing parameter items adjusted by the PID controller in each production process of the product.

[0012] The time series production data is represented by S(t), S(t) = {s1(t), s2(t), ...s i (t), ...s n (t)}, where s i (t) represents the value of the i-th processing parameter item at time t in the time series production data, and n represents the number of processing parameter items.

[0013] The ideal production data is expressed as L(t), L(t) = {l1(t), l2(t), ...l i (t), ...l n (t)},l i (t) represents the value of the i-th processing parameter item at time t in the ideal production data.

[0014] The fluctuation model is expressed as: W(t) = S(t) - L(t), the standardized distance between the time series production data and the ideal production data at time t+1 Among them, S i (t+1) represents the value of the i-th processing parameter item at time t+1 in the time series production data, L i (t+1) represents the value of the i-th processing parameter item at time t+1 in the ideal production data.

[0015] Furthermore, when the standardized distance D(t+1) at time t+1 exceeds the preset distance L0, an early warning request signal is triggered, and the early warning request signal at least includes the early warning request time and the time series production data predicted at time t+1.

[0016] Calculate the difference Δ between each processing parameter item at time t+1 and the ideal data i (t+1), the rate of change Δ i Processing parameter items with θ(t+1) greater than or equal to α are considered abnormal production data: Δ i(t+1)=|S i (t+1)-L i (t+1)|;

[0017]

[0018] Among them, ∈ is a constant term, ∈>0, α is 0.247; if Δ i There are m processing parameter items with θ(t+1)≥α. The change rate between the m abnormal production data at time t+1 and the ideal data is weighted averaged to obtain the abnormality degree between the abnormal production data at time t+1 and the ideal production data:

[0019] Furthermore, when the abnormality does not exceed the preset threshold, i The processing parameter item θ(t+1)≥α is adjusted. The specific adjustment method is:

[0020]

[0021] Wherein, b is the multiplication factor, 0.5≤b≤1.5. If the standardized distance still exceeds the preset distance after one adjustment, the value of the multiplication factor is adjusted until the standardized distance does not exceed the preset distance.

[0022] Furthermore, when the abnormality exceeds the preset threshold, the n groups of PID controller parameters before the current moment are used as the initial population, and each group of parameters includes a proportional gain K P , integration time T i and the derivative time T d , select the inverse of the standardized distance D(t+1) as the fitness function Fitness,

[0023] Perform selection, crossover and mutation operations in sequence; perform selection through the roulette wheel selection method to determine the next generation of individuals. Suppose there are N individuals in the population, and the fitness value of each individual is Fitness i , i = 1, 2, ..., N; total fitness value F total is the sum of all individual fitness values,

[0024] The probability P of each individual being selected i The ratio of its fitness value to the total fitness value: Calculate the cumulative probability Cumulative probability C i , cumulative probability C i The sum of the selection probabilities for the current individual and all previous individuals, Where 1≤j≤N, C N =1;

[0025] Generate a random number r in the range [0, 1] and select individuals according to the cumulative probability:

[0026]

[0027] The selected individuals are cross-operated to generate new individuals, and the cross-generated individuals are mutated. The selection operation, cross-operation and mutation operation are repeated until the maximum number of iterations is reached. After the iteration, the individual with the highest fitness function value is selected as the optimal processing parameter of the PID controller at the next moment; the maximum number of iterations is in the range of 20 to 30 times.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] The present invention can timely discover and adjust processing parameters by real-time collection and analysis of production data, ensuring that products always meet technical trade compliance testing standards. When abnormal data is detected, the system can automatically trigger an early warning and optimize parameters to reduce manual intervention. Through automated and intelligent monitoring methods, production efficiency and product quality are improved, and production costs are reduced. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention is described clearly and completely below. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] A method for technical trade compliance testing and standardization certification of electromechanical products, comprising the following steps:

[0032] Step 1: Take the electromechanical products that have passed trade compliance testing and standardization certification as target products, retrieve the historical production data of the target products, and generate ideal production data based on the historical production data. The production data at least includes the processing parameter items adjusted by the PID controller in each production process of the product.

[0033] In international trade, technical trade compliance testing and standardization certification of electromechanical products are important links to ensure that products comply with the laws and standards of the importing country. The main documents and standards include: International Electrotechnical Commission (IEC) standards, International Organization for Standardization (ISO) standards, EU CE certification, etc. This document and standard covers safety, electromagnetic compatibility, performance and other aspects; for example, a smart washing machine produced by a home appliance manufacturing company needs to be exported to the EU market, and the washing machine must meet the EU CE certification standards and related technical trade requirements.

[0034] To ensure that the products always meet these requirements throughout the entire production process, the company first identified washing machine models that have passed CE certification and trade compliance testing as target products, and retrieved the historical production data of these target products, including the processing parameters adjusted by the PID controller in each production process (such as motor speed, water temperature control, washing time, etc.); based on historical data, data analysis tools were used to generate an ideal production data model, which reflects the parameter settings under optimal production conditions.

[0035] PID controller is the most commonly used controller in industrial control systems. It adjusts the control variables to make the system output reach the expected value. PID controller can accurately control the temperature, pressure, flow, speed and other parameters in the production process. For example, in chemical production, PID controller can accurately control the temperature of the reactor to ensure the stability of chemical reactions.

[0036] Step 2: collect the time series production data of electromechanical products in real time, and construct a fluctuation model between the time series production data and the ideal production data; predict the standardized distance between the time series production data and the ideal production data at the next moment based on the fluctuation model, and when the standardized distance exceeds the preset distance, trigger the generation of an early warning request signal.

[0037] The fluctuation model is used to describe the difference between the time series production data and the ideal production data. A multivariate nonlinear regression model is used to fit the time-based curve function of the time series production data. The production data at least includes the processing parameter items adjusted by the PID controller in each production process of the product.

[0038] The time series production data is represented by S(t), S(t) = {s1(t), s2(t), ...s i (t), ...s n (t)}, where s i (t) represents the value of the i-th processing parameter item at time t in the time series production data, and n represents the number of processing parameter items.

[0039] The ideal production data is expressed as L(t), L(t) = {l1(t), l2(t), ...l i (t), ...l n (t)},l i (t) represents the value of the i-th processing parameter item at time t in the ideal production data.

[0040] The fluctuation model is expressed as: W(t) = S(t) - L(t), the standardized distance between the time series production data and the ideal production data at time t+1 Among them, S i(t+1) represents the value of the i-th processing parameter item at time t+1 in the time series production data, L i (t+1) represents the value of the i-th processing parameter item at time t+1 in the ideal production data.

[0041] When the standardized distance D(t+1) at time t+1 exceeds the preset distance L0, an early warning request signal is triggered to be generated, and the early warning request signal at least includes the early warning request time and the time series production data predicted at time t+1.

[0042] Assume that a certain electromechanical product has three key processing parameters in the production process: temperature (T), pressure (P) and speed (V), historical production data: T = 200 ° C, P = 100 kPa, V = 500 rpm; ideal production data: T = 200 ° C, P = 100 kPa, V = 500 rpm; real-time collection of time series production data: time series production data: T = 205 ° C, P = 102 kPa, V = 505 rpm.

[0043] Fluctuation model: W(t)=S(t)-L(t), standardized distance: D(t+1)≈7.35, preset distance L0=5, since D(t+1)>L0, triggers the generation of an early warning request signal.

[0044] Step three: extracting abnormal production data from the early warning request signal, comparing the abnormal production data with the ideal production data, and obtaining the degree of abnormality between the abnormal production data and the ideal production data.

[0045] Calculate the difference Δ between each processing parameter item at time t+1 and the ideal data i (t+1), the rate of change Δ i Processing parameter items with θ(t+1) greater than or equal to α are considered abnormal production data: Δ i (t+1)=|S i (t+1)-L i (t+1)|;

[0046]

[0047] Among them, ∈ is a constant term, ∈>0, α is 0.247; if Δ i There are m processing parameter items with θ(t+1)≥α. The change rate between the m abnormal production data at time t+1 and the ideal data is weighted averaged to obtain the abnormality degree between the abnormal production data at time t+1 and the ideal production data:

[0048] Assume that a certain electromechanical product has three key processing parameters in the production process: temperature (T), pressure (P) and speed (V), historical production data: T = 200 ° C, P = 100 kPa, V = 500 rpm; ideal production data; then, the difference: ΔT = |205-200| = 5, ΔP = |102-100| = 2, ΔV = |505-500| = 5; rate of change: ΔTθ = 5 / (200+ε), ΔPθ = 2 / (100+ε), ΔVθ = 5 / (500+ε); abnormality: Δθ(t+1) = (ΔTθ+ΔPθ+ΔVθ) / 3.

[0049] Step 4: When the abnormality does not exceed the preset threshold, the processing parameter items are adjusted according to the time series production data at the current moment and the time series production data predicted at the next moment until the standardized distance does not exceed the preset distance.

[0050] When the abnormality does not exceed the preset threshold, i The processing parameter item θ(t+1)≥α is adjusted. The specific adjustment method is:

[0051]

[0052] Wherein, b is the multiplication factor, 0.5≤b≤1.5. If the standardized distance still exceeds the preset distance after one adjustment, the value of the multiplication factor is adjusted until the standardized distance does not exceed the preset distance.

[0053] Step 5: When the abnormality exceeds a preset threshold, the n groups of PID controller parameters before the current moment are used as the initial population, and the processing parameters of the PID controller at the next moment are optimized based on the genetic algorithm.

[0054] Furthermore, when the abnormality exceeds the preset threshold, the n groups of PID controller parameters before the current moment are used as the initial population, and each group of parameters includes a proportional gain K P , integration time T i and the derivative time T d , select the inverse of the standardized distance D(t+1) as the fitness function Fitness,

[0055] Perform selection, crossover and mutation operations in sequence; perform selection through the roulette wheel selection method to determine the next generation of individuals. Suppose there are N individuals in the population, and the fitness value of each individual is Fitness i , i = 1, 2, ..., N; total fitness value F total is the sum of all individual fitness values,

[0056] The probability P of each individual being selectedi The ratio of its fitness value to the total fitness value: Calculate the cumulative probability Cumulative probability C i , cumulative probability C i The sum of the selection probabilities for the current individual and all previous individuals, Where 1≤j≤N, C N =1;

[0057] Generate a random number r in the range [0, 1] and select individuals according to the cumulative probability:

[0058]

[0059] Assume there are 4 individuals, whose fitness values ​​are Fitness1 = 10, Fitness2 = 20, Fitness3 = 30 and Fitness4 = 40. The total fitness value F total =10+20+30+40=100.

[0060] Calculate the selection probability of each individual,

[0061] The cumulative probability is C1 = P1 = 0.1; C2 = P1 + P2 = 0.1 + 0.2 = 0.3; C3 = P1 + P2 + P3 = 0.1 + 0.2 + 0.3 = 0.6; C4 = P1 + P2 + P3 + P4 = 0.1 + 0.2 + 0.3 + 0.4 = 1.0. Generate a random number r, for example r = 0.45, and select the third individual according to the cumulative probability, because C2 ≤ r ≤ C3.

[0062] The selected individuals are cross-operated to generate new individuals, and the cross-generated individuals are mutated. The selection operation, cross-operation and mutation operation are repeated until the maximum number of iterations is reached. After the iteration, the individual with the highest fitness function value is selected as the optimal processing parameter of the PID controller at the next moment; the maximum number of iterations is in the range of 20 to 30 times.

[0063] Through the above steps, the system can monitor and adjust production parameters in real time to ensure that electromechanical products always meet the requirements of technical trade compliance and standardization certification.

[0064] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for technical trade compliance testing and standardization certification of electromechanical products, characterized in that: The method comprises: Taking an electromechanical product that has passed trade compliance testing and standardization certification as a target product, retrieving historical production data of the target product, and generating ideal production data based on the historical production data, wherein the production data at least includes processing parameter items adjusted by a PID controller in each production process of the product; Collect the time series production data of electromechanical products in real time, and construct a fluctuation model between the time series production data and the ideal production data; predict the standardized distance between the time series production data and the ideal production data at the next moment according to the fluctuation model, and trigger the generation of an early warning request signal when the standardized distance exceeds a preset distance; Extracting abnormal production data from the early warning request signal, comparing the abnormal production data with the ideal production data, and obtaining an abnormality degree between the abnormal production data and the ideal production data; When the abnormality does not exceed the preset threshold, the processing parameter item is adjusted according to the time series production data at the current moment and the time series production data predicted at the next moment until the standardized distance does not exceed the preset distance; When the abnormality exceeds a preset threshold, the parameters of the n groups of PID controllers before the current moment are used as the initial population, and the processing parameters of the PID controller at the next moment are optimized based on the genetic algorithm; The fluctuation model is used to describe the difference between the time series production data and the ideal production data, and a multivariate nonlinear regression model is used to fit the time-based curve function of the time series production data. The production data at least includes the processing parameter items adjusted by the PID controller in each production process of the product; The time series production data is represented by S(t), S(t) = {s1(t), s2(t), …s i (t),…s n (t)}, where s i (t) represents the value of the i-th processing parameter item at time t in the time series production data, and n represents the number of processing parameter items; The ideal production data is expressed as L(t), L(t) = {l1(t),l2(t),…l i (t),…l n (t)},l i (t) represents the value of the i-th processing parameter item at time t in the ideal production data; The fluctuation model is expressed as: W(t) = S(t) - L(t), the standardized distance between the time series production data and the ideal production data at time t+1 Among them, S i (t+1) represents the value of the i-th processing parameter item at time t+1 in the time series production data, L i (t+1) represents the value of the i-th processing parameter item at time t+1 in the ideal production data; When the standardized distance D(t+1) at time t+1 exceeds the preset distance L0, a warning request signal is triggered to be generated, and the warning request signal at least includes the warning request time and the time series production data predicted at time t+1; Calculate the difference Δ between each processing parameter item at time t+1 and the ideal data i (t+1), the rate of change Δ i Processing parameter items with θ(t+1) greater than or equal to α are considered abnormal production data: Δ i (t+1)=|S i (t+1)-L i (t+1)|; Among them, ∈ is a constant term, ∈>0, α is 0.247; if Δ i There are m processing parameter items with θ(t+1)≥α. The change rate between the m abnormal production data at time t+1 and the ideal data is weighted averaged to obtain the abnormal degree of the abnormal production data at time t+1 and the ideal production data:

2. According to claim 1, a method for technical trade compliance testing and standardization certification of electromechanical products is characterized in that: When the abnormality does not exceed the preset threshold, i The processing parameter item θ(t+1)≥α is adjusted. The specific adjustment method is: Wherein, b is the multiplication factor, 0.5≤b≤1.

5. If the standardized distance still exceeds the preset distance after one adjustment, the value of the multiplication factor is adjusted until the standardized distance does not exceed the preset distance.

3. A method for technical trade compliance testing and standardization certification of electromechanical products according to claim 2, characterized in that: When the abnormality exceeds the preset threshold, the n groups of PID controller parameters before the current moment are used as the initial population, each group of parameters includes the proportional gain K P , integration time T i and the differential time T d , select the inverse of the standardized distance D(t+1) as the fitness function Fitness, Perform selection, crossover and mutation operations in sequence; perform selection through the roulette wheel selection method to determine the next generation of individuals. Suppose there are N individuals in the population, and the fitness value of each individual is Fitness i ,i=1,2,…,N; total fitness value F total is the sum of all individual fitness values, The probability P of each individual being selected i The ratio of its fitness value to the total fitness value: Calculate the cumulative probability Cumulative probability C i , cumulative probability C i The sum of the selection probabilities for the current individual and all previous individuals, Where 1≤j≤N, C N =1; Generate a random number r in the range [0,1] and select individuals according to the cumulative probability: The selected individuals are cross-operated to generate new individuals, and the cross-generated individuals are mutated. The selection operation, cross-operation and mutation operation are repeated until the maximum number of iterations is reached. After the iteration, the individual with the highest fitness function value is selected as the optimal processing parameter of the PID controller at the next moment; the maximum number of iterations is in the range of 20 to 30 times.

Images