A textile yarn intelligent production control system and method
By establishing data integration end and equipment connection channels in the textile yarn production workshop, integrating sensing data and distance sampling data, identifying equipment and personnel characteristics and training neural network models, the problem of dependence on managers in the existing technology is solved. The intelligent control of the equipment is realized and management pressure is reduced.
Patent Information
- Application Number
- CN202411132381.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-19
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2044-08-19
AI Technical Summary
The existing textile yarn intelligent production equipment control process is too dependent on managers, which makes it impossible to achieve real-time adjustment when managers ask for leave, increasing their dependence on managers and work pressure.
By obtaining the location of the production equipment, establishing a connection channel between the data integration end and the equipment, integrating sensing data and distance sampling data, identifying equipment and personnel characteristics, training neural network models to generate control instructions, and realizing intelligent control of the equipment.
Reliance on managers has been reduced, the work pressure on managers has been alleviated, and the automated and intelligent control of the equipment has been realized.
Smart Images

Figure CN118655858B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of intelligent control technology, and in particular to an intelligent production control system and method for textile yarns. Background Art
[0002] The intelligent production workshop of textile yarn is usually equipped with various modern and automated production equipment to improve production efficiency and product quality. Common production equipment includes cotton openers, carding machines, drawing machines, roving machines, spinning machines, automatic winding machines, automatic twisting machines, intelligent testing equipment, automatic packaging machines and some robots. In the intelligent production workshop, products at each stage can be controlled by controlling these devices.
[0003] The control process of existing production equipment is too dependent on the ability of managers. If the managers take leave for some reason, the entire control process will not be able to be adjusted in real time. How to reduce the dependence of the intelligent production process on management and control personnel is the technical problem that the technical solution of the present invention aims to solve. Summary of the invention
[0004] The object of the present invention is to provide a textile yarn intelligent production control system and method to solve the problems raised in the above background technology.
[0005] To achieve the above object, the present invention provides the following technical solutions:
[0006] A textile yarn intelligent production control method, the method comprising:
[0007] Obtaining the location of each production equipment in a textile yarn production workshop, setting a data integration terminal according to the location, and establishing a connection channel between the data integration terminal and the production equipment;
[0008] Based on the data integration end, the collected data of the production equipment is obtained, the collected data of the production equipment is integrated, and the integrated data is fed back to the master control end; the collected data of the production equipment includes the sensor data in the production equipment and the distance including direction obtained by the distance sampling device installed on the production equipment;
[0009] Based on the master control end, the sensor data after data integration is identified to obtain the device characteristics;
[0010] Obtain surveillance video from the monitoring system, and determine the personnel characteristics of each production equipment based on the distance containing directions in the surveillance video and the integrated data;
[0011] According to the location of each production equipment, the equipment characteristics and personnel characteristics are counted as production characteristics, and the output results of each production equipment are obtained as production labels;
[0012] Train a neural network model from production features to production labels, and determine the control instructions for each production equipment based on the trained neural network model.
[0013] As a further solution of the present invention: the steps of obtaining the positions of various production equipment in the textile yarn production workshop, setting a data integration terminal according to the positions, and establishing a connection channel between the data integration terminal and the production equipment include:
[0014] Get the location of each production equipment in the textile yarn production workshop;
[0015] Obtain the impact of each production equipment on the final product;
[0016] In the workshop map, the center of the circle is determined according to the position, and the radius is determined according to the impact range, so as to obtain the circular nodes corresponding to each production equipment;
[0017] Receive the preset quantity, randomly set the number of data integration points in the workshop map, and calculate the total distance of data transmission;
[0018] Cycle execution, when the number of cycle executions reaches a preset threshold, stop executing the cycle process, select the data integration point when the total data transmission distance is the smallest, and install the data integration terminal;
[0019] For each production equipment, connect it to the nearest data integration terminal.
[0020] As a further solution of the present invention: the step of acquiring the collected data of the production equipment based on the data integration terminal, integrating the collected data of the production equipment, and feeding back the integrated data to the master control terminal includes:
[0021] Based on the data integration terminal, the sensor data of each production equipment connected to it is obtained, and the sensor data containing the equipment tag, sensor tag and time tag is obtained;
[0022] Based on the data integration terminal, the distance including the direction obtained by the distance collection device on each production equipment connected to it is obtained to obtain the distance including the equipment tag, time tag and direction; the direction is the direction relative to the production equipment;
[0023] The acquired data is sorted according to the time tag and the sorted data is fed back to the master control end.
[0024] As a further solution of the present invention: the step of identifying the sensor data after data integration based on the master control terminal to obtain the device characteristics includes:
[0025] Read the sorted sensor data of the same device tag and the same sensor tag;
[0026] Perform discrete Fourier transform on the sorted sensor data to obtain the values at each frequency;
[0027] Select the frequency whose value is greater than a preset threshold as the sensor feature;
[0028] Count the sensor features of all sensor tags of the same device tag as the device features;
[0029] The process of discrete Fourier transform is:
[0030] ; In the formula, The sensor data is The value at the frequency point, It is an array of sensor data sorted within a preset time range. Indicates Sensor data, Indicates the total number of elements in the array.
[0031] As a further solution of the present invention: the step of obtaining the monitoring video of the monitoring system and determining the personnel characteristics of each production equipment according to the distance containing the direction in the monitoring video and the integrated data includes:
[0032] Read the distance with direction in chronological order from the same device tag;
[0033] Read the distance with direction in sequence, determine the starting point according to the device tag, and determine the end point based on the starting point and the distance with direction;
[0034] Read the image frame at the corresponding time in the surveillance video and determine the identification area based on the end point;
[0035] Identify the identification area to obtain regional features;
[0036] Statistical regional characteristics as personnel characteristics.
[0037] As a further solution of the present invention: the step of training the neural network model of production features to production labels and determining the control instructions of each production equipment based on the trained neural network model includes:
[0038] Read a preset number of production features and their corresponding production labels to build a sample set;
[0039] A neural network model that trains production features to production labels based on a sample set;
[0040] In the actual application stage, production features are obtained regularly and input into the trained neural network model to obtain predicted production labels;
[0041] The predicted production tag is compared with the preset standard production tag to determine the control instructions of each production equipment; the control instructions are used to adjust the accuracy and power of the production equipment.
[0042] The technical solution of the present invention also provides a textile yarn intelligent production control system, the system comprising:
[0043] A channel establishment module is used to obtain the location of each production equipment in the textile yarn production workshop, set the data integration terminal according to the location, and establish a connection channel between the data integration terminal and the production equipment;
[0044] A data integration module is used to obtain the collected data of the production equipment based on the data integration terminal, integrate the collected data of the production equipment, and feed back the integrated data to the master control terminal; the collected data of the production equipment includes the sensor data in the production equipment and the distance including direction obtained by the distance sampling device installed on the production equipment;
[0045] The device feature extraction module is used to identify the sensor data after data integration based on the master control end to obtain the device features;
[0046] The personnel feature extraction module is used to obtain the monitoring video of the monitoring system and determine the personnel features of each production equipment according to the distance containing the direction in the monitoring video and the integrated data;
[0047] The production label acquisition module is used to count the equipment characteristics and personnel characteristics according to the location of each production equipment as production characteristics, and obtain the output results of each production equipment as production labels;
[0048] The control instruction adjustment module is used to train the neural network model from production features to production labels, and determine the control instructions of each production equipment based on the trained neural network model.
[0049] As a further solution of the present invention: the channel establishment module includes:
[0050] A position acquisition unit, used to acquire the position of each production equipment in a textile yarn production workshop;
[0051] Amplitude reading unit, used to obtain the impact amplitude of each production equipment on the final product;
[0052] A node generation unit is used to determine the center of a circle according to the position and the radius according to the impact range in the workshop map, so as to obtain a circular node corresponding to each production equipment;
[0053] A distance calculation unit is used to receive a preset number, randomly set the number of data integration points in the workshop map, and calculate the total distance of data transmission;
[0054] A loop execution unit, used for loop execution, when the number of loop executions reaches a preset number threshold, stops executing the loop process, selects a data integration point when the total data transmission distance is the smallest, and installs a data integration terminal;
[0055] The channel establishment unit is used to connect each production equipment with the nearest data integration terminal.
[0056] As a further solution of the present invention: the data integration module includes:
[0057] A sensor data acquisition unit is used to acquire sensor data of each production device connected to the data integration terminal based on the data integration terminal, and obtain sensor data containing device tags, sensor tags and time tags;
[0058] A distance acquisition unit is used to acquire the distance containing the direction acquired by the distance acquisition device on each production device connected to the data integration terminal based on the data integration terminal, and obtain the distance containing the device tag, time tag and direction; the direction is the direction relative to the production device;
[0059] The data sorting unit is used to sort the acquired data according to the time tag and feed back the sorted data to the master control end.
[0060] As a further solution of the present invention: the device feature extraction module includes:
[0061] A sensor data reading unit, used for reading sorted sensor data of the same device tag and the same sensor tag;
[0062] A Fourier transform unit, used for performing discrete Fourier transform on the sorted sensor data to obtain values at each frequency;
[0063] A value selection unit, used for selecting a frequency whose value is greater than a preset threshold value as a sensor feature;
[0064] A feature statistics unit, used to count sensor features of all sensor tags of the same device tag as device features;
[0065] The process of discrete Fourier transform is:
[0066] ; In the formula, The sensor data is The value at the frequency point, It is an array of sensor data sorted within a preset time range. Indicates Sensor data, Indicates the total number of elements in the array.
[0067] Compared with the prior art, the beneficial effects of the present invention are: the technical solution of the present invention determines equipment characteristics based on the sensor data of the equipment's own sensors, determines personnel characteristics based on visual information, and trains a neural network model in combination with equipment characteristics and personnel characteristics. Through the existing data of the production workshop during the production process, including the equipment's own data and the positions of front-line workers, intelligent control of the equipment can be achieved without the need for management personnel. On the one hand, it reduces the degree of dependence on management personnel, and on the other hand, it also alleviates the work pressure of management personnel. BRIEF DESCRIPTION OF THE DRAWINGS
[0068] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention.
[0069] Figure 1 This is a flow chart of the intelligent production control method for textile yarns.
[0070] Figure 2 This is the structural block diagram of the intelligent production control system for textile yarns. DETAILED DESCRIPTION
[0071] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0072] Figure 1 The flowchart of the intelligent production control method of textile yarn is shown in the following figure. In the embodiment of the present invention, a method for intelligent production control of textile yarn is provided, and the method comprises:
[0073] Step S100: obtaining the location of each production equipment in a textile yarn production workshop, setting a data integration terminal according to the location, and establishing a connection channel between the data integration terminal and the production equipment;
[0074] There are multiple production equipment in the textile yarn production workshop. As mentioned in the background technology part, the production equipment includes at least cotton openers, carding machines, drawing machines, roving machines, spinning machines, automatic winding machines, automatic twisting machines, intelligent detection equipment, automatic packaging machines and some robots, etc.; the position of each production equipment is obtained, and the position can be relative coordinates in the production workshop. For example, a corner point in the room is used as the origin, and other positions can be coordinates relative to the origin; the positions of all production equipment are analyzed, and then a data integration terminal is set. The data integration terminal can be understood as a data transfer station, which is used to first analyze the data generated by each production equipment in the production process, and then upload it to the main control terminal for integrated analysis. This is a star model. The data integration terminal is equivalent to a front-end data preprocessing module, which is used to simplify data and facilitate data processing by the main control terminal; after setting up the data integration terminal, a connection channel is established between each data integration terminal and the production equipment closest to it, and the two parties who establish the connection channel can transmit data through the connection channel.
[0075] Step S200: acquiring the collected data of the production equipment based on the data integration terminal, integrating the collected data of the production equipment, and feeding back the integrated data to the master control terminal;
[0076] Based on the data integration end, the collected data of the production equipment is obtained, and the collected data of the production equipment is integrated. Data integration is a data preprocessing process. The collected data of the production equipment is integrated and the integrated data is fed back to the main control end. It should be noted that this process requires that the collected data of the production equipment includes the sensor data in the production equipment and the distance containing direction obtained by the distance sampling device installed on the production equipment; the sensor data in the production equipment is the data collected by the sensor built into the production equipment. The existing production equipment is mature equipment, and the number and type of sensors therein are relatively large. The collected data of these sensors are the sensor data; further, a distance sampling device will also be installed on the production equipment, and the distance sampling device can obtain the distance containing direction, that is, how far the obstacles in each direction are from the production equipment. The distance sampling device can use a rotatable rangefinder.
[0077] Step S300: Identify the sensor data after data integration based on the master control terminal to obtain device characteristics;
[0078] There is a data transmission channel between the master control end and the data integration end. The data integration end uploads the integrated data to the master control end, and the master control end identifies the sensor data part in the integrated data to obtain the device characteristics.
[0079] Step S400: Acquire the monitoring video of the monitoring system, and determine the personnel characteristics of each production equipment according to the distance containing the direction in the monitoring video and the integrated data;
[0080] The master control end obtains the surveillance video of the monitoring system, and at the same time, obtains the directional distance in the data after data integration. The target can be located in the video by the directional distance and the position of the production equipment, and then identified. In the technical solution of the present application, the obstacles around the production equipment are mainly people, and the purpose of the directional distance is also to obtain the position of the staff. Therefore, after locating the target in the video, the personnel characteristics are obtained.
[0081] Step S500: Counting equipment characteristics and personnel characteristics according to the location of each production equipment as production characteristics, and obtaining output results of each production equipment as production labels;
[0082] Step S600: training a neural network model from production features to production labels, and determining control instructions for each production device based on the trained neural network model;
[0083] The equipment characteristics and personnel characteristics of each production equipment are counted as production characteristics. At the same time, the output results of each production equipment are obtained. The output results can use preset evaluation indicators to further represent the output of each production equipment in the form of a table. Multiple production equipment correspond to multiple tables as production labels.
[0084] Obtain production features and production labels for different time periods, build a sample set, and split the sample set into a training set and a test set in a 9:1 ratio. The neural network model can be trained. In the actual application stage, the same method is used to obtain the equipment features and personnel features of the current time period. The trained neural network model is input to predict the output results of each production equipment, that is, predict the production label, compare it with the target input by the staff, and then determine the control instructions for each production equipment. The control instructions are mainly used to control accuracy or power. For each production equipment, if the predicted result can meet the preset target of the staff, the accuracy can be appropriately reduced. If it cannot meet the preset target, the accuracy will be improved. Regarding the accuracy adjustment process, the accuracy of the working equipment in different modes is different, and the adjustment instructions can use the instructions of the selected mode.
[0085] As a preferred embodiment of the technical solution of the present invention, the steps of obtaining the location of each production equipment in the textile yarn production workshop, setting the data integration terminal according to the location, and establishing a connection channel between the data integration terminal and the production equipment include:
[0086] Get the location of each production equipment in the textile yarn production workshop;
[0087] Obtain the impact of each production equipment on the final product;
[0088] In the workshop map, the center of the circle is determined according to the position, and the radius is determined according to the impact range, so as to obtain the circular nodes corresponding to each production equipment;
[0089] Receive the preset quantity, randomly set the number of data integration points in the workshop map, and calculate the total distance of data transmission;
[0090] Cycle execution, when the number of cycle executions reaches a preset threshold, stop executing the cycle process, select the data integration point when the total data transmission distance is the smallest, and install the data integration terminal;
[0091] For each production equipment, connect it to the nearest data integration terminal.
[0092] Obtain the location of each production equipment in the textile yarn production workshop, and obtain the impact of each production equipment on the final product. The simplest case is that the impact of each production equipment is the same. In some cases, if there are additional equipment such as cleaning equipment, the impact of the additional equipment will be lower. In the workshop map, the center of the circle is determined according to the location, and the radius is determined according to the impact. A circle is established in the workshop map, which is called a circular node corresponding to each production equipment. Then, a preset number of data integration points are randomly set in the workshop map, and the data integration points are randomly set multiple times. The data integration point with the smallest total data transmission distance is selected as the location for installing the data integration terminal. Finally, for each production equipment, it is connected to the data integration terminal closest to it.
[0093] As a preferred embodiment of the technical solution of the present invention, the steps of acquiring the collected data of the production equipment based on the data integration terminal, integrating the collected data of the production equipment, and feeding back the integrated data to the master control terminal include:
[0094] Based on the data integration terminal, the sensor data of each production equipment connected to it is obtained, and the sensor data containing the equipment tag, sensor tag and time tag is obtained;
[0095] Based on the data integration terminal, the distance including the direction obtained by the distance collection device on each production equipment connected to it is obtained to obtain the distance including the equipment tag, time tag and direction; the direction is the direction relative to the production equipment;
[0096] The acquired data is sorted according to the time tag and the sorted data is fed back to the master control end.
[0097] The above content describes the data integration process in detail. When acquiring sensor data, device tags, sensor tags and time tags are generated, which are respectively used to indicate which device the data is from, which sensor in the device the data is from, and when the data was acquired. When acquiring the distance including direction, device tags, time tags and directions are generated, which are used to indicate which device has staff near it, at what time the staff exists, and in which direction the staff exists.
[0098] The acquired sensor data and the distance including the direction are sorted according to the time tag and fed back to the master control end as the integrated data.
[0099] As a preferred embodiment of the technical solution of the present invention, the step of identifying the sensor data after data integration based on the master control terminal to obtain device characteristics includes:
[0100] Read the sorted sensor data of the same device tag and the same sensor tag;
[0101] Perform discrete Fourier transform on the sorted sensor data to obtain the values at each frequency;
[0102] Select the frequency whose value is greater than a preset threshold as the sensor feature;
[0103] Count the sensor features of all sensor tags of the same device tag as the device features;
[0104] The process of discrete Fourier transform is:
[0105] ; In the formula, The sensor data is The value at the frequency point, It is an array of sensor data sorted within a preset time range. Indicates Sensor data, Indicates the total number of elements in the array.
[0106] The above content limits the process of extracting equipment features. During normal operation, the production process has clear periodicity, which means that each sensor has periodicity. Although this periodicity is not particularly strict, the data of each sensor can still be analyzed from the perspective of periodicity.
[0107] For any sensor of any device, due to their different types, the periods are also different. Read the sorted sensor data of the same device tag and the same sensor tag, perform discrete Fourier transform on the sorted sensor data, and get the values at each frequency. When a certain value reaches the preset threshold, read the corresponding frequency. Frequency and period are two manifestations of the same property. The read frequency can be used as the characteristic of the sensor. This characteristic is related to periodicity. Under different working conditions, the periodicity of sensor data is different, and the characteristics are also different.
[0108] The sensor features of all sensor tags of the same device tag are counted and sorted in a preset order as the device features, wherein the order includes the first letter of the sensor tag and the spatial position of the sensor.
[0109] As a preferred embodiment of the technical solution of the present invention, the step of obtaining the monitoring video of the monitoring system and determining the personnel characteristics of each production equipment according to the distance containing the direction in the monitoring video and the integrated data includes:
[0110] Read the distance with direction in chronological order from the same device tag;
[0111] Read the distance with direction in sequence, determine the starting point according to the device tag, and determine the end point based on the starting point and the distance with direction;
[0112] Read the image frame at the corresponding time in the surveillance video and determine the identification area based on the end point;
[0113] Identify the identification area to obtain regional features;
[0114] Statistical regional characteristics as personnel characteristics.
[0115] The above content specifically limits the personnel feature extraction process. The personnel feature extraction process of the present application is a feature extraction process based on the video. On this basis, the distance containing the direction is obtained. The distance containing the direction also contains the equipment tag and the time tag. The image at the corresponding moment is read in the video by the time tag, and the working equipment is located by the equipment tag. The located working equipment is used as the starting point, and the end point is obtained in combination with the distance containing the direction. With the end point as the reference, the image within the preset range is intercepted to obtain the recognition area, the recognition area is identified, and the regional features are obtained; all the regional features at the corresponding moment are counted as personnel features.
[0116] Among them, the process of identifying the identification area is mainly a personnel identification process. If the identification fails, the default feature is used as the area feature. The default feature indicates that there is an obstacle there, which can be some mobile devices; if the identification is successful, the visual feature of the person is used as the personnel feature; since the present application first determines a small area based on the distance containing the direction, this is only a part of the image, even if a more complex and accurate recognition algorithm is used, it will not consume too many resources. Therefore, the present application can directly introduce a particularly accurate area recognition algorithm, or even directly use a face recognition algorithm to determine the identity of the staff.
[0117] As a preferred embodiment of the technical solution of the present invention, the step of training the neural network model of production features to production labels and determining the control instructions of each production equipment based on the trained neural network model includes:
[0118] Read a preset number of production features and their corresponding production labels to build a sample set;
[0119] A neural network model that trains production features to production labels based on a sample set;
[0120] In the actual application stage, production features are obtained regularly and input into the trained neural network model to obtain predicted production labels;
[0121] The predicted production tag is compared with the preset standard production tag to determine the control instructions of each production equipment; the control instructions are used to adjust the accuracy and power of the production equipment.
[0122] Read a preset number of production features and their corresponding production labels, build a sample set, split the sample set to obtain a training set and a test set, train a neural network model from production features to production labels based on the training set, and verify the accuracy of the neural network model with the test set. When the error is small enough, the training is completed.
[0123] In the actual application stage, production characteristics are acquired at regular intervals, and the production characteristics are input into the trained neural network model to obtain predicted production labels. The predicted production labels are compared with preset standard production labels to determine the control instructions for each production equipment; the control instructions are used to adjust the accuracy and power of the production equipment.
[0124] Figure 2 : is a structural block diagram of a textile yarn intelligent production control system. In an embodiment of the present invention, a textile yarn intelligent production control system, the system 10 includes:
[0125] A channel establishment module 11 is used to obtain the location of each production equipment in the textile yarn production workshop, set the data integration terminal according to the location, and establish a connection channel between the data integration terminal and the production equipment;
[0126] The data integration module 12 is used to obtain the collected data of the production equipment based on the data integration terminal, integrate the collected data of the production equipment, and feed back the integrated data to the master control terminal; the collected data of the production equipment includes the sensor data in the production equipment and the distance including the direction obtained by the distance sampling device installed on the production equipment;
[0127] The device feature extraction module 13 is used to identify the sensor data after data integration based on the master control terminal to obtain device features;
[0128] The personnel feature extraction module 14 is used to obtain the monitoring video of the monitoring system and determine the personnel features of each production equipment according to the distance containing the direction in the monitoring video and the integrated data;
[0129] The production label acquisition module 15 is used to count the equipment characteristics and personnel characteristics according to the location of each production equipment as production characteristics, and obtain the output results of each production equipment as production labels;
[0130] The control instruction adjustment module 16 is used to train the neural network model from production features to production labels, and determine the control instructions of each production equipment based on the trained neural network model.
[0131] Furthermore, the channel establishing module 11 includes:
[0132] A position acquisition unit, used to acquire the position of each production equipment in a textile yarn production workshop;
[0133] Amplitude reading unit, used to obtain the impact amplitude of each production equipment on the final product;
[0134] A node generation unit is used to determine the center of a circle according to the position and the radius according to the impact range in the workshop map, so as to obtain a circular node corresponding to each production equipment;
[0135] A distance calculation unit is used to receive a preset number, randomly set the number of data integration points in the workshop map, and calculate the total distance of data transmission;
[0136] A loop execution unit, used for loop execution, when the number of loop executions reaches a preset number threshold, stops executing the loop process, selects a data integration point when the total data transmission distance is the smallest, and installs a data integration terminal;
[0137] The channel establishment unit is used to connect each production equipment with the nearest data integration terminal.
[0138] Specifically, the data integration module 12 includes:
[0139] A sensor data acquisition unit is used to acquire sensor data of each production device connected to the data integration terminal based on the data integration terminal, and obtain sensor data containing device tags, sensor tags and time tags;
[0140] A distance acquisition unit is used to acquire the distance containing the direction acquired by the distance acquisition device on each production device connected to the data integration terminal based on the data integration terminal, and obtain the distance containing the device tag, time tag and direction; the direction is the direction relative to the production device;
[0141] The data sorting unit is used to sort the acquired data according to the time tag and feed back the sorted data to the master control end.
[0142] Furthermore, the device feature extraction module 13 includes:
[0143] A sensor data reading unit, used for reading sorted sensor data of the same device tag and the same sensor tag;
[0144] A Fourier transform unit, used for performing discrete Fourier transform on the sorted sensor data to obtain values at each frequency;
[0145] A value selection unit, used for selecting a frequency whose value is greater than a preset threshold value as a sensor feature;
[0146] A feature statistics unit, used to count sensor features of all sensor tags of the same device tag as device features;
[0147] The process of discrete Fourier transform is:
[0148] ; In the formula, The sensor data is The value at the frequency point, It is an array of sensor data sorted within a preset time range. Indicates Sensor data, Indicates the total number of elements in the array.
[0149] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for intelligent production control of textile yarns, It is characterized in that The method comprises: Obtaining the location of each production equipment in a textile yarn production workshop, setting a data integration terminal according to the location, and establishing a connection channel between the data integration terminal and the production equipment; Based on the data integration end, the collected data of the production equipment is obtained, the collected data of the production equipment is integrated, and the integrated data is fed back to the master control end; the collected data of the production equipment includes the sensor data in the production equipment and the distance including direction obtained by the distance sampling device installed on the production equipment; Based on the master control end, the sensor data after data integration is identified to obtain the device characteristics; Obtain surveillance video from the monitoring system, and determine the personnel characteristics of each production equipment based on the distance containing directions in the surveillance video and the integrated data; According to the location of each production equipment, the equipment characteristics and personnel characteristics are counted as production characteristics, and the output results of each production equipment are obtained as production labels; Train the neural network model from production features to production labels, and determine the control instructions for each production equipment based on the trained neural network model; The steps of obtaining the positions of various production equipment in the textile yarn production workshop, setting a data integration terminal according to the positions, and establishing a connection channel between the data integration terminal and the production equipment include: Get the location of each production equipment in the textile yarn production workshop; Obtain the impact of each production equipment on the final product; In the workshop map, the center of the circle is determined according to the position, and the radius is determined according to the impact range, so as to obtain the circular nodes corresponding to each production equipment; Receive the preset quantity, randomly set the number of data integration points in the workshop map, and calculate the total distance of data transmission; Cycle execution, when the number of cycle executions reaches a preset threshold, stop executing the cycle process, select the data integration point when the total data transmission distance is the smallest, and install the data integration terminal; For each production equipment, connect it to the nearest data integration terminal; The step of obtaining the monitoring video of the monitoring system and determining the personnel characteristics of each production equipment according to the distance containing the direction in the monitoring video and the integrated data includes: Read the distance with direction in chronological order from the same device tag; Read the distance with direction in sequence, determine the starting point according to the device tag, and determine the end point based on the starting point and the distance with direction; Read the image frame at the corresponding time in the surveillance video and determine the identification area based on the end point; Identify the identification area to obtain regional features; Statistical regional characteristics as personnel characteristics.
2. The intelligent production control method for textile yarn according to claim 1, It is characterized in that The steps of acquiring the collected data of the production equipment based on the data integration terminal, integrating the collected data of the production equipment, and feeding back the integrated data to the master control terminal include: Based on the data integration terminal, the sensor data of each production equipment connected to it is obtained, and the sensor data containing the equipment tag, sensor tag and time tag is obtained; Based on the data integration terminal, the distance including the direction obtained by the distance collection device on each production equipment connected to it is obtained to obtain the distance including the equipment tag, time tag and direction; the direction is the direction relative to the production equipment; The acquired data is sorted according to the time tag and the sorted data is fed back to the master control end.
3. The intelligent production control method for textile yarn according to claim 2, It is characterized in that The step of identifying the sensor data after data integration based on the master control end to obtain the device characteristics includes: Read the sorted sensor data of the same device tag and the same sensor tag; Perform discrete Fourier transform on the sorted sensor data to obtain the values at each frequency; Select the frequency whose value is greater than a preset threshold as the sensor feature; Count the sensor features of all sensor tags of the same device tag as the device features; The process of discrete Fourier transform is: ; In the formula, The sensor data is The value at the frequency point, It is an array of sensor data sorted within a preset time range. Indicates Sensor data, Indicates the total number of elements in the array.
4. The intelligent production control method for textile yarn according to claim 1, It is characterized in that The step of training the neural network model of production features to production labels and determining the control instructions of each production equipment based on the trained neural network model includes: Read a preset number of production features and their corresponding production labels to build a sample set; A neural network model that trains production features to production labels based on a sample set; In the actual application stage, production features are obtained regularly and input into the trained neural network model to obtain predicted production labels; The predicted production tag is compared with the preset standard production tag to determine the control instructions of each production equipment; the control instructions are used to adjust the accuracy and power of the production equipment.
5. An intelligent production control system for textile yarns, It is characterized in that The system comprises: A channel establishment module is used to obtain the location of each production equipment in the textile yarn production workshop, set the data integration terminal according to the location, and establish a connection channel between the data integration terminal and the production equipment; A data integration module is used to obtain the collected data of the production equipment based on the data integration terminal, integrate the collected data of the production equipment, and feed back the integrated data to the master control terminal; the collected data of the production equipment includes the sensor data in the production equipment and the distance including direction obtained by the distance sampling device installed on the production equipment; The device feature extraction module is used to identify the sensor data after data integration based on the master control end to obtain the device features; The personnel feature extraction module is used to obtain the monitoring video of the monitoring system and determine the personnel features of each production equipment according to the distance containing the direction in the monitoring video and the integrated data; The production label acquisition module is used to count the equipment characteristics and personnel characteristics according to the location of each production equipment as production characteristics, and obtain the output results of each production equipment as production labels; A control instruction adjustment module is used to train a neural network model from production features to production labels, and determine the control instructions of each production equipment based on the trained neural network model; The channel establishment module includes: A position acquisition unit, used to acquire the position of each production equipment in a textile yarn production workshop; Amplitude reading unit, used to obtain the impact amplitude of each production equipment on the final product; A node generation unit is used to determine the center of a circle according to the position and the radius according to the impact range in the workshop map, so as to obtain a circular node corresponding to each production equipment; A distance calculation unit is used to receive a preset number, randomly set the number of data integration points in the workshop map, and calculate the total distance of data transmission; A loop execution unit, used for loop execution, when the number of loop executions reaches a preset number threshold, stops executing the loop process, selects a data integration point when the total data transmission distance is the smallest, and installs a data integration terminal; The channel establishment unit is used to connect each production equipment with the nearest data integration terminal.
6. The intelligent production control system for textile yarn according to claim 5, It is characterized in that The data integration module includes: A sensor data acquisition unit is used to acquire sensor data of each production device connected to the data integration terminal based on the data integration terminal, and obtain sensor data containing device tags, sensor tags and time tags; A distance acquisition unit is used to acquire the distance containing the direction acquired by the distance acquisition device on each production device connected to the data integration terminal based on the data integration terminal, and obtain the distance containing the device tag, time tag and direction; the direction is the direction relative to the production device; The data sorting unit is used to sort the acquired data according to the time tag and feed back the sorted data to the master control end.
7. The intelligent production control system for textile yarn according to claim 6, It is characterized in that The device feature extraction module comprises: A sensor data reading unit, used for reading sorted sensor data of the same device tag and the same sensor tag; A Fourier transform unit, used for performing discrete Fourier transform on the sorted sensor data to obtain values at each frequency; A value selection unit, used for selecting a frequency whose value is greater than a preset threshold value as a sensor feature; A feature statistics unit, used to count sensor features of all sensor tags of the same device tag as device features; The process of discrete Fourier transform is: ; In the formula, The sensor data is The value at the frequency point, It is an array of sensor data sorted within a preset time range. Indicates Sensor data, Indicates the total number of elements in the array.
Citation Information
Cited By
Spinning self-adaptive monitoring control system and method
CN120866952A
