Industrial algorithm model scheduling method and system based on industrial big data

Abstract

The present invention relates to the technical field of industrial big data processing, and an industrial algorithm model scheduling method and system based on industrial big data. The original industrial data collected by factory sensors is obtained, and the data is preprocessed; for different types of industrial problems such as equipment failure prediction, production scheduling and energy consumption, corresponding algorithm models such as support vector machines, genetic algorithms and linear regression are constructed; an industrial problem processing model scheduling mechanism is established, the corresponding algorithm model is triggered according to the industrial data, and the data is input into the model for calculation, and the corresponding optimization result is output. The present invention can effectively improve the utilization efficiency of industrial data and optimize the industrial production process by fusing multi-source industrial data, constructing a targeted industrial algorithm model, and realizing adaptive scheduling of the model, and has good applicability and promotion value.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial big data processing, and particularly to an industrial algorithm model scheduling method and system based on industrial big data. Background Art

[0002] With the rapid development of industrial Internet, big data and artificial intelligence technologies, massive amounts of data have been generated in the industrial field. How to effectively utilize industrial big data, improve industrial production efficiency and reduce production costs has become a problem that needs to be urgently solved in the industrial field.

[0003] Traditional industrial production management mainly relies on manual experience and pre-set rules, which is difficult to adapt to the increasingly complex industrial production environment. Problems such as industrial equipment failure, production scheduling and energy management often require a lot of manpower and material resources to deal with, which is inefficient and costly. With the continuous accumulation of industrial data, the use of data-driven intelligent algorithms to optimize industrial production processes has become a research hotspot in the industrial field.

[0004] Therefore, there is an urgent need for an industrial algorithm model scheduling method based on industrial big data. By integrating multi-source heterogeneous industrial data, customized algorithm models for different industrial problems are constructed, industrial data can be analyzed in batches, and the most appropriate algorithm model can be scheduled for specific industrial problems. An efficient model scheduling mechanism can be established to improve industrial production efficiency, reduce production costs, and promote the intelligent development of industry.

[0005] Current existing technologies lack the comprehensive utilization of industrial data. Different types of data, such as equipment data, production data, energy data, etc., are often processed separately, making it difficult to give play to the synergistic effect of data: industrial algorithm models are relatively simple, and there is a lack of customized algorithm models for different industrial scenarios. The algorithm model is disconnected from actual industrial production, and there is a lack of efficient algorithm deployment and scheduling mechanisms, which affects the application effect of the algorithm in the actual industrial environment.

[0006] In view of this, the present invention proposes an industrial algorithm model scheduling method and system based on industrial big data Summary of the invention

[0007] To achieve the above-mentioned object, the present invention provides the following technical solution: an industrial algorithm model scheduling method based on industrial big data, characterized in that it includes: obtaining original industrial data collected by factory sensors and preprocessing the original industrial data;

[0008] Define different types of industrial problems, including equipment failure prediction, production scheduling and energy consumption problems. Construct corresponding industrial problem processing models for different industrial problems, construct a support vector machine (SVM) model to predict equipment failure problems, construct a genetic algorithm model to adjust production scheduling problems, and construct a multivariate linear regression model to calculate energy consumption problems.

[0009] Establish an industrial problem processing model scheduling mechanism. When the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data will be input into the corresponding industrial problem processing model;

[0010] The corresponding industrial problem processing model performs calculations in the model based on the received data and outputs the data results of the corresponding industrial problem.

[0011] Preferably, the original industrial data collected by the factory sensors are obtained, the original industrial data including industrial equipment data, product production data and energy data during industrial production; the industrial equipment data includes temperature data, vibration data and current data of the industrial equipment; the product production data includes production material consumption data and product output data; the energy data includes energy price data and energy consumption data;

[0012] Preprocess the original industrial data to remove missing values, outliers and duplicate values, and normalize the original industrial data.

[0013] Preferably, based on the temperature data, vibration data and current data of the industrial equipment, a support vector machine (SVM) model is constructed to predict equipment failure problems;

[0014] Collect historical operation data of industrial equipment, including temperature data, vibration data and current data, and corresponding timestamps; collect historical fault data of industrial equipment, including the time when the fault occurred and the type of fault, and digitally encode it;

[0015] Label the historical fault data, align the standardized time series data with the fault events, generate fault labels, use -1 to indicate that the industrial equipment is normal, and use +1 to indicate that the industrial equipment is faulty. Randomly divide the data set into training set and test set, and use the training set and test set to train and test the support vector machine (SVM) model.

[0016] The training process of the support vector machine (SVM) model to predict equipment failure is as follows:

[0017] The temperature data T′ of industrial equipment after preprocessing and normalization q , vibration data V′ q and current data I′ q Perform feature extraction and generate feature vector x q , x q =[T′ q , V′ q , I′ q ], q = 1, 2, ..., n, n is the eigenvector x q The number of data;

[0018] Given a training set (x1, y1), (x2, y2), ..., (x n ,y n ), where y q represents the fault label corresponding to the qth feature vector, y q ∈{-1, +1}; the optimization problem of predicting equipment failure by support vector machine SVM model is expressed as:

[0019]

[0020]

[0021] Among them, α q is the Lagrange multiplier, C is the penalty coefficient, which is used to control the load of the model and the tolerance of misclassification. represents the transpose of the eigenvector, and the optimal solution is obtained Construct the decision function of the support vector machine SVM model:

[0022]

[0023] Where x represents the input vector; sign(·) is a sign function; using the training set data, by solving the support vector machine SVM model optimization problem, we get and b* values, construct a support vector machine (SVM) model, and predict the future working status of the equipment through the collected industrial equipment data:

[0024]

[0025] in, represents the predicted fault label, K(xq, x) is the kernel function used to map features to a high-dimensional space; if It is predicted that the industrial equipment will be normal in the future period. It is predicted that industrial equipment will fail in the future and that the industrial equipment needs to be repaired in advance.

[0026] Preferably, a genetic algorithm model is constructed to adjust the production scheduling problem based on production material consumption data and product output data;

[0027] Collecting production material consumption data and product output data during the production process, wherein the production material consumption data includes the consumption, consumption time and inventory of raw materials; the product output data includes the production quantity, production time and quality qualification rate of the products; minimizing production costs and maximizing product quality are the production scheduling goals;

[0028] Define the production decision variable, A ijrepresents the quantity of the jth product produced in the i-th time period; B ik represents the amount of the kth material consumed in the i-th time period; Z j represents the qualified rate of the jth product;

[0029] Define the production objective function, the objective function to minimize production cost is: min(∑ i ∑ k C k ×B ik ), k = 1, 2, ..., K, where C k represents the unit cost of the kth material; the objective function for maximizing product quality is: max(∑ j Z j ×∑ i A ij ), i = 1, 2, ..., I; j = 1, 2, ..., J, the maximum product quality is the weighted sum of product qualification rate and output;

[0030] Define the constraints, the material consumption constraint is: ∑ j (q jk ×A ij )≤B ik , where q jk represents the quantity of the kth material required to produce a jth product; the material inventory constraint is: ∑ i (B ik )≤S k , where S k represents the current inventory of the kth material; production capacity constraint: ∑ j (t j ×A ij )≤T i , where t j represents the time required to produce a j-th product, T i represents the production time in the i-th time period; A ij ≥0, B ik ≥0, Z j ∈[0, 1];

[0031] The steps to construct the genetic algorithm are as follows:

[0032] Coding method: Real number coding is used. Each chromosome consists of (|I|×|J|+|I|×|K|+|J|) genes, where I, J and K correspond to the maximum values ​​of i, j and k.

[0033] Randomly generate N feasible solutions that meet the constraints as the initial population;

[0034] Normalize the objective function of minimizing production cost and the objective function of maximizing product quality, and transform the two objective functions into a single objective fitness function by weighted sum;

[0035] A linear ranking selection operator based on sorting is used to assign the operator a probability of being selected according to the fitness ranking of the individual;

[0036] The simulated binary crossover operator is used to generate two offspring individuals based on two parent individuals, and the polynomial mutation operator is used to perturb each gene with a certain probability;

[0037] The maximum evolutionary number is set as the termination condition, and the genetic algorithm calculation is stopped when the maximum evolutionary number is reached.

[0038] Preferably, based on the temperature data, energy price data and energy consumption data of industrial equipment, a multivariate linear regression model is constructed to calculate the energy consumption problem; the energy is electrical energy;

[0039] Collect electricity price data within the historical time period, including peak electricity prices and valley electricity prices, collect electricity consumption data within the historical time period, including peak electricity consumption and valley electricity consumption, and collect average temperature data within the historical time period as weather factors that affect electricity consumption;

[0040] The peak-valley electricity price difference is calculated as the electricity price feature, the peak-valley electricity consumption difference is calculated as the load feature, and the temperature data is standardized, that is, the mean is subtracted and divided by the standard deviation to obtain the standardized temperature feature; the electricity price difference, load difference and standardized temperature are selected as the input features of the regression model;

[0041] Construct a multiple linear regression model, and the regression equation is:

[0042]

[0043] in, is the electricity consumption in the predicted future time period, m1 is the electricity price difference, m2 is the load difference, and m3 is the standardized temperature. The regression coefficients β1, β2, β3, the intercept term β0 and the error term ∈ of the regression equation are obtained by solving the data in the historical time period.

[0044] Preferably, an industrial problem processing model scheduling mechanism is established. When the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data is input into the corresponding industrial problem processing model, specifically including:

[0045] The trigger mechanism of the support vector machine (SVM) model predicts equipment failure problems as follows:

[0046] Let T represent the temperature data of industrial equipment, T min and T maxRespectively represent the lower and upper limits of the normal temperature range; V represents the vibration data of industrial equipment obtained, V th represents the set vibration change threshold; I represents the acquired industrial equipment current data, I th Indicates the set current change threshold when: Trigger the support vector machine (SVM) model to predict equipment failure problems;

[0047] Genetic algorithm model adjusts the trigger mechanism of production scheduling problem:

[0048] Order I k represents the inventory of the kth production material, THS k represents the safety stock level of the kth production material; Q j represents the quality pass rate of the jth product, Q t Indicates the preset quality pass rate target; when: Trigger the genetic algorithm model;

[0049] The trigger mechanism for calculating energy consumption using the linear regression model is:

[0050] Let Δp denote the difference in electricity prices, Δp th represents the electricity price trigger threshold; Δl represents the load difference, Δl th represents the trigger load difference threshold; t represents temperature data, μ t and σ t Indicates the mean and standard deviation of temperature data in the historical period, when:

[0051] Δp≥Δp th ∨Δl≥Δl th ∨[(t<μ t -2σ t )∨(t>μ t +2σ t )]

[0052] Trigger the linear regression model to calculate energy consumption problem.

[0053] Preferably, through the industrial problem processing model scheduling mechanism, when the collected industrial data changes, the corresponding industrial problem processing model is triggered, and the industrial data is transmitted to the corresponding model. The corresponding industrial problem processing model performs calculations in the model based on the received data and outputs the data results of the corresponding industrial problem.

[0054] An industrial algorithm model scheduling system based on industrial big data, which is implemented based on the industrial algorithm model scheduling method based on industrial big data, includes: a data acquisition and processing module, an industrial model module, an industrial model scheduling module and a data output module;

[0055] The data acquisition and processing module is used to acquire the original industrial data collected by the factory sensors and pre-process the original industrial data;

[0056] The industrial model module is used to define different types of industrial problems, including equipment failure prediction, production scheduling and energy consumption. For different industrial problems, corresponding industrial problem processing models are constructed, including a support vector machine (SVM) model for predicting equipment failure problems, a genetic algorithm model for adjusting production scheduling problems, and a multivariate linear regression model for calculating energy consumption problems.

[0057] The industrial model scheduling module is used to establish an industrial problem processing model scheduling mechanism. When the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data is input into the corresponding industrial problem processing model;

[0058] The data output module and the corresponding industrial problem processing model perform calculations in the model based on the received data and output data results corresponding to the industrial problem.

[0059] An electronic device comprises: a processor and a memory, wherein the memory stores a computer program that can be called by the processor; the processor executes the industrial algorithm model scheduling method based on industrial big data by calling the computer program stored in the memory.

[0060] A computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the industrial algorithm model scheduling method based on industrial big data.

[0061] Beneficial effects of the present invention: The present invention improves the quality and availability of industrial data by preprocessing the original industrial data collected by factory sensors and removing redundant and invalid data; at the same time, different types of industrial data such as equipment data, production data, energy data, etc. are integrated and processed to explore the correlation between data and make full use of the value of industrial data.

[0062] The present invention constructs a targeted algorithm model for typical problems such as equipment failure, production scheduling and energy consumption in the industrial field; adopts support vector machines to predict equipment failure, and uses equipment operation data to perform health status monitoring and fault warning; adopts genetic algorithms to optimize production scheduling, and improves production efficiency and product quality through intelligent scheduling; adopts linear regression models to optimize energy consumption, and realizes economic operation of energy through energy consumption data analysis; compared with general algorithms, targeted industrial algorithm models can better meet the needs of industrial scenarios.

[0063] The present invention establishes an industrial problem processing model scheduling mechanism, which automatically triggers the corresponding algorithm model according to the real-time status of industrial equipment and sensor data; by setting the model trigger threshold, the adaptive calling of the industrial algorithm model is realized, which reduces manual intervention and improves the automation level of the industrial production process.

[0064] The industrial algorithm model of the present invention performs intelligent calculations in the model based on the input industrial data and outputs corresponding optimization control results. It reduces unplanned equipment downtime, shortens product production cycles, and reduces energy consumption costs through measures such as equipment fault early warning, production scheduling optimization, and energy optimization, thereby achieving the purpose of optimizing the industrial production process and improving industrial production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a flow chart of the industrial algorithm model scheduling method based on industrial big data of the present invention;

[0066] Figure 2 This is a structural diagram of the industrial algorithm model scheduling system based on industrial big data of the present invention;

[0067] Figure 3 It is a schematic diagram of the structure of an electronic device provided by the present invention;

[0068] Figure 4 It is a schematic diagram of the structure of the computer-readable storage medium provided by the present invention. DETAILED DESCRIPTION

[0069] In order to better understand the present application, a more detailed description will be made of various aspects of the present application with reference to the accompanying drawings. It should be understood that these detailed descriptions are only descriptions of exemplary embodiments of the present application, and are not intended to limit the scope of the present application in any way. Throughout the specification, the same reference numerals refer to the same elements. The expression "and / or" includes any and all combinations of one or more of the associated listed items.

[0070] In the accompanying drawings, the size, dimensions and shapes of the elements have been slightly adjusted for ease of illustration. The drawings are for illustration only and are not strictly drawn to scale. As used herein, the terms "substantially", "approximately" and similar terms are used as terms of approximation, not as terms of degree, and are intended to illustrate the inherent deviations in measurements or calculations that will be recognized by those of ordinary skill in the art. In addition, in this application, the order in which the steps are described does not necessarily represent the order in which these processes occur in actual operation, unless otherwise specified or can be derived from the context.

[0071] It should also be understood that expressions such as "include", "including", "have", "contain" and / or "comprising" are open rather than closed expressions in this specification, which indicate the presence of the stated features, elements and / or components, but do not exclude the presence of one or more other features, elements, components and / or combinations thereof. In addition, when expressions such as "at least one of..." appear after a list of listed features, they modify the entire list of features rather than just the individual elements in the list. In addition, when describing embodiments of the present application, "may" is used to mean "one or more embodiments of the present application". And, the term "exemplary" is intended to refer to an example or illustration.

[0072] Unless otherwise specified, all words (including engineering terms and scientific and technological terms) used in this article have the same meaning as those commonly understood by ordinary technicians in the field to which this application belongs. It should also be understood that, unless clearly stated in this application, words defined in common dictionaries should be interpreted as having the same meaning as their meaning in the context of the relevant technology, and should not be interpreted in an idealized or overly formal sense.

[0073] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0074] Example 1

[0075] Reference Figure 1 The first embodiment of the present invention provides an industrial algorithm model scheduling method based on industrial big data.

[0076] Step 1: Obtain the original industrial data collected by factory sensors and preprocess the original industrial data.

[0077] Specifically, the original industrial data collected by factory sensors is obtained, and the obtained original industrial data includes industrial equipment data, product production data and energy data during industrial production. The industrial equipment data includes temperature data, vibration data and current data of industrial equipment; the product production data includes production material consumption data and product output data; the energy data includes energy price data and energy consumption data.

[0078] Install temperature, vibration, current and other sensors on industrial equipment to collect equipment operation data in real time; use PLC and DCS industrial control systems to collect and record product production data, including production material consumption and product output; obtain energy price and consumption data from energy management systems and supplier systems.

[0079] The acquired raw industrial data is preprocessed to remove missing values, outliers and duplicate values, and to normalize the raw industrial data.

[0080] Step 2: Define different types of industrial problems, including equipment failure prediction, production scheduling, and energy consumption. For different industrial problems, build corresponding industrial problem processing models, build a support vector machine (SVM) model to predict equipment failure problems, build a genetic algorithm model to adjust production scheduling problems, and build a multivariate linear regression model to calculate energy consumption problems; based on industrial equipment temperature data, vibration data, and current data, build a support vector machine (SVM) model to predict equipment failure.

[0081] Specifically, the historical operation data of the industrial equipment is collected, including temperature data, vibration data and current data, and corresponding timestamps; the historical fault data of the industrial equipment is collected, and the historical fault data includes the time when the fault occurred and the type of fault, and is digitally encoded.

[0082] The historical fault data is labeled, the standardized time series data is aligned with the fault events, and the fault labels are generated. -1 is used to indicate that the industrial equipment is normal, and +1 is used to indicate that the industrial equipment is faulty. The data set is randomly divided into training set and test set. The support vector machine (SVM) model is trained and tested using the training set and test set.

[0083] The training process of the support vector machine (SVM) model to predict equipment failure is as follows:

[0084] The temperature data T′ of industrial equipment after preprocessing and normalization q , vibration data v′ q and current data I′ q Perform feature extraction and generate feature vector x q , x q =[T′ q , V′ q , I′ q ], q = 1, 2, ..., n, n is the eigenvector x q The number of data.

[0085] Given a training set (x1, y1), (x2, y2), ..., (x n ,y n ), where y q represents the fault label corresponding to the qth feature vector, y q ∈{-1, +1}; by finding the best hyperplane w T x+b=0, so that the two types of samples can be correctly separated by the hyperplane, and the distance from the sample point (support vector) closest to the hyperplane to the hyperplane is maximized.

[0086] The optimization problem of predicting equipment failure using the support vector machine (SVM) model is expressed as:

[0087]

[0088] sty q (w T x q +b)≥1,q=1,2,...,n

[0089] Among them, w represents the weight vector, w T represents the transpose of the weight vector, and b represents the bias vector;

[0090] Using the Lagrange multiplier method and duality theory, the optimization problem is transformed into a dual problem:

[0091]

[0092]

[0093] Among them, α q is the Lagrange multiplier, C is the penalty coefficient, which is used to control the load of the model and the tolerance of misclassification. represents the transpose of the eigenvector; solving for the optimal Construct the decision function of the support vector machine SVM model:

[0094]

[0095] Where x represents the input vector; sign(·) is a sign function; using the training set data, by solving the support vector machine SVM model optimization problem, we get and b * Value, build support vector machine SVM model, and predict the future working status of the equipment through the collected industrial equipment data:

[0096]

[0097] in, represents the predicted fault label, K(x q , x) is the kernel function, which is used to map features to high-dimensional space; if It is predicted that the industrial equipment will be normal in the future period. It is predicted that industrial equipment will fail in the future and that the industrial equipment needs to be repaired in advance.

[0098] Furthermore, based on the production material consumption data and product output data, a genetic algorithm model is constructed to adjust the production scheduling problem.

[0099] The production material consumption data and product output data in the production process are collected. The production material consumption data includes the consumption, consumption time and inventory of raw materials; the product output data includes the production quantity, production time and quality qualification rate of the products; and the production scheduling goals are to minimize production costs and maximize product quality.

[0100] Define the production decision variable, A ij represents the quantity of the jth product produced in the i-th time period; B ik represents the amount of the kth material consumed in the i-th time period; Z j Represents the qualified rate of the jth product.

[0101] Define the production objective function, the objective function to minimize production cost is: min(∑ i ∑ k C k ×B ik ), k = 1, 2, ..., K, where C k represents the unit cost of the kth material; the objective function for maximizing product quality is: max(∑ j Z j ×∑ i A ij ), i = 1, 2, ..., I; j = 1, 2, ..., J, the maximized product quality is the weighted sum of the product qualification rate and the output.

[0102] Define the constraints, the material consumption constraint is: ∑ j (q jk ×A ij )≤B ik , where q jk represents the quantity of the kth material required to produce a jth product; the material inventory constraint is: ∑ i (B ik )≤S k , where S k represents the current inventory of the kth material; production capacity constraint: ∑ j (t j ×A ij )≤T i , where t j represents the time required to produce a j-th product, T i represents the production time in the i-th time period; A ij ≥0, B ik ≥0, Z j ∈[0, 1].

[0103] The steps to construct the genetic algorithm are as follows:

[0104] Coding method: Real number coding is used. Each chromosome consists of (|I|×|J|+|I|×|K|+|J|) genes, where I, J and K correspond to the maximum values ​​of i, j and k.

[0105] Initialize the population: randomly generate N feasible solutions that meet the constraints as the initial population;

[0106] Fitness function: Normalize the objective function of minimizing production cost and the objective function of maximizing product quality, and transform the two objective functions into a single objective fitness function by weighted sum;

[0107] Selection operator: A linear ranking selection operator based on sorting is used to assign the operator a probability of being selected according to the fitness ranking of the individual;

[0108] Crossover operator: A simulated binary crossover operator is used to generate two offspring individuals based on two parent individuals, and a polynomial mutation operator is used to perturb each gene with a certain probability;

[0109] Termination condition: Set the maximum evolutionary number as the termination condition, and stop the calculation of the genetic algorithm when the maximum evolutionary number is reached.

[0110] Furthermore, based on the temperature data, energy price data and energy consumption data of industrial equipment, a multivariate linear regression model is constructed to calculate the energy consumption problem; the energy is electrical energy.

[0111] Collect electricity price data within the historical time period, including peak electricity prices and valley electricity prices, collect electricity consumption data within the historical time period, including peak electricity consumption and valley electricity consumption, and collect average temperature data within the historical time period as weather factors affecting electricity consumption.

[0112] The peak-valley electricity price difference is calculated as the electricity price feature, the peak-valley electricity consumption difference is calculated as the load feature, and the temperature data is standardized, that is, the mean is subtracted and divided by the standard deviation to obtain the standardized temperature feature; the electricity price difference, load difference and standardized temperature are selected as the input features of the regression model.

[0113] Construct a multiple linear regression model, and the regression equation is:

[0114]

[0115] in, is the electricity consumption in the predicted future time period, m1 is the electricity price difference, m2 is the load difference, and m3 is the standardized temperature. The regression coefficients β1, β2, β3, the intercept term β0 and the error term ∈ of the regression equation are obtained by solving the data in the historical time period.

[0116] Step 3: Establish an industrial problem processing model scheduling mechanism. When the data collected by industrial equipment and factory sensors trigger the industrial problem processing model scheduling mechanism, the collected data is input into the corresponding industrial problem processing model.

[0117] Specifically, the triggering mechanism of the support vector machine (SVM) model to predict equipment failure is:

[0118] Let T represent the temperature data of industrial equipment, T min and T max Respectively represent the lower and upper limits of the normal temperature range; V represents the vibration data of industrial equipment obtained, V th represents the set vibration change threshold; I represents the acquired industrial equipment current data, I th Indicates the set current change threshold when:

[0119]

[0120] That is, when the temperature data of the industrial equipment obtained is not within the lower limit and upper limit of the normal temperature range, the vibration data of the industrial equipment obtained is not within the vibration data change threshold range of the industrial equipment, and the current data of the industrial equipment obtained is not within the current change threshold, the support vector machine SVM model is triggered to predict the equipment failure problem.

[0121] Genetic algorithm model adjusts the trigger mechanism of production scheduling problem:

[0122] Order I k represents the inventory of the kth production material, THS k represents the safety stock level of the kth production material; Q j represents the quality pass rate of the jth product, Q t Indicates the preset quality pass rate target; when: Trigger the genetic algorithm model to adjust the production scheduling problem.

[0123] The trigger mechanism for calculating energy consumption using the linear regression model is:

[0124] Let Δp denote the difference in electricity prices, Δp th represents the electricity price trigger threshold; Δl represents the load difference, Δl th represents the trigger load difference threshold; t represents temperature data, μ t and σ t Indicates the mean and standard deviation of temperature data in the historical period, when:

[0125] Δp≥Δp th ∨Δl≥Δl th ∨[(t<μ t -2σ t )∨(t>μt +2σ t )]

[0126] That is, when the electricity price difference is greater than or equal to the electricity price trigger threshold, the load difference is greater than or equal to the load difference threshold, and the mean and standard deviation of the temperature data in the historical time period constitute a variation range, the linear regression model is triggered to calculate the energy consumption problem.

[0127] Step 4: Through the industrial problem processing model scheduling mechanism, when the collected industrial data changes, the corresponding industrial problem processing model is triggered, and the industrial data is transmitted to the corresponding model. The corresponding industrial problem processing model performs calculations in the model based on the received data and outputs the data results of the corresponding industrial problem.

[0128] Example 2

[0129] Reference Figure 2 ,The second embodiment of the present invention provides an industrial algorithm model ,scheduling system based on industrial big data.

[0130] The system includes: a data acquisition and processing module, an industrial model module, an industrial model scheduling module and a data output module;

[0131] The data acquisition and processing module is used to acquire the original industrial data collected by the factory sensors and pre-process the original industrial data;

[0132] The industrial model module is used to define different types of industrial problems, including equipment failure prediction, production scheduling and energy consumption. For different industrial problems, corresponding industrial problem processing models are constructed, including a support vector machine (SVM) model for predicting equipment failure problems, a genetic algorithm model for adjusting production scheduling problems, and a multivariate linear regression model for calculating energy consumption problems.

[0133] The industrial model scheduling module is used to establish an industrial problem processing model scheduling mechanism. When the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data is input into the corresponding industrial problem processing model;

[0134] The data output module and the corresponding industrial problem processing model perform calculations in the model based on the received data and output data results corresponding to the industrial problem.

[0135] Example 3

[0136] Figure 3 FIG. 1 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Figure 3As shown, according to another aspect of the present invention, an electronic device 500 is also provided. The electronic device 500 may include one or more processors and one or more memories. The memory stores a computer-readable code, and when the computer-readable code is executed by one or more processors, the industrial algorithm model scheduling method based on industrial big data as described above can be executed.

[0137] The method or system according to the embodiment of the present invention can also be used by Figure 3 The architecture of the electronic device shown is implemented.

[0138] like Figure 3 As shown, the electronic device 500 may include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, and the like.

[0139] The storage device in the electronic device 500, such as ROM 503 or hard disk 507, can store the industrial algorithm model scheduling method based on industrial big data provided by the present invention.

[0140] An industrial algorithm model scheduling method based on industrial big data includes: obtaining original industrial data collected by factory sensors and preprocessing the original industrial data; defining different types of industrial problems, the industrial problems include equipment failure prediction, production scheduling and energy consumption industrial problems, and building corresponding industrial problem processing models for different industrial problems, building a support vector machine (SVM) model to predict equipment failure problems, building a genetic algorithm model to adjust production scheduling problems, and building a multivariate linear regression model to calculate energy consumption problems; establishing an industrial problem processing model scheduling mechanism, when the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data is input into the corresponding industrial problem processing model; the corresponding industrial problem processing model performs calculations in the model based on the received data, and outputs the data results of the corresponding industrial problem.

[0141] Furthermore, the electronic device 500 may also include a user interface 508. Figure 3 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different devices. Figure 3 One or more components of an electronic device are shown.

[0142] Example 4

[0143] Figure 4 It is a schematic diagram of the structure of a computer-readable storage medium provided by an embodiment of the present invention.

[0144] like Figure 4As shown, a computer readable storage medium 600 according to one embodiment of the present invention.

[0145] Computer readable storage medium 600 has computer readable instructions stored thereon.

[0146] When the computer-readable instructions are executed by the processor, the industrial algorithm model scheduling method based on industrial big data according to the embodiment of the present invention described with reference to the above drawings can be executed.

[0147] The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. In addition, according to an embodiment of the present invention, the process described above with reference to the flowchart may be implemented as a computer software program.

[0148] For example, the present invention provides a non-temporary machine-readable storage medium, which stores machine-readable instructions, and the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided by the present invention, for example: obtaining raw industrial data collected by factory sensors and preprocessing the raw industrial data; defining different types of industrial problems, the industrial problems include equipment failure prediction, production scheduling, and energy consumption industrial problems, and constructing corresponding industrial problem processing models for different industrial problems, constructing a support vector machine SVM model to predict equipment failure problems, constructing a genetic algorithm model to adjust production scheduling problems, and constructing a multivariate linear regression model to calculate energy consumption problems; establishing an industrial problem processing model scheduling mechanism, when the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data is input into the corresponding industrial problem processing model; the corresponding industrial problem processing model performs calculations in the model based on the received data, and outputs the data results of the corresponding industrial problem.

[0149] When the computer program is executed by the central processing unit (CPU), the above functions defined in the method of the present invention are performed. The method, apparatus, and device of the present invention may be implemented in many ways. For example, the method, apparatus, and device of the present invention may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware.

[0150] The above sequence for the steps of the method is for illustration only, and the steps of the method of the present invention are not limited to the sequence specifically described above unless otherwise specifically stated.

[0151] In addition, in some embodiments, the present invention can also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present invention. Therefore, the present invention also covers a recording medium storing a program for executing the method according to the present invention.

[0152] In addition, the parts of the above technical solutions provided in the embodiments of the present invention that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0153] The specific implementation modes as described above further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation mode of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. The industrial algorithm model scheduling method based on industrial big data is characterized by: include: Obtain the original industrial data collected by factory sensors and pre-process the original industrial data; The original industrial data includes industrial equipment data, product production data and energy data during industrial production; the industrial equipment data includes temperature data, vibration data and current data of industrial equipment; the product production data includes production material consumption data and product output data; the energy data includes energy price data and energy consumption data; Preprocess the original industrial data, remove missing values, outliers and duplicate values, and normalize the original industrial data; Define different types of industrial problems, including equipment failure prediction, production scheduling and energy consumption problems. Construct corresponding industrial problem processing models for different industrial problems, construct a support vector machine (SVM) model to predict equipment failure problems, construct a genetic algorithm model to adjust production scheduling problems, and construct a multivariate linear regression model to calculate energy consumption problems. Based on the temperature data, vibration data and current data of industrial equipment, a support vector machine (SVM) model is constructed to predict equipment failure problems. Collect historical operating data of industrial equipment, including temperature data, vibration data, and current data, as well as corresponding timestamps; Collect historical fault data of industrial equipment, the historical fault data including the time when the fault occurred and the type of fault, and digitally encode the data; Label the historical fault data, align the standardized time series data with the fault events, generate fault labels, use -1 to indicate that the industrial equipment is normal, and use +1 to indicate that the industrial equipment is faulty. Randomly divide the data set into training set and test set, and use the training set and test set to train and test the support vector machine (SVM) model. The training process of the support vector machine (SVM) model to predict equipment failure is as follows: The temperature data T' of industrial equipment after preprocessing and normalization q , vibration data V' q and current data I' q Perform feature extraction and generate feature vector x q , x q =[T' q ,V' q ,I' q ],q=1,2,...,n, n is the eigenvector x q The number of data; Given a training set (x1,y1),(x2,y2),...,(x n ,y n ), where y q represents the fault label corresponding to the qth feature vector, y q ∈{-1,+1}; The optimization problem of predicting equipment failure by support vector machine SVM model is expressed as: Among them, α q is the Lagrange multiplier, C is the penalty coefficient, which is used to control the load of the model and the tolerance of misclassification. represents the transpose of the eigenvector; solving for the optimal Construct the decision function of the support vector machine SVM model: Where x represents the input vector; sign(·) is a sign function; using the training set data, by solving the support vector machine SVM model optimization problem, we get and b * Value, build support vector machine SVM model, and predict the future working status of the equipment through the collected industrial equipment data: in, represents the predicted fault label, K(x q ,x) is the kernel function, which is used to map features to high-dimensional space; if It is predicted that the industrial equipment will be normal in the future period. It predicts industrial equipment failures in the future time period and requires industrial equipment maintenance in advance; Establish an industrial problem processing model scheduling mechanism. When the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data will be input into the corresponding industrial problem processing model; The corresponding industrial problem processing model performs calculations in the model based on the received data and outputs the data results of the corresponding industrial problem.

2. The industrial algorithm model scheduling method based on industrial big data according to claim 1 is characterized in that: Based on production material consumption data and product output data, a genetic algorithm model is constructed to adjust the production scheduling problem; Collecting production material consumption data and product output data during the production process, wherein the production material consumption data includes the consumption, consumption time and inventory of raw materials; the product output data includes the production quantity, production time and quality qualification rate of the products; minimizing production costs and maximizing product quality are the production scheduling goals; Define the production decision variable, A ij represents the quantity of the jth product produced in the i-th time period; B ik represents the amount of the kth material consumed in the i-th time period; Z j represents the qualified rate of the jth product; Define the production objective function, the objective function to minimize production cost is: min(∑ i ∑ k C k ×B ik ),k=1,2,...,K, where C k represents the unit cost of the kth material; the objective function for maximizing product quality is: max(∑ j Z j ×∑ i A ij ),i=1,2,...,I;j=1,2,...,J, the maximum product quality is the weighted sum of product qualification rate and output; Define the constraints, the material consumption constraint is: ∑ j (q jk ×A ij )≤B ik , where q jk Indicates the quantity of the kth material required to produce a jth product; the material inventory constraint is: Σ i (B ik )≤S k , where S k represents the current inventory of the kth material; production capacity constraint: ∑ j (t j ×A ij )≤T i , where t j represents the time required to produce a j-th product, T i represents the production time in the i-th time period; A ij ≥0, B ik ≥0, Z j ∈[0,1]; The steps to construct the genetic algorithm are as follows: Coding method: Real number coding is used. Each chromosome consists of (|I|×|J|+|I|×|K|+|J|) genes, where I, J and K correspond to the maximum values ​​of i, j and k. Randomly generate N feasible solutions that meet the constraints as the initial population; Normalize the objective function of minimizing production cost and the objective function of maximizing product quality, and transform the two objective functions into a single objective fitness function by weighted sum; A linear ranking selection operator based on sorting is used to assign the operator a probability of being selected according to the fitness ranking of the individual; The simulated binary crossover operator is used to generate two offspring individuals based on two parent individuals, and the polynomial mutation operator is used to perturb each gene with a certain probability; The maximum evolutionary number is set as the termination condition, and the genetic algorithm calculation is stopped when the maximum evolutionary number is reached.

3. The industrial algorithm model scheduling method based on industrial big data according to claim 2 is characterized in that: Based on the temperature data, energy price data and energy consumption data of industrial equipment, a multivariate linear regression model is constructed to calculate the energy consumption problem; the energy is electrical energy; Collect electricity price data within the historical time period, including peak electricity prices and valley electricity prices, collect electricity consumption data within the historical time period, including peak electricity consumption and valley electricity consumption, and collect average temperature data within the historical time period as weather factors that affect electricity consumption; The peak-valley electricity price difference is calculated as the electricity price feature, the peak-valley electricity consumption difference is calculated as the load feature, and the temperature data is standardized, that is, the mean is subtracted and divided by the standard deviation to obtain the standardized temperature feature; the electricity price difference, load difference and standardized temperature are selected as the input features of the regression model; Construct a multiple linear regression model, and the regression equation is: in, is the electricity consumption in the predicted future time period, m1 is the electricity price difference, m2 is the load difference, and m3 is the standardized temperature. The regression coefficients β1, β2, β3, the intercept term β0 and the error term ∈ of the regression equation are obtained by solving the data in the historical time period.

4. The industrial algorithm model scheduling method based on industrial big data according to claim 3 is characterized in that: Establish an industrial problem processing model scheduling mechanism. When the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data is input into the corresponding industrial problem processing model. Specifically, it includes: The trigger mechanism of the support vector machine (SVM) model predicts equipment failure problems as follows: Let T represent the temperature data of industrial equipment, T min and T max Respectively represent the lower and upper limits of the normal temperature range; V represents the vibration data of industrial equipment obtained, V th represents the set vibration change threshold; I represents the acquired industrial equipment current data, I th Indicates the set current change threshold when: Trigger the support vector machine (SVM) model to predict equipment failure problems; Genetic algorithm model adjusts the trigger mechanism of production scheduling problem: Order I k represents the inventory of the kth production material, THS k represents the safety stock level of the kth production material; Q j represents the quality pass rate of the jth product, Q t Indicates the preset quality pass rate target; when: Trigger the genetic algorithm model; The trigger mechanism for calculating energy consumption using the linear regression model is: Let Δp represent the difference in electricity prices, Δp th represents the electricity price trigger threshold; Δl represents the load difference, Δl th represents the trigger load difference threshold; t represents temperature data, μ t and σ t Indicates the mean and standard deviation of temperature data in the historical period, when: Δp≥Δp th ∨Δl≥Δl th ∨[(t<μ t -2s t )∨(t>μ t +2s t )] Trigger the linear regression model to calculate energy consumption problem.

5. The industrial algorithm model scheduling method based on industrial big data according to claim 4 is characterized in that: Through the industrial problem processing model scheduling mechanism, when the collected industrial data changes, the corresponding industrial problem processing model is triggered, and the industrial data is transmitted to the corresponding model. The corresponding industrial problem processing model performs calculations in the model based on the received data and outputs the data results of the corresponding industrial problem.

6. An industrial algorithm model scheduling system based on industrial big data, which is implemented based on the industrial algorithm model scheduling method based on industrial big data according to any one of claims 1 to 5, characterized in that: It includes: data acquisition and processing module, industrial model module, industrial model scheduling module and data output module; The data acquisition and processing module is used to acquire the original industrial data collected by the factory sensors and pre-process the original industrial data; The industrial model module is used to define different types of industrial problems, including equipment failure prediction, production scheduling and energy consumption. For different industrial problems, corresponding industrial problem processing models are constructed, including a support vector machine (SVM) model for predicting equipment failure problems, a genetic algorithm model for adjusting production scheduling problems, and a multivariate linear regression model for calculating energy consumption problems. The industrial model scheduling module is used to establish an industrial problem processing model scheduling mechanism. When the data collected by industrial equipment and factory sensors triggers the industrial problem processing model scheduling mechanism, the collected data is input into the corresponding industrial problem processing model; The data output module and the corresponding industrial problem processing model perform calculations in the model based on the received data and output data results corresponding to the industrial problem.

7. An electronic device, characterized in that: include: A processor and a memory, wherein the memory stores a computer program that can be called by the processor; the processor executes the industrial algorithm model scheduling method based on industrial big data as described in any one of claims 1 to 5 by calling the computer program stored in the memory.

8. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are executed on a computer, the computer executes the industrial algorithm model scheduling method based on industrial big data as described in any one of claims 1 to 5.

Images