Gas volume prediction method, device, equipment and storage medium

Abstract

The present invention discloses a gas volume prediction method, device, equipment and storage medium. The method comprises: obtaining the attribute information of the product to be printed and the offset printing speed information; inputting the attribute information of the product to be printed and the offset printing speed information into the gas volume prediction model to obtain the gas volume prediction result, so as to supply gas according to the gas volume prediction result; wherein the gas volume prediction model is trained based on the sample data acquired in advance, and the sample data is determined based on the detection result of the product image collected during the offset printing process. The technical scheme accurately predicts the gas volume required in the offset printing online detection process through the gas volume prediction model, which can solve the problem that a single gas volume is difficult to meet the gas volume requirements of different products in the offset printing online detection process, thereby achieving the purpose of saving energy and reducing production costs while improving the adaptability of gas supply.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a gas volume prediction method, device, equipment and storage medium. Background Art

[0002] With the continuous advancement of printing inspection technology, quality inspection has penetrated from the single quality inspection stage to all printing processes. The introduction of offset printing online inspection not only saves a lot of paper for users, but also helps to improve the quality of printed products.

[0003] At present, in the offset printing online inspection solution, the product to be printed is running at high speed with a large drum. In order to make the camera capture clearer images, a larger air volume must be used to blow the front end of the product. Printing factories usually use a single larger air volume to meet the air volume requirements of various products to be printed.

[0004] However, due to the different properties of the products to be printed, such as weight and width, a single gas volume cannot adaptably meet the gas volume requirements of different products. It will also lead to a substantial increase in electricity consumption, resulting in high electricity bills, and thus increasing production costs. Summary of the invention

[0005] The present invention provides a gas volume prediction method, device, equipment and storage medium to solve the problem that a single gas volume is difficult to meet the gas volume requirements of different products during offset printing online detection. It can improve the adaptability of gas volume supply while achieving the purpose of saving energy and reducing production costs.

[0006] According to one aspect of the present invention, a gas volume prediction method is provided, the method comprising:

[0007] Obtaining property information of products to be printed and offset printing speed information;

[0008] The attribute information of the product to be printed and the offset printing speed information are input into the gas volume prediction model to obtain the gas volume prediction result, so as to supply gas according to the gas volume prediction result; wherein, the gas volume prediction model is trained based on pre-acquired sample data, and the sample data is determined based on the detection results of the product images collected during the offset printing process.

[0009] According to another aspect of the present invention, there is provided a gas volume prediction device, the device comprising:

[0010] An information acquisition module, used to acquire property information of the product to be printed and offset printing speed information;

[0011] The gas volume prediction result determination module is used to input the attribute information of the product to be printed and the offset printing speed information into the gas volume prediction model to obtain the gas volume prediction result, so as to supply gas according to the gas volume prediction result; wherein, the gas volume prediction model is trained based on pre-acquired sample data, and the sample data is determined based on the detection results of the product images collected during the offset printing process.

[0012] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0013] at least one processor; and

[0014] a memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the gas volume prediction method described in any embodiment of the present invention.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the gas volume prediction method described in any embodiment of the present invention when executed.

[0017] The technical solution of the embodiment of the present invention can accurately predict the gas volume required in the online offset printing detection process through a gas volume prediction model, thereby solving the problem that a single gas volume is difficult to meet the gas volume requirements of different products in the online offset printing detection process, thereby improving the adaptability of the gas volume supply while achieving the purpose of saving energy and reducing production costs.

[0018] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 is a flow chart of a gas volume prediction method provided according to Embodiment 1 of the present invention;

[0021] Figure 2is a flow chart of a method for training a gas volume prediction model provided according to Embodiment 2 of the present invention;

[0022] Figure 3A This is a flow chart of a gas volume prediction method provided according to a specific applicable scenario 1 of the present invention;

[0023] Figure 3B It is a structural schematic diagram of a gas volume prediction model provided according to a specific applicable scenario 1 of the present invention;

[0024] Figure 3C This is a schematic diagram of a genetic encoding method for adjusting parameters provided according to a specific applicable scenario 1 of the present invention;

[0025] Figure 4 is a structural schematic diagram of a gas volume prediction device provided according to Embodiment 3 of the present invention;

[0026] Figure 5 It is a structural schematic diagram of an electronic device for implementing the gas volume prediction method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. The acquisition, storage, use, processing, etc. of data in the technical solution of this application comply with the relevant provisions of national laws and regulations.

[0029] Embodiment 1

[0030] Figure 1A flow chart of a gas volume prediction method is provided for the first embodiment of the present invention. This embodiment is applicable to gas volume prediction scenarios, especially to gas volume prediction in offset printing online detection scenarios. The method can be executed by a gas volume prediction device, which can be implemented in the form of hardware and / or software, and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0031] S110, obtaining property information of the product to be printed and offset printing speed information.

[0032] This solution can be executed by an offset printing online detection system, which can be configured on an offset printing device to monitor the quality of printed products, the operating status of the offset printing device, etc. in real time. The offset printing online detection system can read the attribute information of the product to be printed that is pre-entered or stored in the system, such as paper weight, paper length, and paper width. The offset printing online detection system can communicate with the offset printing device to obtain the attribute information of the product to be printed in real time. The offset printing online detection system can continuously collect offset printing speed information during the operation of the offset printing device through speed collection devices such as encoders.

[0033] S120. Input the attribute information of the product to be printed and the offset printing speed information into the gas volume prediction model to obtain the gas volume prediction result, so as to supply gas according to the gas volume prediction result; wherein, the gas volume prediction model is trained based on pre-acquired sample data, and the sample data is determined based on the detection results of the product images collected during the offset printing process.

[0034] The offset printing online detection system can input the attribute information of the product to be printed and the offset printing speed information as input data of the gas volume prediction model into the gas volume prediction model. It is easy to understand that the offset printing gas consumption is correlated with the attributes of the product to be printed. At the same time, the offset printing gas consumption is also affected by the operating speed of the offset printing equipment. Therefore, there is a mapping relationship between the attribute information of the product to be printed, the offset printing speed information and the offset printing gas consumption, and the gas volume prediction model can be used to describe this mapping relationship. The gas volume prediction model can output the gas volume prediction result based on the attribute information of the product to be printed and the offset printing speed information. The offset printing online detection system can supply gas during the offset printing process through the gas supply equipment according to the gas volume prediction result to ensure that the offset printing online detection obtains a clear product image.

[0035] Among them, the gas volume prediction model can be trained based on pre-acquired sample data, for example, it can be a deep learning model such as a neural network. In order to better meet the needs of actual application scenarios, the offset printing online detection system can extract the printing product attribute information, offset printing speed information and offset printing gas consumption in the historical records as sample data. The printing product attribute information and offset printing speed information are used as independent variables, and the offset printing gas consumption is used as the dependent variable. A mapping relationship between the independent variable and the dependent variable is constructed, so that after inputting the attribute information of the product to be printed and the offset printing speed information, the gas volume prediction result is obtained according to the mapping relationship. It should be noted that the selection of the sample data has certain standards. Specifically, the offset printing online detection system can perform blowing quality detection based on the product images collected during the offset printing process, and the sample data can be selected based on the blowing quality detection results.

[0036] This technical solution accurately predicts the gas volume required during the online offset printing detection process through a gas volume prediction model, which can solve the problem that a single gas volume is difficult to meet the gas volume requirements of different products during the online offset printing detection process, thereby improving the adaptability of the gas volume supply while achieving the purpose of saving energy and reducing production costs.

[0037] Embodiment 2

[0038] Figure 2 1 is a flow chart of a method for training a gas volume prediction model according to Embodiment 2 of the present invention. This embodiment is refined based on the above embodiment. Figure 2 As shown, the training method of the gas volume prediction model may include:

[0039] S210, collecting product images during the offset printing process according to a preset sampling frequency, and synchronously obtaining gas flow information.

[0040] In this solution, the gas volume prediction model is constructed based on a radial basis function (RBF) neural network. The offset printing online detection system may include an image acquisition device, such as a camera. During the operation of the offset printing equipment, the offset printing online detection system may collect product images in the offset printing process at a certain sampling frequency. The offset printing online detection system may also include a position information acquisition device, such as a photoelectric sensor, for obtaining the position information of the product in the offset printing process, so as to collect product images within a reasonable shooting range. The offset printing online detection system may also include a gas flow acquisition device, such as a flow meter. While collecting product images, the offset printing online detection system can synchronously obtain gas flow information through the flow acquisition device.

[0041] S220. Determine at least one set of sample data according to the detection result of the product image; wherein the sample data includes product attribute information associated with the product image, pre-acquired offset printing speed information, and the gas flow information.

[0042] The offset printing online detection system can use the product attribute information associated with the product image, the offset printing speed information of the offset printing equipment, and the gas flow information obtained synchronously with the product image as a set of sample data, and build a sample database by collecting multiple sets of sample data, so as to use the sample data in the sample database to train the gas volume prediction model.

[0043] It should be noted that when training the gas volume prediction model, the offset printing online detection system can use a large amount of sample data, such as hundreds or thousands of groups of sample data, for training to increase the adaptability of the model. When the gas volume prediction model has structural advantages, such as a deep learning structure based on single sample learning or few sample learning, the offset printing online detection system can also use a small amount of sample data, such as a few or dozens of groups of sample data, or even one group of sample data, for training.

[0044] During the sample collection process, the gas volume provided by the gas supply equipment during the offset printing process may not meet the requirements of the offset printing online detection system when collecting product images. In order to ensure the reliability of the sample data, a gas volume prediction model with good training effect is obtained. The offset printing online detection system can select the blowing gas volume information that makes the product flattening condition good as the sample data based on the detection results of the product image. Specifically, the offset printing online detection system can perform target detection on the product image. The detected target can be whether there are areas such as product warping and curling in the product image. The offset printing online detection system can also determine whether the gas flow is appropriate by detecting whether there are changes in the appearance of two adjacent product images, such as detecting whether the paper length, paper width, etc. in two images of the same product are consistent.

[0045] In this solution, optionally, the product image is a three-dimensional image;

[0046] Accordingly, the determining of at least one set of sample data according to the detection result of the product image includes:

[0047] According to the three-dimensional image, detecting whether there is an abnormal flattening area of ​​the product;

[0048] If there is no abnormal flattening area, the product attribute information associated with the three-dimensional image, the pre-acquired offset printing speed information, and the gas flow information acquired synchronously with the three-dimensional image are taken as a set of sample data.

[0049] The offset printing online detection system can take multi-angle photos of products in the offset printing process to obtain a three-dimensional image of the product. Based on the three-dimensional image, the offset printing online detection system can more easily determine whether there are abnormal flattening areas such as depressions and bulges in the product during the offset printing process. If the three-dimensional image detects that the product is well flattened and there are no abnormal product areas, the offset printing online detection system can use the product attribute information associated with the three-dimensional image, the pre-acquired offset printing speed information, and the gas flow information acquired synchronously with the three-dimensional image as a set of sample data for training the gas volume prediction model. Among them, the product attribute information and the offset printing speed information can be used as input data of the gas volume prediction model, and the gas flow information can be used as label data of the gas volume prediction model.

[0050] The above scheme can quickly locate the abnormal area of ​​product flattening through three-dimensional images, improve the reliability of sample data selection, and is conducive to achieving good model training effects.

[0051] S230: Using the sample data to train a radial basis function neural network to obtain a gas volume prediction model.

[0052] The offset printing online detection system can input product attribute information and offset printing speed information into the radial basis neural network. Among them, the structure of the radial basis neural network is similar to that of a multi-layer forward network, and can be a three-layer forward network. The input layer is composed of signal source nodes, for example, product attribute information and offset printing speed information can be used as two signal source nodes of the input layer. The second layer is the hidden layer, and the number of hidden units can be determined according to the specific application needs. The transformation function RBF of the hidden unit is a non-negative nonlinear function that is radially symmetric and attenuated about the center point. The third layer is the output layer, which can respond to the effect of the input data. The transformation from the input space to the hidden layer space is nonlinear, while the transformation from the output layer space of the hidden layer space is linear. The radial basis neural network can extract features from the input data, output gas volume prediction features, and optimize the internal weight parameters of the radial basis neural network by narrowing the gap between the gas flow information and the gas volume prediction features to obtain a gas volume prediction model.

[0053] On the basis of the above solution, optionally, the use of the sample data to train a radial basis function neural network to obtain a gas volume prediction model includes:

[0054] Inputting the product attribute information and the offset printing speed information into a radial basis function neural network as model input data to obtain gas volume prediction features;

[0055] According to the gas volume prediction characteristics and the gas flow information, the radial basis function neural network is optimized until it meets the preset conditions to obtain a gas volume prediction model.

[0056] It can be understood that the gas volume prediction feature can be the output data of the radial basis neural network during the training process, and the offset printing online detection system can encode the gas flow information into the same type of data format as the gas volume prediction feature. The radial basis neural network can use the gas flow information as the ultimate goal of training, and calculate the difference data between the gas volume prediction feature and the gas flow information through the activation function. When the difference data does not meet the preset conditions, the difference data is fed back to the hidden layer of the radial basis neural network to continuously optimize the weight parameters until the difference data meets the preset conditions, and then the gas volume prediction model is output. Among them, the activation function can be a Gaussian function, an inverted S-type function, a quasi-multi-quadratic function, etc. The preset condition can be that the number of training iterations reaches the preset number of iterations, or it can be determined by whether the loss value, accuracy, etc. during the radial basis neural network training process meet the preset evaluation indicators.

[0057] This scheme obtains a gas volume prediction model by training a radial basis function neural network, thereby achieving reliable prediction of gas volume and improving the robustness of the gas volume prediction model.

[0058] In a feasible solution, optionally, the gas volume prediction model includes adjustment parameters;

[0059] After obtaining the gas volume prediction model, the method further includes:

[0060] Optimizing the adjustment parameters by genetic algorithm until the optimization termination condition is met, and outputting the optimization result;

[0061] According to the optimization result, the adjustment parameters of the gas volume prediction model are updated.

[0062] The offset printing online detection system can use a genetic algorithm to optimize the adjustment parameters in the gas volume prediction model to obtain a more accurate gas volume prediction result. The adjustment parameters may include a center vector, a basis width vector, and a weight vector in a radial basis neural network. Through the genetic algorithm, the offset printing online detection system can optimize one type of adjustment parameters separately, or optimize multiple types at the same time. The optimization termination condition may be that the number of iterations of the genetic algorithm reaches a preset number of iterations.

[0063] The above scheme can further optimize the adjustment parameters of the gas volume prediction model, which is conducive to achieving high-precision prediction of offset printing gas consumption.

[0064] On the basis of the above solution, optionally, optimizing the adjustment parameters by a genetic algorithm until the optimization termination condition is met and outputting the optimization result includes:

[0065] Genetically encoding the adjustment parameters to generate an input population;

[0066] Performing selection, crossover and mutation operations on the input population to obtain an output population;

[0067] Calculating the fitness of individuals in the output population, and determining the input population for the next iteration according to the fitness calculation result;

[0068] If the number of iterations reaches the preset number of iterations, the iterations are terminated, and the optimization result of the adjustment parameters is determined according to the output population finally obtained.

[0069] Specifically, the offset printing online detection system can genetically encode the adjustment parameters to be optimized to obtain the initial population, that is, the input population of the first iteration. The initial population is selected, crossed and mutated to achieve population expansion and obtain the output population. The fitness of the individuals in the output population is calculated, and the fitness calculation results are compared with the pre-set selection conditions, and the individuals in the output population are selected to enter the input population of the next iteration. Until the number of iterations reaches the preset number of iterations, the iteration is terminated and the optimization result of the adjustment parameters is output.

[0070] This solution simulates the evolution process of the population and outputs the optimization results of the adjustment parameters, which is conducive to obtaining adjustment parameters that are more suitable for the application scenario and improving the prediction accuracy of the gas volume prediction model.

[0071] In a preferred solution, the adjustment parameters include a center vector and a base width vector;

[0072] The step of genetically encoding the adjustment parameters to generate an input population comprises:

[0073] The individuals are genetically encoded in an alternating manner between the elements in the center vector and the elements in the base width vector to generate the input population.

[0074] In this scheme, the offset printing online detection system can optimize the two types of adjustment parameters, the center vector and the base width vector, through the genetic algorithm. In the genetic encoding process, the frequency and probability of the exchange of center data and width data can be increased by encoding each element in the center vector and each element in the base width vector in an alternating manner, which is conducive to increasing the diversity of the population and thus improving the overall search ability of the genetic algorithm.

[0075] This technical solution accurately predicts the gas volume required during the online offset printing detection process through a gas volume prediction model, which can solve the problem that a single gas volume is difficult to meet the gas volume requirements of different products during the online offset printing detection process, thereby improving the adaptability of the gas volume supply while achieving the purpose of saving energy and reducing production costs.

[0076] Specific application scenario 1

[0077] Figure 3AThis is a flow chart of a gas volume prediction method provided according to a specific application scenario 1 of the present invention. This embodiment is a specific embodiment based on the above embodiment. Figure 3A As shown, the method includes:

[0078] Step 1: Sample collection, input the specifications of the product to be printed into the data collection system; when the offset printing machine is running, the encoder collects the offset printing speed in real time and enters the speed value into the data collection system; when the offset printing machine is in normal production, the flow meter collects the gas consumption in real time and transmits it to the data collection system through RS485 communication. The data collection system collects the specifications of the product to be printed, the offset printing speed and the gas consumption into the sample library.

[0079] Step 2: Data verification, filter the data in the sample library and delete abnormal data; when the offset printing machine is in normal production, take pictures of the products under the gas consumption counted by the flow meter. If the blowing and flattening effect is normal, the gas flow counted at this time is true and retained as sample data; if the blowing and flattening effect is abnormal, the gas flow statistics are false and cannot be used as sample data.

[0080] Step 3: Eliminate invalid sample data from the sample library and divide it into a training library and a test library. Use the sample library to train the RBF network and obtain a preliminary model.

[0081] Step 4: Use Genetic Algorithm (GA) to optimize the parameters of the RBF network to obtain the optimized GA-RBF network model.

[0082] Step 5: Use the optimized GA-RBF network model to predict the gas consumption of the offset printing machine and compare it with the output gas consumption of the test library (i.e., the actual gas consumption). If the error is less than the preset error threshold, the optimization termination condition is met.

[0083] The specific application of the GA-RBF network model is as follows:

[0084] 1. RBF Neural Network

[0085] Figure 3B It is a structural diagram of a gas volume prediction model provided according to the specific applicable scenario 1 of the present invention. In this solution, the RBF neural network selects a Gaussian function as a hidden layer kernel function.

[0086] The output of the hidden layer is a nonlinear activation function h i constitute:

[0087]

[0088] Among them, X represents the input sample vector, C i represents the center vector; b iis the width of the basis function; m is the number of nodes in the hidden layer.

[0089] The weight values ​​between the hidden layer and the output layer in the RBF network and the output threshold value are constantly adjusted under the supervision of supervised learning rules. The key to the RBF network lies in the determination of the center vector, base width vector and weight vector.

[0090] Since the input of the RBF network is a four-dimensional array and the output is a one-dimensional value, there is a nonlinear relationship between the input and output. The nonlinear approximation function of the RBF network is suitable for solving the nonlinear relationship that exists in theory.

[0091] 2. Genetic algorithm optimization of RBF network

[0092] When designing an RBF network, determine the data center vector c k , base width vector b k and the weight vector w k The specific values ​​of these three parameters are the key to improving the speed and accuracy of RBF network problem solving. Genetic algorithm optimizes network parameters, including the following aspects:

[0093] (1) Coding and population initialization

[0094] Figure 3C This is a schematic diagram of a genetic encoding method for adjusting parameters provided according to the specific application scenario 1 of the present invention. The optimization of the traditional RBF by selecting the real number encoding method in this solution is mainly reflected in the following two points:

[0095] First point: Traditional RBF network design is completed in steps. The first step is to use the k-means clustering method to select the data center and base width constant of the network. The second step is to use the least squares method to calculate the edge weights and then obtain the network output. This step-by-step calculation of the solution may not necessarily synchronize the optimal solution, which will affect the superiority of the algorithm. The genetic algorithm puts the encoding of the three types of parameters into a chromosome encoding string to achieve the synchronization of the optimal solution of the parameters.

[0096] Second point: Figure 3C As shown in the figure, the center of the data and the width of the basis function in the RBF neural network affect each other. In order to be closer to the radial basis function itself, the chromosome c k , and b k Alternate, and then arrange w at the end of the chromosome string k The above arrangement method is used to encode individuals. It is beneficial to increase the frequency and probability of center and width exchange, which is crucial to the diversity of the population and can improve the overall search ability of the network.

[0097] (2) Parameter optimization update

[0098] Assume the output of the RBF network is:

[0099]

[0100] Among them, X represents the input sample vector, C i represents the center vector, w i Represents the output weights of the hidden layer nodes.

[0101] The energy function of the RBF network is:

[0102]

[0103] Among them, X p represents the input sample vector; N represents the number of training samples, d p Indicates output expectations.

[0104] C i and b i The update is:

[0105]

[0106]

[0107] Where t represents the number of iterations, λ is the learning efficiency of the basis function center value, and β is the learning efficiency of the basis function width value.

[0108] This technical solution accurately predicts the gas volume required during the online offset printing detection process through a gas volume prediction model, which can solve the problem that a single gas volume is difficult to meet the gas volume requirements of different products during the online offset printing detection process, thereby improving the adaptability of the gas volume supply while achieving the purpose of saving energy and reducing production costs.

[0109] Embodiment 3

[0110] Figure 4 This is a schematic diagram of the structure of a gas volume prediction device provided in Embodiment 3 of the present invention. Figure 4 As shown, the device comprises:

[0111] The information acquisition module 410 is used to acquire the property information of the product to be printed and the offset printing speed information;

[0112] The gas volume prediction result determination module 420 is used to input the attribute information of the product to be printed and the offset printing speed information into the gas volume prediction model to obtain the gas volume prediction result, so as to supply gas according to the gas volume prediction result; wherein, the gas volume prediction model is trained based on pre-acquired sample data, and the sample data is determined based on the detection results of the product images collected during the offset printing process.

[0113] In this solution, optionally, the device further includes:

[0114] An image and flow acquisition module is used to collect product images during the offset printing process according to a preset sampling frequency and simultaneously obtain gas flow information;

[0115] A sample data determination module, configured to determine at least one set of sample data according to the detection result of the product image; wherein the sample data includes product attribute information associated with the product image, pre-acquired offset printing speed information, and the gas flow information;

[0116] The gas volume prediction model generation module is used to train a radial basis neural network using the sample data to obtain a gas volume prediction model.

[0117] In a feasible solution, optionally, the product image is a three-dimensional image;

[0118] Accordingly, the sample data determination module includes:

[0119] an abnormality detection unit, used for detecting whether there is an abnormal flattening area of ​​the product according to the three-dimensional image;

[0120] The sample data determination unit is used to take the product attribute information associated with the three-dimensional image, the pre-acquired offset printing speed information and the gas flow information acquired synchronously with the three-dimensional image as a set of sample data if there is no abnormal flattening area.

[0121] Based on the above solution, optionally, the gas volume prediction model generation module includes:

[0122] A gas volume prediction feature generation unit, used for inputting the product attribute information and the offset printing speed information as model input data into a radial basis function neural network to obtain a gas volume prediction feature;

[0123] The gas volume prediction model generating unit is used to optimize the radial basis function neural network according to the gas volume prediction characteristics and the gas flow information until the preset conditions are met to obtain the gas volume prediction model.

[0124] In another feasible solution, the gas volume prediction model includes adjustment parameters;

[0125] The device also includes:

[0126] An optimization result output module is used to optimize the adjustment parameters by genetic algorithm until the optimization termination condition is met and output the optimization result;

[0127] The adjustment parameter updating module is used to update the adjustment parameters of the gas volume prediction model according to the optimization result.

[0128] Based on the above solution, optionally, the optimization result output module includes:

[0129] An input population generation unit, used for genetically encoding the adjustment parameters to generate an input population;

[0130] An output population generating unit, used for performing selection operation, crossover operation and mutation operation on the input population to obtain an output population;

[0131] A fitness calculation unit, used to calculate the fitness of individuals in the output population, and determine the input population for the next iteration according to the fitness calculation result;

[0132] The optimization result output unit is used to terminate the iteration if the number of iterations reaches a preset number of iterations, and determine the optimization result of the adjustment parameter according to the output population finally obtained.

[0133] In a preferred solution, the adjustment parameters include a center vector and a base width vector;

[0134] The input population generation unit is specifically used to perform gene encoding on individuals in a manner of alternating each element in the center vector and each element in the base width vector to generate an input population.

[0135] The gas volume prediction device provided in the embodiment of the present invention can execute the gas volume prediction method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0136] Embodiment 4

[0137] Figure 5 A schematic diagram of an electronic device 510 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0138] like Figure 5As shown, the electronic device 510 includes at least one processor 511, and a memory connected to the at least one processor 511 in communication, such as a read-only memory (ROM) 512, a random access memory (RAM) 513, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 511 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 512 or the computer program loaded from the storage unit 518 to the random access memory (RAM) 513. In the RAM 513, various programs and data required for the operation of the electronic device 510 can also be stored. The processor 511, the ROM 512, and the RAM 513 are connected to each other via a bus 514. An input / output (I / O) interface 515 is also connected to the bus 514.

[0139] A number of components in the electronic device 510 are connected to the I / O interface 515, including: an input unit 516, such as a keyboard, a mouse, etc.; an output unit 517, such as various types of displays, speakers, etc.; a storage unit 518, such as a disk, an optical disk, etc.; and a communication unit 519, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 519 allows the electronic device 510 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0140] The processor 511 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 511 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 511 executes the various methods and processes described above, such as the gas volume prediction method.

[0141] In some embodiments, the gas volume prediction method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 518. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 510 via the ROM 512 and / or the communication unit 519. When the computer program is loaded into the RAM 513 and executed by the processor 511, one or more steps of the gas volume prediction method described above may be performed. Alternatively, in other embodiments, the processor 511 may be configured to execute the gas volume prediction method in any other appropriate manner (e.g., by means of firmware).

[0142] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0143] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0144] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0145] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0146] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0147] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0148] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0149] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A gas volume prediction method, characterized in that: The method comprises: Obtaining property information of products to be printed and offset printing speed information; Inputting the property information of the product to be printed and the offset printing speed information into the gas volume prediction model to obtain a gas volume prediction result, so as to supply gas according to the gas volume prediction result; The training process of the gas volume prediction model includes: Collect product images during offset printing according to the preset sampling frequency and obtain gas flow information simultaneously; Determine at least one set of sample data according to the detection result of the product image; wherein the sample data includes product attribute information associated with the product image, pre-acquired offset printing speed information, and the gas flow information; The sample data is used to train a radial basis function neural network to obtain a gas volume prediction model.

2. The method according to claim 1, characterized in that The product image is a three-dimensional image; Accordingly, the determining of at least one set of sample data according to the detection result of the product image includes: According to the three-dimensional image, detecting whether there is an abnormal flattening area of ​​the product; If there is no abnormal flattening area, the product attribute information associated with the three-dimensional image, the pre-acquired offset printing speed information, and the gas flow information acquired synchronously with the three-dimensional image are taken as a set of sample data.

3. The method according to claim 2, characterized in that The method of using the sample data to train a radial basis function neural network to obtain a gas volume prediction model includes: Inputting the product attribute information and the offset printing speed information into a radial basis function neural network as model input data to obtain gas volume prediction features; According to the gas volume prediction characteristics and the gas flow information, the radial basis function neural network is optimized until it meets the preset conditions to obtain a gas volume prediction model.

4. The method according to claim 1, characterized in that: The gas volume prediction model includes adjustment parameters; After obtaining the gas volume prediction model, the method further includes: Optimizing the adjustment parameters by genetic algorithm until the optimization termination condition is met, and outputting the optimization result; According to the optimization result, the adjustment parameters of the gas volume prediction model are updated.

5. The method according to claim 4, characterized in that The step of optimizing the adjustment parameters by using a genetic algorithm until an optimization termination condition is met and outputting an optimization result includes: Genetically encoding the adjustment parameters to generate an input population; Performing selection, crossover and mutation operations on the input population to obtain an output population; Calculating the fitness of individuals in the output population, and determining the input population for the next iteration according to the fitness calculation result; If the number of iterations reaches the preset number of iterations, the iterations are terminated, and the optimization result of the adjustment parameters is determined according to the output population finally obtained.

6. The method according to claim 5, characterized in that The adjustment parameters include a center vector and a base width vector; The step of genetically encoding the adjustment parameters to generate an input population comprises: The individuals are genetically encoded in an alternating manner between the elements in the center vector and the elements in the base width vector to generate the input population.

7. A gas volume prediction device, characterized in that: include: An information acquisition module, used to acquire property information of the product to be printed and offset printing speed information; A gas volume prediction result determination module, used for inputting the property information of the product to be printed and the offset printing speed information into the gas volume prediction model to obtain the gas volume prediction result, so as to supply gas according to the gas volume prediction result; Wherein, the device further comprises: An image and flow acquisition module is used to collect product images during the offset printing process according to a preset sampling frequency and simultaneously obtain gas flow information; A sample data determination module, configured to determine at least one set of sample data according to the detection result of the product image; wherein the sample data includes product attribute information associated with the product image, pre-acquired offset printing speed information, and the gas flow information; The gas volume prediction model generation module is used to train a radial basis neural network using the sample data to obtain a gas volume prediction model.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the gas volume prediction method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the gas volume prediction method according to any one of claims 1 to 6 when executed.

Images