Real-time data monitoring system in knitted fabric production process

Abstract

The present invention provides a real-time data monitoring system in a knitted fabric production process, which relates to the field of data monitoring technology, and includes: a real-time monitoring module, which is used to directly collect actual production data from equipment and the environment of a production workshop through various sensors; a data analysis module, which is used to compare the initial prediction value of a prediction model with the actual production data to output the final prediction data, and generate optimization parameters according to the final prediction data; a prediction production module, which is used to establish a virtual production model according to the actual production data, and the prediction model predicts the initial prediction value of the output variable based on the input data; a dynamic adjustment module, which is used to generate a dynamic deviation threshold according to deviation trend data, and generate a corresponding adjustment strategy according to whether the deviation value falls within the deviation threshold. The present invention improves the accuracy and stability of the production process and reduces the influence of human factors.

Description

Technical Field

[0001] The invention relates to the technical field of knitted fabric production, in particular to a real-time data monitoring system in the production process of knitted fabrics. Background Art

[0002] In the production process of knitted fabrics, warp knitting equipment is commonly used as production equipment for mass production of high-quality knitted fabrics. The parameters of the warp knitting equipment need to be set and debugged according to the production process and production process. Accurate parameter settings can maximize the production capacity of the equipment while ensuring product quality.

[0003] At present, equipment parameters often rely on manual operation and experience judgment, and need to be adjusted in time according to the product's production process and production quality. When the input parameters are wrong, it is easy to cause loss of production capacity and quality problems. The quality of the product is greatly affected by human factors, and it is difficult to make adaptive adjustments to the parameters in time as the product quality changes. At the same time, it is not easy to issue timely warnings when abnormal data appears in the equipment production process, which can easily lead to the quality of the products produced being difficult to maintain good stability. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a real-time data monitoring system in the knitted fabric production process, which improves the accuracy and stability of the production process through refined data management, real-time monitoring and automated intelligent feedback.

[0005] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0006] Real-time data monitoring system during knitted fabric production, including:

[0007] Real-time monitoring module, used to collect actual production data directly from the equipment and the production workshop environment through various sensors;

[0008] The prediction production module builds a prediction model based on real-time production data, inputs the preprocessed real-time data into the trained prediction model, and the prediction model predicts the initial prediction value of the output variable based on the input data;

[0009] Data analysis module: compares the initial prediction value of the prediction model with the actual production data, calculates the deviation value at the same timestamp, performs time series analysis on the deviation value, obtains the deviation trend data, adjusts the prediction data according to the deviation trend data to output the final prediction data, and generates optimization parameters according to the final prediction data;

[0010] The dynamic adjustment module is used to generate a dynamic deviation threshold according to the deviation trend data, and to generate a corresponding adjustment strategy according to whether the deviation value falls within the deviation threshold, so that the deviation value returns to the range of the dynamic deviation threshold, and to issue an alarm for abnormal deviation values ​​to guide manual intervention;

[0011] The wireless communication module is used for data interaction between each module and the production equipment. The wireless communication module includes a transmitting end and a receiving end.

[0012] Furthermore, the initial prediction value of the prediction model is compared with the actual production data, and the deviation value at the same timestamp is calculated, including:

[0013] pass Calculate the deviation between the actual production data and the predicted data at the same timestamp, where: for The deviation value at time, and denote the actual and predicted values ​​of the additional state variables, respectively. is the weight matrix of state variables, is the adjustment factor, is the adjustment index, For the The actual value of the state variable, For the The initial predicted values ​​of the state variables, is the number of state variables.

[0014] Furthermore, the deviation values ​​are analyzed in time series to obtain deviation trend data, including:

[0015] pass Calculate the deviation trend value within the set time step, where is the deviation trend value at the current moment, is the length of the time window, is the weight coefficient, , , , is the adjustment factor, , , is the adjustment index, is the deviation value at the current moment, is the deviation value of the previous moment, is the rate of change of the current deviation value, is the acceleration of the deviation change, For the current moment.

[0016] Further, the forecast data is adjusted according to the deviation trend data to output the final forecast data, including:

[0017] pass Calculate the final prediction value, where For the The final predicted value of the state variable, For the The initial predicted values ​​of the state variables, , , , , , is the adjustment factor, , , is the adjustment index, is the rate of change of the deviation trend, is the acceleration of the deviation trend, is the offset constant.

[0018] Furthermore, optimization parameters are generated based on the final prediction data, including:

[0019] pass Calculate the optimized parameters, where Indicates state variables to control parameters sensitivity, For the current moment, For the next moment, is the control parameter vector of the device at the current moment, is the optimization parameter vector for the next moment, is the learning rate, is the weight of each production status parameter; is the regularization coefficient, It is The actual value of the state variable, is the target parameter vector.

[0020] Furthermore, a dynamic deviation threshold is generated based on the deviation trend data, including:

[0021] pass Calculate the dynamic deviation threshold, where is the dynamic deviation threshold, , , , is the adjustment coefficient, is the offset, is the adjustment index.

[0022] Furthermore, according to whether the deviation value falls within the deviation threshold, a corresponding adjustment strategy is generated, including:

[0023] If the deviation value at the current moment Exceeds the current dynamic deviation threshold range, an adjustment strategy is generated, which is to optimize the parameters As The equipment parameters at each moment are input into the equipment control system. If they are not exceeded, no adjustment measures are taken. If they are continuously exceeded, an alarm signal is issued to remind the operator to intervene.

[0024] Furthermore, the wireless communication module includes a transmitting end and a receiving end, including:

[0025] The sending end is integrated into the control system of the production equipment, and the receiving end is connected to the monitoring platform. The data transmission adopts Wi-Fi technology for data transmission, participates in the data interaction between various modules and devices, and realizes data encapsulation and transmission through the IP protocol.

[0026] The above solution of the present invention includes at least the following beneficial effects:

[0027] The above solution of the present invention, through the adjustment of the dynamic deviation threshold, the system can adaptively adjust the control parameters when the production process changes, especially when the raw materials, environmental factors or production conditions fluctuate, the system can still maintain a stable control effect;

[0028] The system only makes adjustments when the deviation exceeds the threshold range, avoiding the interference of frequent parameter adjustments on the production process and improving the stability and continuity of production;

[0029] It realizes automatic monitoring and adjustment of deviations, reduces manual intervention in the production process, improves the level of production automation and reduces human error factors;

[0030] When the deviation continues to exceed the threshold range, the system will issue an early warning signal to prompt the operator to check or intervene, reducing the risk in the production process, preventing production abnormalities caused by out-of-control deviations, and ensuring stable product quality;

[0031] Through dynamic threshold control and early warning mechanisms, the system can quickly respond to fluctuations in the production process, reduce the need for downtime adjustments caused by accumulated production deviations, and ensure the stability of production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of a real-time data monitoring system in a knitted fabric production process provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0033] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0034] like Figure 1 As shown, an embodiment of the present invention provides a real-time data monitoring system in a knitted fabric production process, comprising:

[0035] Real-time monitoring module, used to collect actual production data directly from the equipment and the production workshop environment through various sensors;

[0036] The prediction production module builds a prediction model based on real-time production data, inputs the preprocessed real-time data into the trained prediction model, and the prediction model predicts the initial prediction value of the output variable based on the input data;

[0037] Data analysis module: compares the initial prediction value of the prediction model with the actual production data, calculates the deviation value at the same timestamp, performs time series analysis on the deviation value, obtains the deviation trend data, adjusts the prediction data according to the deviation trend data to output the final prediction data, and generates optimization parameters according to the final prediction data;

[0038] The dynamic adjustment module is used to generate a dynamic deviation threshold according to the deviation trend data, and to generate a corresponding adjustment strategy according to whether the deviation value falls within the deviation threshold, so that the deviation value returns to the range of the dynamic deviation threshold, and to issue an alarm for abnormal deviation values ​​to guide manual intervention;

[0039] The wireless communication module is used for data interaction between each module and the production equipment. The wireless communication module includes a transmitting end and a receiving end.

[0040] In an embodiment of the present invention, the real-time monitoring module collects key actual production data in real time through sensors installed in textile equipment and production workshops. Specifically, the actual production data may include ambient temperature data, ambient air pressure data, equipment vibration data, equipment temperature data, ambient humidity data, equipment speed data, equipment tension data, equipment voltage data, equipment current data, knitted fabric density data, knitted fabric air permeability data, knitted fabric tensile strength data, knitted fabric elongation data, equipment noise data and knitted fabric pilling performance data, and transmits the data collected above to the control system;

[0041] The forecast production module builds a forecast model. Specifically, the construction method may include:

[0042] Step 1, obtain time series data from the real-time monitoring module, including state vector, control input parameters and the corresponding state at the next moment. Define state variables according to the key state variables of the production process, select state variables that can effectively reflect production quality and equipment performance according to the characteristics of the production process, and quantify them.

[0043] Step 2: Combine the selected state vectors into a state vector , ,in Indicates The value quantified by a state variable.

[0044] Step 3, state transfer matrix Describe the interaction between state variables, reflect the dynamic process of transition from one state to the next, determine the relationship between state variables based on experience or experimental data, and convert these relationships into transfer matrices. It represents the transfer relationship from one state to another, that is, the dynamic connection between different state variables. The formula is: ,in, Indicates the current moment, The control input matrix represents the impact of external control factors on the production status (such as equipment temperature, ambient humidity, equipment speed, equipment tension, equipment voltage and equipment current, etc.). is the control parameter vector at the current moment, corresponding to the control parameters of the current device (such as device temperature, ambient humidity, device speed, device tension, device voltage and device current, etc.), that is, the state at the next moment By current status And the control parameter vector at the current moment Decide;

[0045] Specifically, the state transfer matrix Specifically, they may include:

[0046] Step 101, align the state vector with time, expressed as , , construct the state transfer matrix The structure, Among them, the state transfer matrix Each element of Indicates the current moment The state variable has an effect on the next moment The influence degree of a state variable, if It means the The state variable has an effect on the next moment The state variables have no direct effect.

[0047] Step 102: Determine the matrix by fitting historical data The elements of the matrix are determined using historical data Each element in The specific steps may include:

[0048] Step 1021, obtain a state variable data set containing multiple groups of different time steps. The data set format can be expressed as ,in, is the total number of time steps.

[0049] Step 1022, for linear state transition relationships, construct a linear equation, Expressed as The linear combination of is used to fit the equation for each state variable using the least squares method. For each equation, the least squares method is used to solve The optimal value is found by minimizing the error value, making the predicted Close to the actual observed value, use the test data to verify the matrix The accuracy is, specifically, the state in the test data Substitute into the state transfer equation and calculate the predicted , and the actual By contrast, the prediction error is calculated by the mean square error method. If the error is small, it means that the matrix It can effectively describe the interaction between state variables, otherwise it needs to be adjusted or refitted. The physical meaning of each element in can be expressed as: >0, indicating right Have a positive impact; <0, indicating right There are negative impacts; =0, indicating right There is no direct impact;

[0050] Step 1023, when the state transfer relationship is nonlinear, the state transfer equation is expressed as: ,in, Is a nonlinear function used to describe the current state and the control parameter vector The impact on the next state, is the process disturbance vector, representing the external disturbance.

[0051] Constructing nonlinear regression based on the neural network model can include:

[0052] The input is the state vector at the current moment and the control parameter vector at the current moment , the output is the state vector at the next moment ;

[0053] The input layer accepts the current state vector and the current control parameter vector The hidden layer can contain multiple layers of neurons and use the nonlinear activation function ReLU to capture the complex nonlinear relationship between state variables. The number of neurons in the output layer should be the same as the state vector The dimensions are consistent;

[0054] The mean square error (MSE) loss function is used to measure the predicted value. and the true value The error between: ,in, is the sample size;

[0055] Use the training data to train the model, minimize the loss function through the gradient descent method, and adjust the model parameters to fit the nonlinear state transfer relationship;

[0056] Use historical data to train the model and obtain the nonlinear state transfer function During the training process, the model learns the complex nonlinear relationship between the current state and the control parameters for the next moment. The gradient descent method will gradually adjust the model parameters so that the state predicted by the model Will be as close to the real next moment state as possible ;

[0057] Validate the fitted nonlinear model using a test dataset The prediction accuracy of and Input into the trained model and calculate the predicted , the predicted With actual For comparison, the mean square error (MSE) is used for error calculation. If the error is small, it means that the prediction effect of the model is good. If the error is large, the model structure or parameters may need to be adjusted.

[0058] Step 103, determine the control input matrix , the control parameter vector It is expressed as: ,in, Indicates The control variables are The set value at time aligns the state vector with time and is expressed as , , is the number of state variables, the control matrix It can be expressed as,

[0059] ,in, express For state variables The degree of influence of , using historical data, by least squares fitting matrix Each element of , in order to quantify the impact of each control input parameter on the state, if it is known that some control inputs have a clear physical relationship with specific state variables, they can be directly copied, otherwise they need to be fitted based on the data. The specific fitting method can refer to the matrix The fitting method.

[0060] Validate the matrix using test data The accuracy of the control input ensures that the impact of the control input on the state can be accurately predicted and adjusted. The value of makes the model prediction more consistent with the actual situation.

[0061] The initial prediction value of the prediction model Compared with actual production data Compare and calculate the deviation value at the same timestamp , perform time series analysis on the deviation values ​​and obtain deviation trend data , based on the deviation trend data Initial prediction data Adjust to output the final prediction data , according to the final forecast data Generate optimization parameters .

[0062] Introducing dynamic deviation threshold , generate dynamic deviation thresholds based on deviation trend data , depending on whether the deviation value falls within the dynamic deviation threshold Generate the corresponding adjustment strategy to make the deviation value Return to dynamic deviation threshold It will also send out alarms for abnormal deviation values, prompting manual intervention in production.

[0063] Data interaction between modules and devices is achieved through wireless communication modules.

[0064] It should be noted that the installation and use of knitting equipment are well known in the prior art and will not be described in detail here. The selection, installation and collection of related data of sensors are well known in the prior art and will not be described in detail here. The specific state parameters and control input parameters can be adjusted accordingly according to different production processes.

[0065] Compare the initial prediction value of the prediction model with the actual production data and calculate the deviation value at the same timestamp, including:

[0066] pass Calculate the deviation between the actual production data and the predicted data at the same timestamp, where: for The deviation value at time, and denote the actual and predicted values ​​of the additional state variables, respectively. is the weight matrix of state variables, is the adjustment factor, is the adjustment index, For the The actual value of the state variable, For the The initial predicted values ​​of the state variables, is the number of state variables.

[0067] In another preferred embodiment of the present invention, the calculation goal of the deviation value is to measure the difference between the initial predicted value of the prediction model and the actual value. Specifically, by calculating the deviation between the predicted value and the actual value of each state variable, the accuracy of the model can be evaluated and a reference can be provided for further optimization.

[0068] The introduction of external state variables will affect the production process significantly, such as ambient temperature, external air pressure, and vibration factors, which cannot be directly controlled. ;

[0069] For the deviation part of the state variables, we first calculate the difference between the actual value and the predicted value of each state variable, that is, , using the weighting matrix The difference of each state variable is weighted so that different state variables are given different importance in the total deviation value, that is, ,in is a vector Transpose to ensure that the dimensions of the matrix multiplication match, sum the weighted errors of all state variables, and take the square root to get the overall state variable deviation part ;

[0070] For the deviation between external state variables (such as ambient temperature, external air pressure, vibration disturbance, etc.), the difference between the actual value and the predicted value of the external state variable is calculated, that is, , average the errors of all external state variables to get the average difference of all external state variables, that is, , the average deviation of the external state variables is adjusted exponentially, that is , using the adjustment factor Adjust the result and finally get the deviation part of the external state variable, that is, , combine the deviation of the state variable and the deviation of the external state variable to obtain the deviation value at the same time .

[0071] The calculated deviation value , can be used for real-time monitoring, if the deviation value If the smaller the prediction model is, the more accurate it is. If it is large, the difference between the surface model prediction and the actual production data is large, and the model needs to be adjusted.

[0072] Perform time series analysis on the deviation values ​​to obtain deviation trend data, including:

[0073] pass Calculate the deviation trend value within the set time step, where is the deviation trend value at the current moment, is the length of the time window, is the weight coefficient, , , , is the adjustment factor, , , is the adjustment index, is the deviation value at the current moment, is the deviation value of the previous moment, is the rate of change of the current deviation value, is the acceleration of the deviation change.

[0074] In a preferred embodiment of the present invention, the historical data of the deviation, the current deviation value, and the variation of the deviation are comprehensively considered to analyze and obtain the overall trend of the deviation at the current moment, which may specifically include:

[0075] Take from arrive All deviation values ​​in time , that is, in the past The deviation within a time step is calculated by the weight coefficient Each deviation value is weighted to reflect the different impacts of deviations at different time steps on the average value, and all weighted deviation values ​​are averaged, that is, divided by the length of the time window , by adjusting the index Nonlinearly enlarge or reduce the result, and then multiply it by the adjustment coefficient , and obtain the final weighted historical deviation average, that is, , which reflects the overall trend of deviations over the past period of time and helps smooth out current deviation fluctuations.

[0076] Calculate the deviation value of the previous moment of The power is used to make nonlinear adjustments to the deviation of the previous moment. If >1, the deviation of the previous moment will be magnified, which means that if the deviation of the previous moment is large, its weight in the deviation trend value will also be increased. <1, the deviation of the previous moment will be reduced, so that the deviation value of the previous moment will have a relatively small impact on the calculation of the deviation trend. Through this nonlinear adjustment, the deviation value of the previous moment can be flexibly processed to meet different application requirements. ,pass Adjust the influence of the deviation of the previous moment on the deviation trend value of the current moment. If The larger the value, the higher the deviation of the previous moment is in the current moment. If If the value is small, the influence of the previous moment deviation on the current deviation trend value is weak, that is, By adjusting the deviation value of the previous moment and its nonlinearity, the influence of short-term historical deviation on the deviation trend of the current moment is increased.

[0077] The deviation value of the previous moment reflects the short-term trend before the current moment. Introducing it into the calculation of the deviation trend value helps to capture short-term fluctuations and fine-tune the current deviation trend. and adjustment coefficient , the influence of the previous moment deviation on the current moment deviation trend value can be flexibly adjusted according to actual needs. For example, when focusing on short-term fluctuations, and The value of the previous moment can magnify the impact of the deviation at the previous moment; incorporating the deviation at the previous moment into the calculation of the deviation trend can make the deviation trend calculation smoother and avoid the drastic fluctuation caused by the single deviation value at the current moment.

[0078] First Derivative Indicates the rate of increase or decrease of the deviation value at the current moment. A positive value indicates that the deviation is increasing, a negative value indicates that the deviation is decreasing, and a zero value indicates that the deviation remains unchanged. By introducing the first-order derivative, the speed of the deviation change can be reflected in real time. When the deviation change rate is large, it means that the deviation value is changing rapidly; when the change rate is small, it means that the deviation is relatively stable.

[0079] Introduction , used to control the deviation change rate in the overall deviation trend The weight in The size of determines the impact of the deviation change rate on the overall deviation trend. The larger the deviation, the greater the impact of the change rate on the trend value. If the value is small, the influence of this item on the deviation trend is weak. According to actual needs, it can be adjusted The size of the deviation trend makes it more sensitive to rapidly changing deviations, or relatively smooth to avoid excessive fluctuations.

[0080] That is, through By calculating the rate of change of the current deviation and combining it with the adjustment coefficient It can reflect the dynamic change trend of deviation over time, by adjusting The influence of the deviation change rate on the overall deviation trend can be controlled so that the deviation trend value More sensitive to increases and decreases in deviation, thereby improving the ability to respond to rapid fluctuations.

[0081] pass Reflects the change in the rate of change of the deviation, that is, the acceleration of the deviation change. yes The second-order derivative with respect to time indicates how the deviation accelerates or decelerates over time. If the second-order derivative is positive, it means that the rate of change of the deviation is increasing, that is, the deviation is accelerating; if the second-order derivative is negative, it means that the rate of change of the deviation is decreasing, that is, the deviation is decelerating. By adding the acceleration term, the sensitivity to the deviation change trend can be enhanced, especially when the deviation changes at an accelerated rate.

[0082] By introducing , nonlinearly adjust the second-order derivative to control The impact on the deviation trend is >1, the second-order derivative term is amplified, indicating that it is more sensitive to the change trend of acceleration. <1, the second-order derivative term is reduced and the impact on the angular velocity is reduced. Through this nonlinear adjustment, the contribution of the acceleration term to the deviation trend can be flexibly adjusted to meet different dynamic requirements.

[0083] By introducing Used to control the acceleration item in the overall deviation trend value The weight in The value of determines The impact on the overall deviation trend is , the influence of acceleration on the deviation trend will increase, and it will be more sensitive to changes in deviation acceleration or deceleration; if If it is smaller, the influence of acceleration will be weaker, and the calculation of deviation trend will be smoother, avoiding excessive influence of acceleration changes.

[0084] The second-order derivative reflects the increase or decrease in the rate of change of the deviation, thereby helping to determine whether the deviation is accelerating or decelerating. This is very useful for identifying sudden intensification or deceleration of deviation changes. Adding the acceleration term to the deviation trend value calculation formula can make the deviation trend value more sensitive to the acceleration or deceleration of the deviation change, thereby promptly reflecting the rapid changes or abnormal trends that may exist in production. By adjusting the index and adjustment coefficient , the impact of the acceleration term can be flexibly controlled to achieve a balance between the ability to respond to rapid fluctuations and trend smoothing.

[0085] The forecast data is adjusted according to the deviation trend data to output the final forecast data, including:

[0086] pass Calculate the final prediction value, where For the The final predicted value of the state variable, , , , , , is the adjustment factor, , , is the adjustment index, is the rate of change of the deviation trend, is the acceleration of the deviation trend, is the offset constant.

[0087] In an embodiment of the present invention, Represents the initial prediction value Linear adjustment of It is used to control the weight of the initial prediction value in the final prediction value, so that the final prediction value is adjusted based on the initial prediction; represents a nonlinear adjustment to the initial prediction value, For the initial prediction Perform a power transformation so that the impact on the initial prediction value can be nonlinear. Control the impact of this term on the final predicted value by The initial prediction value is nonlinearly enlarged or reduced to increase the flexibility of the prediction. Nonlinear adjustment makes the impact on the initial prediction value more flexible. For example, when >1, the impact of the initial prediction value is amplified; when 0< When <1, the impact of the initial prediction value is reduced. This item provides flexibility for the model, allowing it to be flexibly adjusted according to actual needs.

[0088] pass Reflect the impact of the deviation trend on the final prediction, so that the prediction result can take into account the current deviation trend, Used to control the linear impact of the deviation trend value on the final prediction. Nonlinearly amplify or reduce the impact of deviation trends to increase forecast flexibility. Used to control the weight of this item in the final prediction, For deviation trend A power transformation is performed so that the effect of the deviation trend can be nonlinear.

[0089] pass This makes the final prediction value more sensitive to the dynamic changes of the deviation, thereby improving the accuracy of the prediction. Indicates deviation trend The rate of change over time reflects the speed at which the deviation trend increases or decreases. Used to control the impact of the deviation trend change rate on the final predicted value.

[0090] pass Helps capture the acceleration or deceleration trend of deviation changes, so that the forecast can respond to the rapidly changing deviation trend in a timely manner. Indicates the deviation trend value The acceleration reflects the change of the deviation trend change rate. It is used to make nonlinear adjustments to the deviation acceleration so that the item can be enlarged or reduced. Used to control the impact of deviation acceleration on the final predicted value.

[0091] A benchmark offset is provided so that the forecast value can more accurately reflect the actual situation and the final forecast value can be adjusted as a whole according to the needs.

[0092] Right now The value of not only considers the initial prediction result, but also corrects it through the deviation trend value and its dynamic changes to obtain a more accurate and more dynamically responsive final prediction result. This method enables the prediction result to better adapt to the changes in various state quantities in the production process and provide support for precise control.

[0093] Generate optimization parameters based on the final prediction data, including:

[0094] pass Calculate the optimized parameters, where Indicates The state variables control the input parameters sensitivity, For the current moment, For the next moment, is the control input parameter vector at the current moment, is the optimization parameter vector for the next moment, is the learning rate, is the weight of each production status parameter; is the regularization coefficient, It is The actual value of the state variable, is the target parameter vector.

[0095] In the embodiment of the present invention, by Calculate the total effect of the deviation of each state variable on the control input, Used to indicate the relative importance of each state variable. If a state variable has a greater impact on the production process, it is given a higher weight so that the control parameters can be adjusted first when the deviation is large;

[0096] Indicates the sensitivity of the control input to the state variable. This item is used to measure the influence of the control parameter on each state variable. When the value is high, it is highly sensitive, and even a small adjustment will have a greater impact on the state quantity, thereby more effectively reducing the deviation. When the value is low, it means low sensitivity, and the adjustment strength of the control input can be appropriately increased.

[0097] pass Balance the changes in control parameters to avoid over-adjustment, Regularization coefficient, used to balance the stability of control parameters, control parameters With target parameters This item encourages the control parameters to move closer to the target value to ensure the stability of production.

[0098] Learning Rate Determines the step size of the adjustment. The larger the step size, the faster the control parameters are adjusted, but it may cause instability; the smaller the step size, the smoother the adjustment, but the convergence speed becomes slower.

[0099] By dynamically generating optimization parameters Make the production process more stable and precise, and effectively deal with changes in deviations.

[0100] Generate dynamic deviation thresholds based on deviation trend data, including:

[0101] pass Calculate the dynamic deviation threshold, where is the dynamic deviation threshold, , , , is the adjustment coefficient, is the offset, is the adjustment index.

[0102] In the embodiment of the present invention, the deviation threshold is dynamically adjusted through the deviation trend and the derivative information of the deviation, so that the deviation change can be better coped with in the production process to achieve the purpose of stable control.

[0103] The current deviation can be flexibly adjusted to play different roles in the dynamic threshold calculation. Used to nonlinearly enlarge or reduce the current deviation.

[0104] Acts as a baseline offset to balance and calibrate the overall level of the dynamic deviation threshold.

[0105] Adjustment coefficient , , , and adjustment index The choice should be adjusted according to the characteristics of the actual production process. and The influence of the deviation trend value and the current deviation in the threshold can be enhanced to make the threshold more sensitive to the deviation. and This will make the threshold more sensitive to the rate of change and acceleration of the deviation, thus responding faster to dynamic changes in the deviation. It can be used to adjust the nonlinearity of the current deviation term and select the appropriate nonlinear amplification or reduction effect according to actual needs.

[0106] Depending on whether the deviation value falls within the deviation threshold, a corresponding adjustment strategy is generated, including:

[0107] If the deviation value at the current moment Dynamic Deviation Threshold Exceeded , then an adjustment strategy is generated, which is to optimize the parameters As The equipment parameters at each moment are input into the equipment control system. If they are not exceeded, no adjustment measures are taken. If they are continuously exceeded, an alarm signal is issued to remind the operator to intervene.

[0108] In the embodiment of the present invention, the deviation value of the current moment is calculated in real time at each moment. , and determine whether it is within the dynamic deviation threshold If the deviation value at the current moment is Exceeds the current dynamic deviation threshold , then generate the corresponding adjustment strategy. The adjustment strategy includes two aspects:

[0109] Adjustment direction: Determines whether the control parameter is adjusted upward or downward to bring the deviation back within the threshold range.

[0110] Adjustment step size: Depending on the degree of deviation, the adjustment step size can be varied to be larger or smaller to ensure that the deviation can be effectively controlled.

[0111] Adjustment strategy will be The generated adjustment strategy is applied to the input of control parameters to adjust the operating status of production equipment in order to bring the deviation value back to the dynamic deviation threshold at the next moment. If the deviation value at the current moment At dynamic deviation threshold No adjustment measures will be taken to ensure stability; if the deviation value at the current moment is If the dynamic deviation threshold is continuously exceeded, it indicates that an abnormal situation may have occurred in the production process, such as equipment failure, raw material fluctuation, etc. The alarm device will sound an alarm to remind the operator to intervene to prevent the failure from expanding. This design is intended to prevent the deviation from existing for a long time without being detected, thereby preventing the production process from deviating from expectations.

[0112] The wireless communication module includes a transmitting end and a receiving end, including:

[0113] The sending end is integrated into the control system of the production equipment, and the receiving end is connected to the monitoring platform. The data transmission adopts Wi-Fi technology for data transmission, participates in the data interaction between various modules and devices, and realizes data encapsulation and transmission through the IP protocol.

[0114] In the embodiment of the present invention, the control system of the production equipment obtains the operating status and key parameters of the equipment in real time through sensors and data acquisition modules, and the collected analog signals are converted into digital signals through the data acquisition modules to ensure that the data can be transmitted in the wireless network. After the data conversion is completed, the data is processed by encoding and a check code is added. The check code is used to detect the integrity and accuracy of the data packet during the data transmission process to improve the reliability of data transmission.

[0115] The encoded data is encapsulated into an IP data packet by the sender, and the header information required for transmission is attached, including the source address, destination address, data type and checksum. The system uses Wi-Fi technology for wireless data transmission, which has a high transmission speed and can meet the production data transmission requirements with high real-time requirements. The sender sends the encapsulated IP data packet to the receiving end of the monitoring platform through the Wi-Fi network. The high transmission rate and coverage of the Wi-Fi network enable data to be transmitted quickly and stably from the production equipment to the monitoring platform, meeting the requirements of real-time data monitoring in the factory environment.

[0116] After the data is transmitted to the receiving end, the receiving end first decapsulates the data packet and extracts the real-time status data of the production equipment. During the decapsulation process, the data packet is decoded and the integrity of the data is verified using a checksum. If the check result shows that the data packet is erroneous or incomplete, the system will automatically send a retransmission request to the sender to ensure that the received data is complete and accurate. The retransmission mechanism can ensure the stability and reliability of data transmission when interference or signal fluctuations occur.

[0117] After decoding is completed, the receiver stores the decoded production data in the database of the monitoring platform. The database will update the latest production data to the monitoring interface in real time, and the operator can view the real-time status and production parameters of the production equipment at any time through the interface. The real-time monitoring interface displays the operation status of each device through charts and data, allowing the operator to understand the performance and production status of the production equipment at the first time, providing a basis for timely decision-making and adjustment.

[0118] If the Wi-Fi signal strength is unstable or interference occurs during data transmission, the system will automatically trigger an alarm prompt to remind the operator to check the wireless communication environment to ensure the stability of the wireless network. When a data packet is lost or erroneous, the receiving end will automatically issue a retransmission request, requiring the sending end to resend the relevant data packet. This exception handling mechanism enhances the system's anti-interference ability and ensures the integrity and stability of data transmission. To ensure security during data transmission, the system supports the configuration of advanced encryption protocols such as WPA2 encryption to encrypt data transmission and prevent production data from being intercepted or tampered with during wireless transmission. The combination of encryption algorithm and checksum further ensures data security.

[0119] The high transmission speed of the Wi-Fi network and the real-time data update of the system enable the production monitoring system to meet high real-time requirements and ensure that production data can be accurately monitored at any time. Through WPA2 encryption, checksum and retransmission mechanism, the system has high data transmission security and reliability to prevent data leakage or tampering. The retransmission mechanism and signal monitoring function ensure the stable transmission of data in different production environments and improve the anti-interference ability of the system. Wireless transmission eliminates the physical limitations of wired connections and facilitates the flexible layout and expansion of equipment on the production line. The system is compatible with more production equipment to meet future expansion needs.

[0120] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. Real-time data monitoring system in the knitted fabric production process, characterized in that: include: Real-time monitoring module, used to collect actual production data directly from the equipment and the production workshop environment through various sensors; The prediction production module is used to build a prediction model based on real-time production data, input the pre-processed real-time data into the trained prediction model, and the prediction model predicts the initial prediction value of the output variable based on the input data; Data analysis module: compares the initial prediction value of the prediction model with the actual production data, calculates the deviation value at the same timestamp, performs time series analysis on the deviation value, obtains the deviation trend data, adjusts the prediction data according to the deviation trend data to output the final prediction data, and generates optimization parameters according to the final prediction data; The dynamic adjustment module is used to generate a dynamic deviation threshold according to the deviation trend data, and to generate a corresponding adjustment strategy according to whether the deviation value falls within the deviation threshold, so that the deviation value returns to the range of the dynamic deviation threshold, and to issue an alarm for abnormal deviation values ​​to guide manual intervention; A wireless communication module, used for data exchange between each module and the production equipment, the wireless communication module includes a transmitting end and a receiving end; Compare the initial prediction value of the prediction model with the actual production data and calculate the deviation value at the same timestamp, including: pass Calculate the deviation between the actual production data and the predicted data at the same timestamp, where: for The deviation value at time, and denote the actual and predicted values ​​of the additional state variables, respectively. is the weight matrix of state variables, is the adjustment factor, is the adjustment index, For the The actual value of the state variable, For the The initial predicted values ​​of the state variables, is the number of state variables, is the transpose symbol, indicating is a vector The transpose of Perform time series analysis on the deviation values ​​to obtain deviation trend data, including: pass Calculate the deviation trend value within the set time step, where is the deviation trend value at the current moment, is the length of the time window, is the weight coefficient, , , , is the adjustment factor, , , is the adjustment index, is the deviation value at the current moment, is the deviation value of the previous moment, is the rate of change of the current deviation value, is the acceleration of the deviation change, for the current moment; The forecast data is adjusted according to the deviation trend data to output the final forecast data, including: pass Calculate the final prediction value, where For the The final predicted value of the state variable, For the The initial predicted values ​​of the state variables, , , , , , is the adjustment factor, , , is the adjustment index, is the rate of change of the deviation trend, is the acceleration of the deviation trend, is the offset constant; Generate optimization parameters based on the final prediction data, including: pass Calculate the optimized parameters, where Indicates state variables to control parameters sensitivity, For the current moment, For the next moment, is the control parameter vector of the device at the current moment, is the optimization parameter vector for the next moment, is the learning rate, is the weight of each production status parameter; is the regularization coefficient, It is The actual value of the state variable, is the target parameter vector; Generate dynamic deviation thresholds based on deviation trend data, including: pass Calculate the dynamic deviation threshold, where is the dynamic deviation threshold, , , , is the adjustment coefficient, is the offset, is the adjustment index; Depending on whether the deviation value falls within the deviation threshold, a corresponding adjustment strategy is generated, including: If the deviation value at the current moment Dynamic deviation threshold exceeded , then an adjustment strategy is generated, which is to optimize the parameters As The equipment parameters at the moment are input into the equipment control system. If they are not exceeded, no adjustment measures are taken. If they continue to exceed the dynamic deviation threshold , an alarm signal is issued to remind the operator to intervene; The wireless communication module includes a transmitting end and a receiving end, including: The sending end is integrated into the control system of the production equipment, and the receiving end is connected to the monitoring platform. The data transmission adopts Wi-Fi technology for data transmission, participates in the data interaction between various modules and devices, and realizes data encapsulation and transmission through the IP protocol.

Images