Intelligent regulation and control method and system for cotton processing quality

Abstract

The invention provides an intelligent regulation and control method and system for cotton processing quality. The method comprises the following steps: S1, acquiring collected cotton impurity rate data, moisture regain data, impurity grain number and rolling quality grade; s2, presetting a cotton impurity rate standard value, a moisture regain standard value, an impurity grain number standard value and a rolling quality grade standard value; s3, respectively preprocessing the cotton impurity rate data, the moisture regain data, the number of impurity-containing grains and the rolling quality grade, and S4, taking the preprocessed cotton impurity rate data, moisture regain data, the number of impurity-containing grains and the rolling quality grade as input quantities, making a decision through a cotton rolling quality decision model, and outputting the frequency of a main cotton ginning motor and the frequency of a cotton ginning and feeding motor; and S5, performing regulation and control by utilizing the frequency of the cotton ginning main motor and the frequency of the cotton ginning and feeding motor output by the cotton ginning quality decision model. According to the invention, intelligent decision-making of cotton rolling is realized through the cotton rolling quality decision-making model, and the cotton rolling quality is effectively improved.

Description

Technical Field

[0001] The invention belongs to the technical field of cotton processing quality detection and regulation, and in particular relates to an intelligent regulation method and system for cotton processing quality. Background Art

[0002] Machine-picked cotton has the characteristics of high moisture regain and high impurity content. During the cotton processing process, there are few online detection methods for parameters such as cotton moisture regain, impurity content, and ginning quality that affect the cotton processing quality, and the stability of the detection device is poor. The cotton processing quality control is completely dependent on the experience of the processing personnel, resulting in poor consistency of cotton ginning. Furthermore, the consistency of cotton ginning is mainly guaranteed by adjusting the frequency of the cotton ginning motor, but the frequency of the cotton ginning motor is currently adjusted entirely by hand, and the accuracy cannot be guaranteed. At the same time, for cotton with different moisture regain and impurity content, in order to ensure its consistency and uniformity, the frequency of the cotton ginning motor that needs to be adjusted is different, which will undoubtedly increase the difficulty of manual adjustment and make it difficult to maintain the stability of cotton processing quality. Therefore, in order to improve the quality of cotton processing, there is an urgent need for a method that can intelligently control the frequency of the cotton ginning motor. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art, automatically adjust the speed of the cotton ginning motor, improve the consistency of cotton ginning, improve the accuracy of cotton ginning control, and provide an intelligent control method and system for cotton processing quality.

[0004] In order to achieve the above object, the present invention adopts the following technical solution:

[0005] An intelligent control method for cotton processing quality comprises the following steps:

[0006] S1: Obtain the collected cotton trash content data, moisture regain data, trash grain count and ginning quality grade;

[0007] S2: preset standard values ​​for cotton trash content, moisture regain, number of trash particles and ginning quality grade;

[0008] S3: Pre-process the cotton trash content data, moisture regain data, trash grain count and ginning quality grade respectively.

[0009] S4: taking the pre-processed cotton trash content data, moisture regain data, trash grain count and ginning quality grade as input, and making a decision through a cotton ginning quality decision model, wherein the cotton ginning quality decision model is based on a multi-layer perceptron and outputs a cotton ginning main motor frequency and a cotton ginning feeding motor frequency;

[0010] S5: The cotton ginning main motor frequency and the cotton ginning feeding motor frequency output by the cotton ginning quality decision model are used to regulate the cotton ginning main motor and the cotton ginning feeding motor.

[0011] The present invention realizes intelligent decision-making on cotton ginning quality through a cotton ginning quality decision model. The pre-processed cotton trash content data, moisture regain data, trash particle count and ginning quality grade are taken as input quantities, and the model is used to output the speed of the cotton ginning machine feeding motor and the speed of the main motor. The optimized motor speed is obtained by training the model parameters, thereby improving the consistency of cotton ginning and effectively improving the quality of cotton ginning.

[0012] In order to ensure that the input and output data can be processed at the same level, the input and output samples are normalized. The steps of preprocessing the cotton trash content data, moisture regain data, number of trash particles and ginning quality grade in S3 include: normalizing the cotton trash content data, moisture regain data, number of trash particles and ginning quality grade, and the expressions are:

[0013]

[0014] Where: It indicates the change rate of cotton trash content, a1 indicates the set value of trash content, k1 indicates the detected value of trash content, represents the cotton regain rate change rate, a2 represents the regain setting value, k2 represents the regain detection value, Indicates the rate of change of impurity content, b3 indicates the current impurity content detection value, b2 indicates the last impurity content detection value, and b1 indicates the last impurity content detection value. It indicates the cotton ginning quality grade, and P is the cotton ginning quality grade setting value.

[0015] Preferably, the step of making a decision through the cotton ginning quality decision model in S4 includes:

[0016] S4.1: Definition: A multilayer perceptron consists of an input layer, a hidden layer, and an output layer. The input layer includes four neurons, the output layer includes two neurons, and the hidden layer includes multiple hidden units.

[0017] S4.2: The change rate of cotton trash content, the change rate of cotton regain, the change rate of trash particles and the cotton ginning quality grade are taken as the input of the input layer, and the output of the output layer is the frequency of the cotton ginning main motor and the frequency of the cotton ginning feeding motor respectively;

[0018] S4.3: Use the activation function to transform the output of the hidden layer and define the output relationship of the cotton ginning quality decision model as:

[0019] O(X)=XW h W o +b h W o +b o

[0020] Where: O is the output variable, h is the number of hidden units, W h is the weight parameter of the hidden layer, b h is the bias parameter of the hidden layer, W o is the weight parameter of the output layer, b o is the bias parameter of the output layer.

[0021] S4.4: Set the hyperparameter iteration cycle and learning rate for cotton ginner quality decision model training, train the cotton ginner quality decision model based on the historical data set, and update the weights and biases.

[0022] Preferably, the activation function adopts the ReLU activation function. The expression is:

[0023]

[0024] z=W·X+b

[0025] Where: z is the result of linear transformation, X is the input variable, W is the weight parameter, and b is the bias parameter.

[0026] By utilizing the nonlinear processing capability of the ReLU function, we can effectively avoid the gradient vanishing problem, ensure the training effect of the cotton ginning quality decision model, and handle the complex nonlinear relationship of cotton ginning quality control.

[0027] Preferably, S4 also includes the following steps: S4.5: Based on the historical data set, the cotton ginning quality decision model is tested, a loss function is defined, and the cotton ginning quality decision model is evaluated.

[0028] Preferably, the loss function uses the coefficient of determination, mean absolute error, and mean square error to measure the error between the output of the cotton ginning quality decision model and the actual cotton ginning quality control target, and its expression is:

[0029]

[0030] Among them, R 2 is the coefficient of determination, MAE is the mean absolute error, MSE is the mean square error, n is the number of samples, is the predicted value of the i-th sample, y i is the true value of the i-th sample, is the average of the actual values.

[0031] The present invention uses the mean square error loss function to measure the difference between the model prediction value and the true value. The mean square error reflects the average square error between the predicted value and the true value of each sample. By minimizing MSE, it can ensure that the model outputs a more accurate motor speed control value.

[0032] Preferably, S4 also includes the following steps: S4.6: Iteratively train the cotton ginning quality decision model using the control results. Through iterative training of the multi-layer neural network, the weight matrix and the bias term will be gradually updated to obtain an optimized motor speed control decision, ensure cotton ginning consistency, and effectively improve the cotton ginning quality.

[0033] A cotton processing quality intelligent control system is used to implement the cotton processing quality intelligent control method, comprising: a collection mechanism, an intelligent control module, a cotton ginning main motor, and a cotton ginning feeding motor. The collection mechanism is arranged on a cotton collecting pipeline and is used to collect cotton trash content data, moisture regain data, trash particle count and ginning quality grade; the intelligent control module is arranged on an operating table and is respectively connected to the collection mechanism, the cotton ginning main motor, and the cotton ginning feeding motor.

[0034] The present invention collects trash content and moisture regain data through a trash content data collection mechanism and a moisture regain data collection mechanism, processes the trash content and moisture regain data through an intelligent control module, controls the rotation speed of a cotton ginning main motor and a cotton feeding motor, and effectively improves the cotton ginning quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0036] Figure 1 is a block diagram of the method of the present invention;

[0037] Figure 2 is a schematic diagram of the cotton ginning quality decision model of the present invention;

[0038] Figure 3 It is a schematic diagram of the structure of the system of the present invention.

[0039] Figure numerals: 1. Cotton ginning feeding motor, 2. Cotton ginning main motor, 3. Collection mechanism, 4. Intelligent control module. DETAILED DESCRIPTION

[0040] like Figure 1 As shown, the present invention provides an intelligent control method for cotton processing quality, comprising the following steps:

[0041] S1: Obtain the collected cotton trash content data, moisture regain data, trash grain count and ginning quality grade;

[0042] S2: preset standard values ​​for cotton trash content, moisture regain, number of trash particles and ginning quality grade;

[0043] S3: pre-processing the cotton trash content data, moisture regain data, trash grain count and ginning quality grade respectively;

[0044] S4: The pre-processed cotton trash content data, moisture regain data, trash grain count and ginning quality grade are used as inputs, and a cotton ginning quality decision model is used to make a decision. The cotton ginning quality decision model is based on a multi-layer perceptron and outputs the cotton ginning main motor frequency and the cotton ginning feeding motor frequency;

[0045] S5: The cotton ginning main motor frequency and the cotton ginning feeding motor frequency output by the cotton ginning quality decision model are used to regulate the cotton ginning main motor and the cotton ginning feeding motor.

[0046] The present invention realizes intelligent decision-making on cotton ginning quality through a cotton ginning quality decision model. The pre-processed cotton trash content data, moisture regain data, trash particle count and ginning quality grade are taken as input quantities, and the model is used to output the speed of the cotton ginning machine feeding motor and the speed of the main motor. The optimized motor speed is obtained by training the model parameters, thereby improving the consistency of cotton ginning and effectively improving the quality of cotton ginning.

[0047] In this embodiment, in order to ensure that the input and output data can be processed at the same level, the input and output samples are normalized. The steps of pre-processing the cotton trash content data, moisture regain data, number of trash particles and rolling quality grade in S3 include: normalizing the cotton trash content data, moisture regain data, number of trash particles and rolling quality grade, and the expressions are:

[0048]

[0049] Where: It indicates the change rate of cotton trash content, a1 indicates the set value of trash content, k1 indicates the detected value of trash content, represents the cotton regain rate change rate, a2 represents the regain setting value, k2 represents the regain detection value, Indicates the rate of change of impurity content, b3 indicates the current impurity content detection value, b2 indicates the last impurity content detection value, and b1 indicates the last impurity content detection value. It indicates the cotton ginning quality grade, and P is the cotton ginning quality grade detection value.

[0050] In this embodiment, Figure 2 As shown, the steps of making decisions through the cotton ginning quality decision model in S4 include:

[0051] S4.1: Definition A multilayer perceptron consists of an input layer, a hidden layer, and an output layer. The input layer includes four neurons, the output layer includes two neurons, and the hidden layer includes six hidden units.

[0052] S4.2: The change rate of cotton trash content, the change rate of cotton regain, the change rate of trash particles and the cotton ginning quality grade are taken as the input of the input layer, and the output of the output layer is the frequency of the cotton ginning main motor and the frequency of the cotton ginning feeding motor.

[0053] S4.3: Use the activation function to transform the output of the hidden layer and define the output relationship of the cotton ginning quality decision model as:

[0054] O(X)=XW h W o +b h W o +b o

[0055] Where: O is the output variable, h is the number of hidden units, W h is the weight parameter of the hidden layer, b h is the bias parameter of the hidden layer, W o is the weight parameter of the output layer, b o is the bias parameter of the output layer.

[0056] In this embodiment, the activation function adopts the ReLU activation function. The expression is:

[0057]

[0058] z=W·X+b

[0059] Where: z is the result of linear transformation, X is the input variable, W is the weight parameter, and b is the bias parameter.

[0060] The activation function is a key component for introducing nonlinear characteristics into a neural network, allowing the model to learn complex input-output relationships. In the cotton ginning quality decision model of the present invention, the activation function uses a rectified linear unit (ReLU function for short), which has nonlinear processing capabilities, can effectively avoid the gradient vanishing problem, ensure the training effect of the cotton ginning quality decision model, and handle the complex nonlinear relationship of cotton ginning quality control.

[0061] S4.4: Set the hyperparameter iteration cycle and learning rate for cotton ginner quality decision model training, train the cotton ginner quality decision model based on the historical data set, and update the weights and biases.

[0062] In this embodiment, S4 also includes the following steps: S4.5: Based on the historical data set, the cotton ginning quality decision model is tested, a loss function is defined, and the cotton ginning quality decision model is evaluated.

[0063] In this embodiment, the loss function uses the determination coefficient, mean absolute error, and mean square error to measure the error between the output of the cotton ginning quality decision model and the actual cotton ginning quality control target, and its expression is:

[0064]

[0065] Among them, R 2 is the coefficient of determination, MAE is the mean absolute error, MSE is the mean square error, n is the number of samples, is the predicted value of the i-th sample, y i is the true value of the i-th sample, is the average of the actual values.

[0066] In this embodiment, the cotton ginning quality decision model is evaluated by performance evaluation indicators. The MAE of the model is 0.5047, MSE is 0.4796, and R 2 =0.6084, which shows that the model can better capture the relationship between input and output and has strong predictive ability.

[0067] In this embodiment, the mean square error loss function is used to measure the difference between the model prediction value and the true value. The mean square error reflects the average square error between the predicted value and the true value of each sample. By minimizing MSE, it can be ensured that the model outputs a more accurate motor speed control value.

[0068] In this embodiment, it also includes: S4.6: iteratively training the cotton ginning quality decision model using the control results. Through iterative training of the multi-layer neural network, the weight matrix and the bias term will be gradually updated, thereby obtaining an optimized motor speed control decision, ensuring the consistency of cotton ginning, and effectively improving the quality of cotton ginning.

[0069] like Figure 3 As shown, the present invention provides an intelligent control system for cotton processing quality, which is used to realize the above-mentioned intelligent control method for cotton processing quality, including: a collecting mechanism 3, an intelligent control module 4, a cotton ginning main motor 2, and a cotton ginning feeding motor 1. The collecting mechanism 3 is arranged on the cotton collecting pipeline, and is used to collect cotton impurity content data, regain data, impurity particle count and ginning quality grade; the intelligent control module 4 is arranged on the operating table, and the intelligent control module 4 is respectively connected to the collecting mechanism 3, the cotton ginning main motor 2, and the cotton ginning feeding motor 1.

[0070] The present invention collects trash content and moisture regain data through a trash content data collection mechanism and a moisture regain data collection mechanism, processes the trash content and moisture regain data through an intelligent control module, controls the rotation speed of a cotton ginning main motor and a cotton feeding motor, and effectively improves the cotton ginning quality.

[0071] The above are only preferred embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope defined by the claims.

Claims

1. An intelligent control method for cotton processing quality, characterized in that: The steps include: S1: Obtain the collected cotton trash content data, moisture regain data, trash grain count and ginning quality grade; S2: preset standard values ​​for cotton trash content, moisture regain, number of trash particles and ginning quality grade; S3: Pre-process the cotton trash content data, moisture regain data, trash grain count and rolling quality grade respectively. S4: taking the pre-processed cotton trash content data, moisture regain data, trash grain count and ginning quality grade as input, and making a decision through a cotton ginning quality decision model, wherein the cotton ginning quality decision model is based on a multi-layer perceptron and outputs a cotton ginning main motor frequency and a cotton ginning feeding motor frequency; S5: The cotton ginning main motor frequency and the cotton ginning feeding motor frequency output by the cotton ginning quality decision model are used to regulate the cotton ginning main motor and the cotton ginning feeding motor.

2. The intelligent control method for cotton processing quality according to claim 1, characterized in that: The step of pre-processing the cotton trash content data, moisture regain data, trash particle count and ginning quality grade in S3 comprises: normalizing the cotton trash content data, moisture regain data, trash particle count and ginning quality grade, and the expression is: Where: It indicates the change rate of cotton trash content, a1 indicates the set value of trash content, k1 indicates the detected value of trash content, It indicates the change rate of cotton regain, a2 indicates the set value of regain, k2 indicates the detection value of regain, Indicates the rate of change of impurity content, b3 indicates the current impurity content detection value, b2 indicates the last impurity content detection value, and b1 indicates the last impurity content detection value. It indicates the cotton ginning quality grade, and P is the cotton ginning quality grade setting value.

3. The intelligent control method for cotton processing quality according to claim 2, characterized in that: The steps of making decisions through the cotton ginning quality decision model in S4 include: S4.1: Definition: A multilayer perceptron consists of an input layer, a hidden layer, and an output layer. The input layer includes four neurons, the output layer includes two neurons, and the hidden layer includes multiple hidden units. S4.2: The change rate of cotton trash content, the change rate of cotton regain, the change rate of trash particles and the cotton ginning quality grade are taken as the input of the input layer, and the output of the output layer is the frequency of the cotton ginning main motor and the frequency of the cotton ginning feeding motor respectively; S4.3: Use the activation function to transform the output of the hidden layer and define the output relationship of the cotton ginning quality decision model as: O(X)=XW h W o +b h W o +b o Where: O is the output variable, h is the number of hidden units, W h is the weight parameter of the hidden layer, b h is the bias parameter of the hidden layer, W o is the weight parameter of the output layer, b o is the bias parameter of the output layer; S4.4: Set the hyperparameter iteration cycle and learning rate for cotton ginner quality decision model training, train the cotton ginner quality decision model based on the historical data set, and update the weights and biases.

4. The intelligent control method for cotton processing quality according to claim 3, characterized in that: The activation function adopts the ReLU activation function. The expression is: z=W·X+b Where: z is the result of linear transformation, X is the input variable, W is the weight parameter, and b is the bias parameter.

5. The intelligent control method for cotton processing quality according to claim 3, characterized in that: S4 also includes the following steps: S4.5: Based on the historical data set, the cotton ginning quality decision model is tested, the loss function is defined, and the cotton ginning quality decision model is evaluated.

6. The intelligent control method for cotton processing quality according to claim 5, characterized in that: The loss function uses the coefficient of determination, mean absolute error, and mean square error to measure the error between the output of the cotton ginning quality decision model and the actual cotton ginning quality control target, and its expression is: Among them, R 2 is the coefficient of determination, MAE is the mean absolute error, MSE is the mean square error, n is the number of samples, is the predicted value of the i-th sample, y i is the true value of the i-th sample, is the average of the actual values.

7. The intelligent control method for cotton processing quality according to claim 6, characterized in that: S4 also includes the following steps: S4.6: Iteratively train the cotton ginning quality decision model using the control results.

8. An intelligent control system for cotton processing quality, used to implement the intelligent control method for cotton processing quality according to any one of claims 1 to 7, characterized in that: include: Collection mechanism, intelligent control module, cotton ginning main motor, cotton ginning feeding motor, The collecting mechanism is arranged on the cotton collecting pipeline and is used to collect cotton trash content data, moisture regain data, trash particle count and rolling quality grade; The intelligent control module is arranged on the operating table, and the intelligent control module is respectively connected with the collecting mechanism, the cotton ginning main motor, and the cotton ginning feeding motor.

Images