Industrial equipment operation state control method and device and computing equipment

Abstract

The invention provides a control method and device for an operation state of industrial equipment and computing equipment. The control method comprises the following steps: acquiring real-time state data of the industrial equipment in an actual working state in an industrial production process; performing feature extraction processing on the real-time state data to obtain feature state data; inputting the feature state data into a trained industrial equipment defect detection model for defect detection to obtain a detection result; the industrial equipment defect detection model is obtained by training a preset adjacent neural network model according to historical operation state data of industrial equipment; determining defect probability data according to a detection result; according to the defect probability data, industrial equipment adjustment parameters are obtained; and according to the industrial equipment adjustment parameters, controlling the industrial equipment to operate according to the industrial equipment adjustment parameters. The device operation parameters can be corrected in real time, the comprehensive production efficiency is improved, and the product defect probability is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment status control methods, and in particular to a control method, device and computing device for the operating status of industrial equipment. Background Art

[0002] After years of development, device communication technology has become mature and widely used in the manufacturing industry. Through device communication technology, data communication between IoT terminal hardware and production equipment can be realized, and the communication cycle can reach milliseconds. By obtaining device operating parameters in real time, abnormal device parameters can be monitored in real time.

[0003] In production activities, in addition to the specific equipment processing process, there is also the coordination of auxiliary equipment and quality inspection equipment. The auxiliary equipment provides production-related raw materials, consumables, auxiliary gases, etc. The quality inspection equipment conducts quality inspection on the finished products and coordinates with the equipment. Those with unqualified quality are processed again to complete the entire process of production-quality inspection-repair.

[0004] At present, the entire production process and efficiency optimization, from production equipment, auxiliary equipment to quality inspection equipment, are mostly analyzed afterwards, which cannot provide substantial and timely help for improving subsequent processing efficiency. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a control method, device and computing device for the operating status of industrial equipment, which can correct the equipment operating parameters in real time, improve the overall production efficiency and reduce the probability of product defects.

[0006] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0007] A method for controlling the operating state of industrial equipment, comprising:

[0008] Obtain real-time status data of industrial equipment under actual working conditions in industrial production processes;

[0009] Performing feature extraction processing on the real-time status data to obtain feature status data;

[0010] The characteristic state data is input into a trained industrial equipment defect detection model to perform defect detection and obtain a detection result; the industrial equipment defect detection model is obtained by training a preset neighboring neural network model according to the historical operating state data of the industrial equipment;

[0011] Determining defect probability data based on the detection results;

[0012] Obtaining industrial equipment adjustment parameters according to the defect probability data;

[0013] According to the industrial equipment adjustment parameters, the industrial equipment is controlled to operate according to the industrial equipment adjustment parameters.

[0014] Optionally, performing feature extraction processing on the real-time status data to obtain feature status data includes:

[0015] Cleaning the real-time status data to obtain first data;

[0016] Preprocessing the first data to obtain second data;

[0017] Perform feature extraction processing on the second data to obtain feature state data.

[0018] Optionally, the training process of the industrial equipment defect detection model includes:

[0019] Acquire historical status data of industrial equipment within a preset time period, wherein the historical status data includes defect status data;

[0020] Extracting features from the defect state data to obtain defect state feature data;

[0021] Inputting the defect state characteristic data into the first training module of the preset proximity neural network model for processing to obtain a first processing result;

[0022] Inputting the first processing result into the second training module of the preset proximity neural network model for processing to obtain a second processing result;

[0023] The defect type is determined according to the second processing result to obtain an industrial equipment defect detection model.

[0024] Optionally, the characteristic state data is input into a trained industrial equipment defect detection model to perform defect detection to obtain a detection result, including:

[0025] Inputting the characteristic state data into a first processing module in an industrial equipment defect detection model for processing to obtain a first output;

[0026] The first output is input into a second processing module in an industrial equipment defect detection model for defect detection processing to obtain a detection result.

[0027] Optionally, determining defect probability data according to the detection result includes:

[0028] Normalizing the detection result to obtain a normalized result;

[0029] Defect probability data is determined based on the normalized result.

[0030] Optionally, determining defect probability data according to the normalization result includes:

[0031] According to the normalization result, the likelihood probability, prior probability, and marginal probability of the feature state data are obtained;

[0032] Defect probability data is obtained according to the likelihood probability, prior probability and marginal probability.

[0033] Optionally, obtaining industrial equipment adjustment parameters according to the defect probability data includes:

[0034] When the defect probability data is greater than a first preset value, the characteristic state data is compared with a standard value of an equipment operation parameter to obtain an industrial equipment adjustment parameter;

[0035] Otherwise, the industrial equipment adjustment parameters remain unchanged.

[0036] An embodiment of the present invention further provides a device for controlling the operating status of industrial equipment, comprising:

[0037] An acquisition module is used to obtain real-time status data of industrial equipment under actual working conditions in industrial production processes;

[0038] A processing module, used for performing feature extraction processing on the real-time status data to obtain feature status data;

[0039] The characteristic state data is input into a trained industrial equipment defect detection model to perform defect detection and obtain a detection result; the industrial equipment defect detection model is obtained by training a preset neighboring neural network model according to the historical operating state data of the industrial equipment;

[0040] Determining defect probability data based on the detection results;

[0041] Obtaining industrial equipment adjustment parameters according to the defect probability data;

[0042] According to the industrial equipment adjustment parameters, the industrial equipment is controlled to operate according to the industrial equipment adjustment parameters.

[0043] An embodiment of the present invention further provides a computing device, comprising:

[0044] one or more processors;

[0045] The storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the control method of the operating status of industrial equipment described in the present invention.

[0046] An embodiment of the present invention further provides a computer-readable storage medium, wherein a program is stored in the computer-readable storage medium, and when the program is executed by a processor, the method for controlling the operating status of industrial equipment described in the present invention is implemented.

[0047] The above technical solution of the present invention has at least the following technical effects:

[0048] The control method of the operating state of industrial equipment of the present invention obtains the real-time state data of the industrial equipment in the actual working state in the industrial production process; performs feature extraction processing on the real-time state data to obtain feature state data; inputs the feature state data into a trained industrial equipment defect detection model to perform defect detection and obtain a detection result; the industrial equipment defect detection model is obtained by training a preset neighboring neural network model according to the historical operating state data of the industrial equipment; according to the detection result, the defect probability data is determined; according to the defect probability data, the industrial equipment adjustment parameters are obtained; according to the industrial equipment adjustment parameters, the industrial equipment is controlled to operate according to the industrial equipment adjustment parameters. It is possible to correct the equipment operating parameters in real time, improve the overall production efficiency, and reduce the probability of product defects. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a flow chart of a method for controlling the operating status of industrial equipment of the present invention;

[0050] Figure 2 It is a working process diagram of the method for controlling the operating state of industrial equipment of the present invention;

[0051] Figure 3 It is an Internet of Things structure diagram of an embodiment of a method for controlling the operating status of industrial equipment of the present invention;

[0052] Figure 4 It is a schematic diagram of the module structure of the control device for the operating status of industrial equipment of the present invention. DETAILED DESCRIPTION

[0053] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0054] like Figure 1 As shown, an embodiment of the present invention provides a method for controlling the operating state of industrial equipment, including:

[0055] Step S1, obtaining real-time status data of industrial equipment in actual working state in industrial production process;

[0056] Step S2, performing feature extraction processing on the real-time status data to obtain feature status data;

[0057] Step S3, inputting the characteristic state data into a trained industrial equipment defect detection model to perform defect detection and obtain a detection result; the industrial equipment defect detection model is obtained by training a preset neighboring neural network model according to the historical operating state data of the industrial equipment;

[0058] Step S4, determining defect probability data according to the detection results;

[0059] Step S5, obtaining industrial equipment adjustment parameters according to the defect probability data;

[0060] Step S6, according to the industrial equipment adjustment parameters, controlling the industrial equipment to operate according to the industrial equipment adjustment parameters.

[0061] In this embodiment, Figure 1 As shown, based on the equipment communication technology, the production equipment, auxiliary equipment, quality inspection equipment, thermometer and hygrometer, etc. are networked and communicated, and the corresponding operation data and test results of the equipment are obtained in real time. The real-time status data of the equipment is obtained through the equipment communication technology. After collection and accumulation within a predetermined time range, a predetermined number of implementation status data are obtained, and then the implementation status data is subjected to feature extraction processing to obtain multiple groups of feature status data. With the help of machine learning algorithms, the historical operation status data of product defects are analyzed again, a product defect model is constructed, an algorithm is formed, and continuous training and prediction evaluation are carried out to gradually improve the accuracy of the algorithm. The product defect algorithm is applied to the actual production process of the equipment, and real-time calculation and prediction of various data are performed. According to the prediction probability results, various parameters in the equipment production process are controlled and corrected in advance to reduce the probability of product defects, thereby improving the equipment processing efficiency.

[0062] In an optional embodiment of the present invention, step S2 performs feature extraction processing on the real-time status data to obtain feature status data, including:

[0063] Step S21, cleaning the real-time status data to obtain first data;

[0064] Step S22, preprocessing the first data to obtain second data;

[0065] Step S23, performing feature extraction processing on the second data to obtain feature state data.

[0066] In this embodiment, real-time status data of production equipment is collected, which may include real-time operating status data and quality inspection data. Communication is established with each device through the IoT terminal hardware. The terminal hardware is installed in a non-invasive manner, with millisecond-level two-way data communication, and the data is cleaned, pre-processed, and persisted to ensure the accuracy and objectivity of the data.

[0067] Data cleaning is to set a data rationality threshold range for each real-time status data, discard the data that exceeds the range, and prevent data fluctuations from affecting the model algorithm.

[0068] Data preprocessing performs preprocessing calculations on real-time status data. For example, a certain pressure parameter value (102.85Mpa) is composed of two original data (102 and 85), which represent the integer part and decimal part of the pressure value respectively. The terminal preprocesses the data according to the rules and reports the value of 102.85.

[0069] Data persistence is to store the real-time status data after cleaning and preprocessing in the database for persistent storage. Among them, the time series database is specially used to store and process time series data, emphasizing high-performance time range query, aggregation and analysis. It is particularly suitable for scenarios such as the Internet of Things, monitoring systems, sensor data, etc. that need to process a large amount of time series data. In this solution, it is used to store data collected by devices in real time; the relational database uses row storage, each record contains multiple fields, supports complex data relationships and diverse data types, and is mainly suitable for storing structural business data. In this solution, data such as products and defects are stored;

[0070] Feature extraction processing mainly extracts features from real-time status data parameters related to equipment operation to obtain status data directly related to product defects.

[0071] In an optional embodiment of the present invention, the training process of the industrial equipment defect detection model in step S3 includes:

[0072] Step S31, obtaining historical status data of the industrial equipment within a preset time period, wherein the historical status data includes defect status data;

[0073] Step S32, extracting features from the defect state data to obtain defect state feature data;

[0074] Step S33, inputting the defect state characteristic data into the first training module of the preset proximity neural network model for processing to obtain a first processing result;

[0075] Step S34, inputting the first processing result into the second training module of the preset proximity neural network model for processing to obtain a second processing result;

[0076] Step S35, determining the defect type according to the second processing result, and obtaining an industrial equipment defect detection model.

[0077] In this embodiment, for different product defects, the state data when the product defects occur is extracted, and after a certain amount of accumulation, a training data set is formed, and then the K-Nearest Neighbors (KNN) algorithm is used for training; the K-Nearest Neighbors (KNN) algorithm is an instance-based learning method, which occupies an important position in the field of machine learning with its concise and clear ideas and wide applicability. The core idea of ​​the algorithm is: for a new, unknown data point, by comparing its distance with the data points in the known category training set, find the K nearest neighbors, and determine the category of the new data point based on the majority category of these K neighbors;

[0078] Specifically, after accumulation within a preset time period, historical status data with complete elements are obtained, especially defect status data containing a predetermined number of defect status data, and then feature extraction is performed on the defect status data to obtain defect status feature data of key parameters of the equipment, and these accumulated defect status feature data are used as training data sets for defect detection models, and the defect status feature data are input into the first training module of the preset proximity neural network model for processing, and the first training module is used to calculate the distance between the specified data point and each data point in the existing training data set, and the distance is expressed by the sum of absolute wheelbases on the standard coordinate system to obtain a first processing result; secondly, the first processing result is input into the second training module of the preset proximity neural network model for processing, and the second training module is used to sort the first processing result to obtain K data points closest to the specified data point, where K is a preset value, and the obtained K data points are the second processing results; finally, the defect type corresponding to the specified data point is determined according to the second processing result, and after multiple rounds of iterations, the defect type of each data point is determined, and then the industrial equipment defect detection model is obtained;

[0079] The specific model training process can be seen in the following embodiment:

[0080] 1. Training Dataset

[0081] Mold temperature(℃) Clamping pressure (T) Defect Classification 230 80 Flow marks and patterns 250 80 Flow marks and patterns 280 80 Web-like hairy wings (cracks) 250 70 Unclear outline 250 75 Sharp corners are not fully filled 250 80 Metal flow traces

[0082] 2. Test sample: [Mold temperature: 245, clamping pressure: 80]

[0083] 3. Set K = 3

[0084] Output:

[0085] Bring the above parameters into the model for training. The algorithm will calculate the distance between the test sample and each data point in the training set, and then select the three nearest data points. Assuming that the three nearest neighbors are two data points of the categories [flow marks and patterns] and one data point of the category [net-like hair wings (turtle cracks)], then according to the majority voting principle, the test sample will be classified into the category [flow marks and patterns].

[0086] Through the above algorithm and training process, all collected data are included in the calculation, and after learning and training the data samples in the previous step, various types of product defect model data are generated.

[0087] In an optional embodiment of the present invention, in step S3, the characteristic state data is input into a trained industrial equipment defect detection model to perform defect detection to obtain a detection result, including:

[0088] Step S36, inputting the characteristic state data into a first processing module in the industrial equipment defect detection model for processing to obtain a first output;

[0089] Step S37: input the first output into a second processing module in the industrial equipment defect detection model for defect detection processing to obtain a detection result.

[0090] In this embodiment, the trained product defect model data is used to input the real-time characteristic status data of the equipment into the first processing module in the industrial equipment defect detection model for processing. The first processing module is used to calculate the distance between the real-time data point and each data point in the model data set. The distance is represented by the sum of the absolute wheelbases on the standard coordinate system to obtain a first output. The first output is then input into the second processing module in the industrial equipment defect detection model for processing. The second processing module is used to sort the first output to obtain the K data points closest to the real-time data points, where K is a preset value. The obtained K data points are the detection results.

[0091] In an optional embodiment of the present invention, step S4 determines defect probability data according to the detection result, including:

[0092] Step S41, normalizing the detection result to obtain a normalized result;

[0093] Step S42, determining defect probability data according to the normalization result.

[0094] In this embodiment, the detection results are normalized. The main purpose of normalization is to simplify calculations and improve the efficiency and accuracy of data processing. The normalization method converts dimensional expressions into dimensionless expressions so that the sum of the probabilities of all possible events in the detection results is equal to 1, thereby ensuring the integrity and consistency of the probability distribution.

[0095] In an optional embodiment of the present invention, in step S42, determining defect probability data according to the normalization result includes:

[0096] Step S421, obtaining likelihood probability, prior probability and marginal probability of feature state data according to the normalization result;

[0097] Step S422, obtaining defect probability data according to the likelihood probability, prior probability and marginal probability.

[0098] In this embodiment, the naive Bayes method is used to establish a computational model of the relationship between the likelihood probability of product defects, the prior probability of product defects, and the marginal probability of the state data; the naive Bayes method is a corresponding simplification on the basis of the Bayes algorithm, that is, it is assumed that the attributes are conditionally independent of each other when the target value is given, that is, no attribute variable occupies a larger proportion of the decision result, and no attribute variable occupies a smaller proportion of the decision result. Although this simplification method reduces the classification effect of the Bayesian classification algorithm to a certain extent, in actual application scenarios, it greatly simplifies the complexity of the Bayesian method and improves the computational efficiency.

[0099] Specifically, first, according to the product defect model, the likelihood probability of the real-time data having a product defect is obtained, which is expressed as P(X|C), where P represents probability, X represents the vector of real-time data, and C represents the product defect classification; when a product defect occurs, the distribution of the state data is a normal distribution, the mean of the state data is μ, and the variance is σ 2 , P(X|C) is expressed by To calculate; use P(C) to represent the prior probability of product defects, P(X) to represent the marginal probability of real-time data, and use the formula P(C|X)=P(X|C)P(C) / P(X) to obtain the probability P(C|X) that product defects will occur in the real-time data, and then calculate the probability P(C|X) that product defects will occur based on the real-time data.

[0100] In an optional embodiment of the present invention, step S5 obtains industrial equipment adjustment parameters according to the defect probability data, including:

[0101] Step S51, when the defect probability data is greater than a first preset value, the characteristic state data is compared with the standard value of the equipment operation parameter to obtain the industrial equipment adjustment parameter; otherwise, the industrial equipment adjustment parameter remains unchanged.

[0102] In this embodiment, the normal range standard value of the equipment operating parameters is first obtained. When the defect probability data is greater than the first preset value, in order to prevent product defects, it is necessary to intervene and adjust the equipment operating parameters, compare the real-time feature status data with the standard value of the equipment operating parameters, and use the difference between the real-time feature status data and the standard value of the equipment operating parameters as the industrial equipment adjustment parameter; when the defect probability data is not greater than the first preset value, the product defect is within an acceptable controlled range, the industrial equipment adjustment parameters remain unchanged, and the equipment operates normally to produce products.

[0103] like Figure 2 As shown, the specific implementation process of the above method of the present invention is described below by taking the production process of a die-casting machine as an example:

[0104] The parameters involved in die casting machine production are as follows:

[0105] Die casting machine parameters: clamping pressure, injection pressure, injection speed;

[0106] Mold parameters: holding time, cooling time, mold temperature;

[0107] Material parameters: alloy composition, melting temperature, melting temperature;

[0108] Environmental parameters: temperature, humidity;

[0109] Take NGType1 = Surface has patterns and metal flow marks. This defect is taken as an example, and the processing process is described as follows:

[0110] 1. If Figure 3 As shown, the terminal performs millisecond-level two-way communication with each device through the communication driver to obtain real-time data.

[0111] The terminal runs the acquisition driver of the corresponding device controller, such as the Siemens S7-200smart driver. The acquisition driver communicates with the Siemens PLC in real time through the S7 protocol, collects data related to the device operation into the terminal and performs the next step of preprocessing.

[0112] 2. Clean, preprocess, and persist real-time data, and finally store it in the time series database and relationship database.

[0113] Cleaning: Set a data rationality threshold range for each parameter, discard data that exceeds the range, and prevent data fluctuations from affecting model calculations.

[0114] Preprocessing: For some data that need to be calculated, preprocessing calculations are performed. For example, a certain pressure parameter value (102.85Mpa) is composed of two original data (102 and 85), which represent the integer part and decimal part of the pressure value respectively. The terminal preprocesses the data according to the rules and reports the value of 102.85.

[0115] Persistence: The cleaned and preprocessed data is stored in the database for persistence.

[0116] Time series database: It is specially used to store and process time series data, emphasizing high-performance time range query, aggregation and analysis. It is particularly suitable for scenarios that need to process large amounts of time series data, such as the Internet of Things, monitoring systems, and sensor data. It is used in this solution to store data collected by devices in real time.

[0117] Relational database: uses row storage, each record contains multiple fields, supports complex data relationships and diverse data types. It is mainly suitable for storing structural business data. This solution stores product and defect data.

[0118] 3. Machine learning algorithm, continuously obtains training sample data from the database.

[0119] The defect model library adopts supervised learning algorithm, specifically using K-nearest neighbor algorithm and naive Bayes algorithm.

[0120] The K-Nearest Neighbors (KNN) algorithm is an instance-based learning method that occupies an important position in the field of machine learning for its concise and clear ideas and wide applicability. The core idea of ​​the algorithm is: for a new data point of unknown category, by comparing its distance with the data points in the training set of known categories, find its nearest K neighbors, and determine the category of the new data point based on the majority category of these K neighbors.

[0121] enter:

[0122] 1. Training Dataset

[0123]

[0124]

[0125] 2. Test sample: [Mold temperature: 245, clamping pressure: 80]

[0126] 3. Set K = 3

[0127] Output:

[0128] Bring the above parameters into the model for training. The algorithm will calculate the distance between the test sample and each data point in the training set, and then select the three nearest data points. Assuming that the three nearest neighbors are two data points of the categories [flow marks and patterns] and one data point of the category [net-like hair wings (turtle cracks)], then according to the majority voting principle, the test sample will be classified into the category [flow marks and patterns].

[0129] Through the above algorithm and training process, all collected data are included in the calculation.

[0130] After learning and training the data samples in the previous step, various types of product defect model data are generated.

[0131] 4. Store various product defect data calculated by the KNN algorithm into the database.

[0132] The various defects calculated by KNN, when a certain parameter data of the equipment is in a certain range, the possible defect types that may occur, all the data and defect classifications obtained after training are stored in the relational database for use in the next stage.

[0133] 5. By using the Naive Bayes method, the actual collected data and defect model data are calculated in real time to obtain the parameter adjustment of the equipment.

[0134] The probability of each defect type and the conditional probability of the features are calculated by the Naive Bayes method.

[0135] The Naive Bayes method is a corresponding simplification of the Bayesian algorithm, that is, it assumes that the attributes are conditionally independent when the target value is given. In other words, no attribute variable has a large proportion of the decision result, and no attribute variable has a small proportion of the decision result. Although this simplification method reduces the classification effect of the Bayesian classification algorithm to a certain extent, it greatly simplifies the complexity of the Bayesian method in actual application scenarios.

[0136] Processing process:

[0137] Bayes' theorem can be stated as:

[0138] P(C|X)=P(X|C)P(C) / P(X)

[0139] Where C is a defect in the product defect model dataset

[0140] x is a feature vector: the mold temperature value collected in real time (245℃)

[0141] P(C|X) is the posterior probability of category C (defect classification) after observing feature X (mold temperature).

[0142] P(X|C) is the likelihood probability of feature X (mold temperature) under category C (defect classification).

[0143] P(C) is the prior probability of category C (defect classification).

[0144] P(X) is the marginal probability of feature X (mold temperature).

[0145] Through the calculation of the Naive Bayes method, it can be concluded that when the real-time data of a certain parameter is input, the output can be obtained after calculation: P(C|X), which outputs the probability of occurrence of defects in real time according to the real-time equipment parameters.

[0146] 6. Obtain real-time data of equipment operation from the real-time database in the terminal.

[0147] The Naive Bayes method obtains data from the terminal in real time and stores the device operation data in the time series database (TSDB) and relational database (MYSQL).

[0148] 7. Send the equipment parameter adjustment range to the production equipment through the terminal.

[0149] Assuming the temperature parameter is 235 degrees Celsius, the probability of defect (X) occurring is calculated by the naive Bayes method to be P(C|X)=60%. If it exceeds 50%, it means that this temperature is likely to cause product defects (X).

[0150] By comparing the temperature parameters with the standard (230 degrees Celsius) in the defect model database, it is determined whether the temperature is too high or too low, and compared with the standard temperature value, the temperature difference (5 degrees Celsius) is obtained. The temperature difference is used as the parameter adjustment range to modify the real-time parameters of the equipment. This information is displayed to the user through the terminal.

[0151] 8. The cloud platform manages and updates the data training algorithms in the terminal. It continuously improves the equipment processing model based on the actual equipment acquisition and interference conditions, and makes timely adjustments after discovering new influencing factors, thereby improving the accuracy and execution efficiency of the model.

[0152] 9. The product defect model and feature data in the terminal are uploaded to the cloud platform for management, and platform users view and manage the data.

[0153] 10. The cloud platform manages the real-time calculation algorithms and updates the versions of defects within the terminal.

[0154] like Figure 4 As shown, an embodiment of the present invention further provides a control device 40 for the operating status of industrial equipment, comprising:

[0155] An acquisition module 41 is used to acquire real-time status data of industrial equipment in an actual working state in an industrial production process;

[0156] The processing module 42 is used to perform feature extraction processing on the real-time status data to obtain feature status data; input the feature status data into a trained industrial equipment defect detection model to perform defect detection and obtain a detection result; the industrial equipment defect detection model is obtained by training a preset proximity neural network model based on the historical operating status data of the industrial equipment; based on the detection result, defect probability data is determined; based on the defect probability data, industrial equipment adjustment parameters are obtained; based on the industrial equipment adjustment parameters, the industrial equipment is controlled to operate according to the industrial equipment adjustment parameters.

[0157] Optionally, performing feature extraction processing on the real-time status data to obtain feature status data includes:

[0158] Cleaning the real-time status data to obtain first data;

[0159] Preprocessing the first data to obtain second data;

[0160] Perform feature extraction processing on the second data to obtain feature state data.

[0161] Optionally, the training process of the industrial equipment defect detection model includes:

[0162] Acquire historical status data of industrial equipment within a preset time period, wherein the historical status data includes defect status data;

[0163] Extracting features from the defect state data to obtain defect state feature data;

[0164] Inputting the defect state characteristic data into the first training module of the preset proximity neural network model for processing to obtain a first processing result;

[0165] Inputting the first processing result into the second training module of the preset proximity neural network model for processing to obtain a second processing result;

[0166] The defect type is determined according to the second processing result to obtain an industrial equipment defect detection model.

[0167] Optionally, the characteristic state data is input into a trained industrial equipment defect detection model to perform defect detection to obtain a detection result, including:

[0168] Inputting the characteristic state data into a first processing module in an industrial equipment defect detection model for processing to obtain a first output;

[0169] The first output is input into a second processing module in an industrial equipment defect detection model to perform defect detection processing to obtain a detection result.

[0170] Optionally, determining defect probability data according to the detection result includes:

[0171] Normalizing the detection result to obtain a normalized result;

[0172] Defect probability data is determined based on the normalized result.

[0173] Optionally, determining defect probability data according to the normalization result includes:

[0174] According to the normalization result, the likelihood probability, prior probability, and marginal probability of the feature state data are obtained;

[0175] Defect probability data is obtained according to the likelihood probability, prior probability and marginal probability.

[0176] Optionally, obtaining industrial equipment adjustment parameters according to the defect probability data includes:

[0177] When the defect probability data is greater than a first preset value, the characteristic state data is compared with a standard value of an equipment operation parameter to obtain an industrial equipment adjustment parameter;

[0178] Otherwise, the industrial equipment adjustment parameters remain unchanged.

[0179] It should be noted that the device is a device that corresponds one-to-one to the above method, and all implementation methods in the above method embodiments can be applied to the embodiments of the device and can achieve the same technical effects.

[0180] An embodiment of the present invention further provides a computing device, comprising:

[0181] one or more processors;

[0182] A storage device is used to store one or more programs. When the one or more programs are executed by the one or more processors, the one or more processors implement the control method of the operating state of industrial equipment described in the present invention. All implementation methods in the above method embodiments are applicable to the embodiments of the device and can achieve the same technical effect.

[0183] An embodiment of the present invention further provides a computing device readable storage medium, comprising:

[0184] The computing device readable storage medium stores a program, which, when executed by the processor, implements the method for controlling the operating state of industrial equipment described in the present invention. All implementations in the above method embodiments are applicable to the embodiments of the computing device readable storage medium, and can achieve the same technical effects.

[0185] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0186] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0187] In the embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0188] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0189] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0190] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks.

[0191] In addition, it should be noted that in the apparatus and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. Moreover, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order, and some steps can be performed in parallel or independently of each other. For those of ordinary skill in the art, it is understood that all or any steps or components of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or a network of computing devices in hardware, firmware, software or a combination thereof, which can be achieved by those of ordinary skill in the art using their basic programming skills after reading the description of the present invention.

[0192] Therefore, the purpose of the present invention can also be achieved by running a program or a group of programs on any computing device. The computing device can be a well-known general-purpose device. Therefore, the purpose of the present invention can also be achieved by simply providing a program product containing a program code for implementing a method or device. That is to say, such a program product also constitutes the present invention, and a storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any well-known storage medium or any storage medium developed in the future. It should also be pointed out that in the device and method of the present invention, it is obvious that each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. In addition, the steps of performing the above-mentioned series of processing can naturally be performed in chronological order according to the order of description, but it is not necessary to perform them in chronological order. Some steps can be performed in parallel or independently of each other.

[0193] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for controlling the operating status of industrial equipment, characterized in that: include: Obtain real-time status data of industrial equipment under actual working conditions in industrial production processes; Performing feature extraction processing on the real-time status data to obtain feature status data; The characteristic state data is input into a trained industrial equipment defect detection model to perform defect detection and obtain a detection result; the industrial equipment defect detection model is obtained by training a preset neighboring neural network model according to the historical operating state data of the industrial equipment; Determining defect probability data based on the detection results; Obtaining industrial equipment adjustment parameters according to the defect probability data; According to the industrial equipment adjustment parameters, the industrial equipment is controlled to operate according to the industrial equipment adjustment parameters.

2. The method for controlling the operating status of industrial equipment according to claim 1, characterized in that: Performing feature extraction processing on the real-time status data, Get feature status data, including: Cleaning the real-time status data to obtain first data; Preprocessing the first data to obtain second data; Perform feature extraction processing on the second data to obtain feature state data.

3. The method for controlling the operating status of industrial equipment according to claim 1, characterized in that: The training process of the industrial equipment defect detection model includes: Acquire historical status data of industrial equipment within a preset time period, wherein the historical status data includes defect status data; Extracting features from the defect state data to obtain defect state feature data; Inputting the defect state characteristic data into the first training module of the preset proximity neural network model for processing to obtain a first processing result; Inputting the first processing result into the second training module of the preset proximity neural network model for processing to obtain a second processing result; The defect type is determined according to the second processing result to obtain an industrial equipment defect detection model.

4. The method for controlling the operating status of industrial equipment according to claim 1, characterized in that: The characteristic state data is input into a trained industrial equipment defect detection model to perform defect detection and obtain a detection result, including: Inputting the characteristic state data into a first processing module in an industrial equipment defect detection model for processing to obtain a first output; The first output is input into a second processing module in an industrial equipment defect detection model for defect detection processing to obtain a detection result.

5. The method for controlling the operating status of industrial equipment according to claim 1, characterized in that: Determine defect probability data based on the test results, including: Normalizing the detection result to obtain a normalized result; Defect probability data is determined based on the normalized result.

6. The method for controlling the operating status of industrial equipment according to claim 5, characterized in that: Determine defect probability data according to the normalized result, including: According to the normalization result, the likelihood probability, prior probability and marginal probability of the feature state data are obtained; Defect probability data is obtained according to the likelihood probability, prior probability and marginal probability.

7. The method for controlling the operating status of industrial equipment according to claim 1, characterized in that: According to the defect probability data, industrial equipment adjustment parameters are obtained, including: When the defect probability data is greater than a first preset value, the characteristic state data is compared with the standard value of the equipment operation parameter to obtain the industrial equipment adjustment parameter; otherwise, the industrial equipment adjustment parameter remains unchanged.

8. A control device for the operating status of industrial equipment, characterized in that: include: An acquisition module is used to obtain real-time status data of industrial equipment under actual working conditions in industrial production processes; A processing module is used to perform feature extraction processing on the real-time status data to obtain feature status data; input the feature status data into a trained industrial equipment defect detection model to perform defect detection and obtain a detection result; the industrial equipment defect detection model is obtained by training a preset neighboring neural network model according to the historical operating status data of the industrial equipment; and determine defect probability data according to the detection result; According to the defect probability data, industrial equipment adjustment parameters are obtained; according to the industrial equipment adjustment parameters, the industrial equipment is controlled to operate according to the industrial equipment adjustment parameters.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the method as claimed in any one of claims 1 to 7.

10. A computing device readable storage medium, characterized in that: The computing device readable storage medium stores a program, which implements the method according to any one of claims 1 to 7 when executed by a processor.

Images