A method and system for adjusting the tightening force of a strapping machine

Abstract

The present application relates to the technical field of automatic control, and especially relates to a tightening force adjustment method and system of a strapping machine. The method comprises the following steps: obtaining parameter items corresponding to each sample in a tightening force test set, the parameter items including volume, weight and friction; calculating the volume weight, weight weight and friction weight of the sample; obtaining a fluctuation index of the parameter items based on the weighted mean and weighted value of each parameter item of the sample; calculating a penalty parameter of the parameter items, the penalty parameter being negatively correlated with the fluctuation index of the parameter items; and training a support vector regression model based on the tightening force of the sample, the weight of each parameter item and the penalty parameter, so as to realize tightening force adjustment of the strapping machine, thereby effectively improving the accuracy of tightening force adjustment of the strapping machine.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automatic control, and in particular to a tightening force adjustment method and system of a strapping machine. BACKGROUND

[0002] The strapping machine can be applied to packaging, logistics, warehousing and other industries, and is mainly used for bundling packaging articles, stabilizing goods or materials, etc.

[0003] The tightening force is one of the key parameters in the bundling process. By setting the tightening force, the bundling of the articles to be strapped can be achieved. The size of the tightening force directly affects the stability, appearance and safety of the packaging articles during transportation. If the tightening force is too large, the packaging articles may be deformed or damaged. If the tightening force is too small, the packaging articles may not be tightly bundled, increasing the risk of damage to the goods during transportation. Therefore, when bundling different types of articles, a method for accurately controlling the tightening force is needed to meet the bundling requirements of different articles.

[0004] During the bundling process, the volume, friction, weight, etc. of the articles to be strapped will affect the bundling effect. The requirement for the tightening force will change according to the different characteristics of the articles. For example, for light-weight articles, a smaller bundling force can be set, while for heavy or irregular articles, a stronger bundling force needs to be set.

[0005] Therefore, how to intelligently adjust the tightening force of the strapping machine according to the relevant parameters of the articles after obtaining the relevant parameters of the articles to be strapped, so that the strapping machine can achieve the purpose of bundling the articles while avoiding excessive compression or damage to the articles, is a problem that needs to be solved by those skilled in the art. SUMMARY

[0006] To solve the technical problem of how to intelligently adjust the tightening force of the strapping machine according to the relevant parameters of the articles after obtaining the relevant parameters of the articles to be strapped, so that the strapping machine can achieve the purpose of bundling the articles while avoiding excessive compression or damage to the articles, the present application provides a tightening force adjustment method and system of a strapping machine.

[0007] In a first aspect, the present application provides a tightening force adjustment method of a strapping machine, which adopts the following technical solution:

[0008] A tightening force adjustment method of a strapping machine, comprising the steps of:

[0009] The parameter items corresponding to each sample in the bundling force test set are acquired, the parameter items including volume, weight and friction force; the volume index of the sample is obtained through the difference between the volume of the sample and the volume of other samples, the weight index and the friction force index of the sample are acquired; the volume weight of the sample is obtained according to the sum of the volume index, the weight index and the friction force index of the sample, the weight weight and the friction force weight of the sample are acquired; the weighted mean of each parameter item of the sample is acquired, the fluctuation index of the parameter item is calculated according to the weighted value and the weighted mean of each parameter item of the sample; the penalty parameter of the parameter item is calculated, the penalty parameter is negatively correlated with the fluctuation index of the parameter item; and the support vector regression model is trained based on the bundling force of the sample, the weight of each parameter item and the penalty parameter, so as to realize the bundling force adjustment of the banding machine.

[0010] The present application considers that the relationship between the parameter items of the sample and the bundling force may be linear or nonlinear, and therefore uses the SVR model to adapt to such data changes to construct the bundling force prediction model. In the process of constructing the bundling force prediction model, the present application considers that the contribution degrees of different types of parameter items are different when the bundling force of the sample is acquired; based on this, the present application obtains the weight of each parameter item by analyzing the importance of each parameter item compared with other parameter items, and uses the weight of each parameter item of the sample to construct the model, which can effectively improve the accuracy of the predicted value of the bundling force. On this basis, the present application also considers that the volatility of each parameter item in the bundling force test set will abnormally affect the model construction, and based on this, the present application calculates the corresponding penalty parameter through the volatility of each parameter item, so as to reduce the abnormal influence of the parameter item with strong volatility on the model construction, improve the accuracy of the bundling force prediction model construction, and effectively improve the accuracy of the control of the bundling force adjustment of the banding machine.

[0011] According to the bundling force adjustment method of the banding machine provided by the present application, the parameter items corresponding to each sample in the bundling force test set are acquired, and the method further comprises the following steps: presetting the number of sample categories, and obtaining the bundling force test set after standard normalization processing of the volume data, the weight data and the friction force data of each sample.

[0012] The present application considers that the data values of different parameter items of the sample are greatly different, and the model construction may tend to the parameter item with a larger data value, so the standard normalization processing is performed on each parameter data collected, so as to prepare for the subsequent model construction.

[0013] The method comprises the following steps: taking the cumulative sum of the difference between one sample volume and other sample volumes as the volume index of the sample; taking the cumulative sum of the difference between one sample weight and other sample weights as the weight index of the sample; and taking the cumulative sum of the difference between one sample friction and other sample frictions as the friction index of the sample.

[0014] The method comprises the following steps: taking the cumulative sum of the difference between one sample volume and other sample volumes as the volume index of the sample; taking the cumulative sum of the difference between one sample weight and other sample weights as the weight index of the sample; and taking the cumulative sum of the difference between one sample friction and other sample frictions as the friction index of the sample.

[0015] The method comprises the following steps: taking the cumulative sum of the difference between one sample volume and other sample volumes as the volume index of the sample; taking the cumulative sum of the difference between one sample weight and other sample weights as the weight index of the sample; and taking the cumulative sum of the difference between one sample friction and other sample frictions as the friction index of the sample.

[0016] The method comprises the following steps: taking the cumulative sum of the difference between one sample volume and other sample volumes as the volume index of the sample; taking the cumulative sum of the difference between one sample weight and other sample weights as the weight index of the sample; and taking the cumulative sum of the difference between one sample friction and other sample frictions as the friction index of the sample.

[0017] The method comprises the following steps: taking the cumulative sum of the difference between one sample volume and other sample volumes as the volume index of the sample; taking the cumulative sum of the difference between one sample weight and other sample weights as the weight index of the sample; and taking the cumulative sum of the difference between one sample friction and other sample frictions as the friction index of the sample.

[0018] The method comprises the following steps: taking the cumulative sum of the difference between one sample volume and other sample volumes as the volume index of the sample; taking the cumulative sum of the difference between one sample weight and other sample weights as the weight index of the sample; and taking the cumulative sum of the difference between one sample friction and other sample frictions as the friction index of the sample.

[0019] The constructed bundling force prediction model comprehensively considers the importance and fluctuation degree of each parameter item, calculates the weight and penalty parameter of the parameter item, and trains the bundling force prediction model through the weighted result and the penalty parameter of the parameter item, so that the accuracy of the construction of the bundling force prediction model can be effectively improved.

[0020] According to the bundling force adjustment method of the bundling machine provided by the application, the support vector regression model is trained based on the bundling force of the sample, the weight of each parameter item and the penalty parameter, so as to realize the bundling force adjustment of the bundling machine, and the method comprises the following steps: inputting each parameter item of the to-be-bundled article into the bundling force prediction model to obtain the required bundling force of the to-be-bundled article, and adjusting the bundling force of the bundling machine based on the difference between the actual bundling force of the bundling machine and the required bundling force of the to-be-bundled article.

[0021] According to the bundling force adjustment method of the bundling machine provided by the application, the support vector regression model is trained based on the bundling force of the sample, the weight of each parameter item and the penalty parameter, so as to realize the bundling force adjustment of the bundling machine, and the method comprises the following steps: inputting each parameter item of the to-be-bundled article into the bundling force prediction model to obtain the required bundling force of the to-be-bundled article, and adjusting the bundling force of the bundling machine based on the difference between the actual bundling force of the bundling machine and the required bundling force of the to-be-bundled article.

[0022] After the difference between the actual bundling force of the bundling machine and the required bundling force of the to-be-bundled article is obtained, the control signal can be automatically generated according to the difference between the actual bundling force of the bundling machine and the required bundling force of the to-be-bundled article to adjust the bundling force, so as to reduce the manual intervention and effectively improve the bundling efficiency of the article.

[0023] In the second aspect, the application provides a bundling force adjustment system of a bundling machine, which adopts the following technical scheme:

[0024] The bundling force adjustment system of the bundling machine comprises a processor and a memory, and the memory stores computer program instructions.

[0025] By adopting the above technical scheme, the bundling force adjustment method of the bundling machine is generated into a computer program and stored in the memory, so as to be loaded and executed by the processor, so as to manufacture a terminal device according to the memory and the processor, and facilitate use.

[0026] The application has the following technical effects:

[0027] Based on the technical scheme, the method and system for adjusting the bundling force of the bundling machine are provided, when the bundling force of the bundling machine is adjusted, the SVR prediction model is constructed based on the relationship between the parameter items of the sample and the bundling force, and the bundling force of the bundling machine can be accurately controlled based on the bundling force prediction result. In this process, the application considers that the contribution degrees of different types of parameter items are different when the sample bundling force is obtained. Based on this, the application obtains the weight of each parameter item by analyzing the importance of each parameter item compared with other parameter items, and the model is constructed using the weight of each parameter item of the sample, which can effectively improve the accuracy of the bundling force prediction value. On this basis, the application also considers that there is volatility in each parameter item in the bundling force test set. Based on this, the application calculates the corresponding penalty parameter through the volatility of each parameter item, so as to reduce the abnormal influence of the parameter item with strong volatility on the model construction, improve the accuracy of the bundling force prediction model construction, and effectively improve the accuracy of the control of the bundling force adjustment of the bundling machine. BRIEF DESCRIPTION OF DRAWINGS

[0028] Figure 1 A flowchart of a bundling force adjustment method of a bundling machine is provided for the embodiments of the application. DETAILED DESCRIPTION

[0029] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are part of the embodiments of the application, rather than all the embodiments of the application.

[0030] In order to ensure the bundling effect and quality of the goods, the bundling force of the bundling machine needs to be automatically adjusted.

[0031] Based on this, the embodiments of the application disclose a bundling force adjustment method of a bundling machine, which analyzes the relationship between the sample parameter items and the corresponding bundling force by training a support vector regression (SVR) model, predicts the bundling force of the goods to be bundled, adjusts the bundling force of the bundling machine, and effectively improves the accuracy of the bundling force adjustment of the bundling machine.

[0032] For details, please refer to Figure 1 as shown, Figure 1 A flowchart of a bundling force adjustment method of a bundling machine is provided for the embodiments of the application, which specifically includes the following steps.

[0033] S1: Obtain the parameter items corresponding to each sample in the bundling force test set, and the parameter items include volume, weight and friction.

[0034] It should be noted that the bulky article is more susceptible to external force during transportation or storage, and therefore requires greater tightening force to maintain its stability. The heavy article will generate a greater inertia force when subjected to external force, and if the tightening force is small, it may cause the article to slip or fall during transportation. The greater the friction on the surface of the article, the more difficult it is to move relative to the article, and therefore the tightening force is enhanced.

[0035] Based on this, the embodiment of the present application constructs a tightening force prediction model by analyzing the relationship between the volume, weight and friction of the article and the tightening force, to realize the tightening force prediction of the article to be bound, and controls the tightening force of the binding machine based on the tightening force prediction value of the article to be bound.

[0036] For example, in the embodiment of the present application, the parameter items corresponding to each sample in the tightening force test set are obtained, and before that, the number of sample categories is preset, and after the volume data, weight data and friction data of each sample are standardized and normalized, the tightening force test set is obtained.

[0037] The number of sample categories in the tightening force test set and the number of samples of each category can be set according to actual needs, and the embodiment of the present application does not make too many limitations here.

[0038] For example, after the volume data, weight data and friction data are standardized and normalized, missing data processing and the like can also be performed, which can be set according to actual needs.

[0039] It should be noted that based on the above steps, the parameter items corresponding to each sample can be obtained, but when calculating the tightening force of each type of sample, the contribution degree of each parameter item to the tightening force is different, the tightening force of part of the sample is mainly determined by its volume, and the tightening force of part of the sample may be based on its weight, etc., so the relationship between the tightening force and the parameter items of the sample can be linear or nonlinear.

[0040] Based on this, the embodiment of the present application obtains the tightening force prediction model by training the support vector regression model, and comprehensively analyzes the weight of each parameter item in the training process, so as to accurately obtain the tightening force prediction value of the article to be bound, and realize the adjustment of the tightening force of the binding machine, that is, the following steps are performed.

[0041] S2: obtain the volume index of the sample by the difference between the volume of the sample and the volume of other samples, obtain the weight index and the friction index of the sample; normalize the ratio of the sum of the volume index, the weight index and the friction index of the sample, to obtain the volume weight of the sample, and obtain the weight weight and the friction weight of the sample.

[0042] It should be noted that if the weight of each parameter item is not analyzed, the baling force prediction model obtained by directly training the baling force test set based on the support vector regression model will be more inclined to the relationship between the volume, weight and friction of the sample as a whole and the baling force, and the relationship between the volume, weight and friction of the single sample and the baling force is ignored, thereby leading to different bundling effects of different articles in the subsequent baling force adjustment process.

[0043] Based on this, the embodiment of the present application obtains the parameter index of the parameter item in the sample by analyzing the difference between one parameter item of the sample and other similar parameter items, and obtains the weight of each parameter item in the sample through the proportion of one parameter index and the sum of the other two parameter indexes.

[0044] For example, in the embodiment of the present application, the volume index of the sample is obtained through the difference between the volume of the sample and the volumes of other samples, the weight index and the friction index of the sample are obtained, including: taking the cumulative sum of the difference between the volume of one sample and the volumes of other samples as the volume index of the sample; taking the cumulative sum of the difference between the weight of one sample and the weights of other samples as the weight index of the sample; and taking the cumulative sum of the difference between the friction of one sample and the frictions of other samples as the friction index of the sample.

[0045] After obtaining each index based on the above steps, the weight of each item can be calculated.

[0046] Specifically, the volume weight of the sample is obtained according to the sum of the volume index, the weight index and the friction index.

[0047] For example, in the embodiment of the present application, the weight weight and the friction weight of the sample are obtained, including: normalizing the ratio of the weight index to the sum of the volume index and the friction index to obtain the weight weight of the sample; and normalizing the ratio of the friction index to the sum of the volume index and the weight index to obtain the friction weight of the sample.

[0048] For the convenience of understanding, the embodiment of the present application is described by taking the calculation of the volume weight of the sample as an example, but it does not mean that the embodiment of the present application is limited to this.

[0049] For example, the volume weight of the sample can be calculated according to the following relationship:

[0050] ;

[0051] The volume weight of the i th sample is The number of samples in the baling force test set is The volume of the i th sample is The volume of the i th sample is a volume of the i-th sample, a weight of the i-th sample, a weight of the i-th sample, a weight of the i-th sample, a friction of the i-th sample, a friction of the i-th sample, a friction of the i-th sample, is a standard normalization function.

[0052] In the above formula, represents a volume index of the i-th sample.

[0053] represents a weight index of the i-th sample.

[0054] represents a friction index of the i-th sample.

[0055] The greater the volume index of the i-th sample, the greater the difference between the volume of the i-th sample and the volumes of the other samples; the greater the weight index of the i-th sample, the greater the difference between the weight of the i-th sample and the weights of the other samples; the greater the friction index of the i-th sample, the greater the difference between the friction of the i-th sample and the frictions of the other samples. If the difference between the volume of the i-th sample and the volumes of the other samples is greater than the proportion of the differences between the weight and friction of the i-th sample and the weights and frictions of the other samples, it means that the i-th sample has a greater difference in volume compared to the other samples, and therefore the possibility that the bundling force of the article needs to be adapted to the volume of the sample is greater, i.e. the volume of the article needs to be taken into account during the bundling process, and therefore the weight corresponding to the volume of the test article is greater.

[0056] Similarly, if the difference between the weight of the i-th sample and the weights of the other samples is greater than the proportion of the differences between the volume and friction of the i-th sample and the volumes and frictions of the other samples, it means that the i-th sample has a greater difference in weight compared to the other samples, and therefore the possibility that the bundling force of the article needs to be adapted to the weight of the sample is greater, i.e. the weight of the article needs to be taken into account during the bundling process, and therefore the weight corresponding to the weight of the test article is greater.

[0057] If the difference between the friction of the i-th sample and the frictions of the other samples is greater than the proportion of the differences between the weight and volume of the i-th sample and the weights and volumes of the other samples, it means that the i-th sample has a greater difference in friction compared to the other samples, and therefore the possibility that the bundling force of the article needs to be adapted to the friction of the sample is greater, i.e. the friction of the article needs to be taken into account during the bundling process, and therefore the weight corresponding to the friction of the test article is greater.

[0058] Similarly, the weight weight and the friction weight of the sample can be obtained, and after the weight of each parameter item of each sample is obtained based on the above steps, the following steps are continued to be executed.

[0059] S3: obtaining a weighted mean of each parameter item of the sample, obtaining a fluctuation index of the parameter item through the weighted value and the weighted mean of each parameter item of the sample; and calculating a penalty parameter of the parameter item, the penalty parameter being negatively correlated with the fluctuation index of the parameter item.

[0060] It should be noted that the weight of each parameter item is analyzed based on the above steps, so that the relationship between a single parameter item and the tightening force can be considered in the SVR model training process. In addition, each parameter item in the tightening force test set also has fluctuations, and a parameter item with greater fluctuations may introduce noise or uncertainty in the model training. In the SVR model training process, the greater the penalty parameter, the more the model tends to fit the training data, which may cause overfitting. The smaller the penalty parameter, the simpler the model, and the generalization ability may be enhanced, but the model may be underfitting.

[0061] Based on this, the penalty parameter of the parameter item in the SVR model can be adjusted based on the fluctuation degree of the parameter item in the embodiment of the application. The penalty parameter value of the parameter item with greater fluctuations can be reduced, so that the model can pay more attention to the key features that have an important influence on the prediction result, and the negative influence of the parameter item with greater fluctuations on the model training can be reduced, thereby improving the accuracy of the model prediction.

[0062] For example, in the embodiment of the application, the fluctuation index of the parameter item is calculated, including: taking the absolute value of the difference between the weighted value and the weighted mean of each sample parameter item as the index of the exponential function with e as the base, to obtain the fluctuation index of the parameter item.

[0063] Wherein, the product of the value of the sample parameter item and the corresponding weight is taken as the weighted value of the sample parameter item; and the mean of the weighted values of all samples of the same parameter item is taken as the weighted mean of the parameter item.

[0064] For example, in the embodiment of the application, the penalty parameter of the parameter item is calculated, including: taking the difference between 1 and the fluctuation index of the parameter item as the penalty parameter of the parameter item.

[0065] Wherein, if the parameter item is volume, the penalty parameter of the volume can be obtained based on the above steps; if the parameter item is weight, the penalty parameter of the weight can be obtained based on the above steps; and if the parameter item is friction, the penalty parameter of the friction can be obtained based on the above steps.

[0066] For ease of understanding, the penalty parameter of the volume in the tightening force test set is taken as an example for description in the embodiment of the application, but the embodiment of the application is not limited to this.

[0067] For example, the penalty parameter for calculating the volume can be found in the following formula:

[0068] ;

[0069] The penalty parameter is for volume. The number of samples in the tightness test set. Let i be the volume weight of the i-th sample. Let be the volume of the i-th sample. For the first set of binding force tests Volume weight of each sample For the first set of binding force tests Volume of each sample It is an exponential function with base e. It is the absolute value symbol.

[0070] In the above formula, The weighted value for the volume of the i-th sample is... For the first The weighted value of each sample volume. This is the volume-weighted mean of the binding force test set.

[0071] This represents the volume fluctuation index in the binding force test set. The larger the value, the greater the difference between the volume weighted value and the volume weighted mean in the binding force test set, and the greater the fluctuation of the volume data. In order to reduce the negative impact of the large volume fluctuation data on model training, the penalty parameter needs to be reduced, so the corresponding penalty parameter is also smaller.

[0072] Similarly, by analyzing the fluctuations in weight and friction during the binding force test separately, the penalty parameters for weight and friction can be accurately obtained, and the penalty parameters for each parameter can be used to train the model.

[0073] S4: Based on the sample's binding force, the weights of each parameter, and the penalty parameter, a support vector regression model is trained to adjust the binding force of the strapping machine.

[0074] For example, in an embodiment of the present invention, training a support vector regression model based on the binding force of a sample, the weights of each parameter item, and the penalty parameter includes: weighting the parameter items based on the weights of each parameter item of the sample to obtain weighted parameters; using each weighted parameter of each sample as a set of feature vectors to obtain a feature matrix of the binding force test set; using the binding force of each sample as a target vector, and using the feature matrix, target vector, and penalty parameter to train the SVR model to obtain a binding force prediction model.

[0075] The kernel function in the SVR model can be a radial basis kernel function, and can be set according to actual needs. The specific steps of training the SVR model by using the feature matrix, the target vector and the penalty parameter can be implemented by using the prior art, and will not be repeated here.

[0076] For example, in the embodiment of the present application, the support vector regression model is trained based on the tightening force of the sample, the weight of each parameter item and the penalty parameter to realize the tightening force adjustment of the strapping machine, which includes: inputting each parameter item of the to-be-strapped article into the tightening force prediction model to obtain the required tightening force of the to-be-strapped article, and adjusting the tightening force of the strapping machine based on the difference between the actual tightening force of the strapping machine and the required tightening force of the to-be-strapped article.

[0077] The difference between the actual tightening force of the strapping machine and the required tightening force of the to-be-strapped article can be obtained by the difference between the actual tightening force and the required tightening force of the to-be-strapped article.

[0078] For example, in the embodiment of the present application, the tightening force of the strapping machine is adjusted based on the difference between the actual tightening force of the strapping machine and the required tightening force of the to-be-strapped article, which includes: generating a control signal based on the difference between the actual tightening force of the strapping machine and the required tightening force of the to-be-strapped article, and adjusting the packaging belt by the tensioning component based on the control signal to complete the strapping of the to-be-strapped article.

[0079] The step of adjusting the packaging belt by the tensioning component based on the control signal can be implemented by using a proportional-integral-derivative controller, and will not be repeated here.

[0080] For example, in the embodiment of the present application, the tightening force adjustment of the strapping machine is realized, and then further includes: monitoring the strapping quality of the article after strapping.

[0081] For example, when the strapping quality of the article after strapping is monitored, a high-resolution industrial camera can be installed at the packaging position of the strapping machine to collect images of the article after strapping to construct an image set of the article; the shape of the strap and the corresponding defects in the image set of the article are manually labeled to obtain a labeled image set of the article; and the labeled image set of the article is input into a convolutional neural network for training to obtain a defect recognition model that can be used to identify defects of the strap.

[0082] The image of the article after the latest strapping is input into the defect recognition model to obtain the corresponding strapping quality monitoring result.

[0083] The specific steps of inputting the labeled image set of the article into the convolutional neural network for training to obtain the defect recognition model that can be used to identify defects of the strap can be implemented by using the prior art, and will not be repeated here.

[0084] It can be seen that in the embodiment of the present application, when the bundling force of the bundling machine is adjusted, the parameter items corresponding to each sample in the bundling force test set can be obtained, the parameter items including volume, weight and friction; the volume index of the sample is obtained through the difference between the volume of the sample and the volume of other samples, the weight index and the friction index of the sample are obtained; the volume weight of the sample is obtained according to the sum of the volume index, the weight index and the friction index of the sample, the weight weight and the friction weight of the sample are obtained; the weighted mean of each parameter item of the sample is obtained, the fluctuation index of the parameter item is calculated according to the weighted value and the weighted mean of each parameter item of the sample; the penalty parameter of the parameter item is calculated, the penalty parameter is negatively correlated with the fluctuation index of the parameter item; the support vector regression model is trained based on the bundling force of the sample, the weight of each parameter item and the penalty parameter, so as to realize the bundling force adjustment of the bundling machine, so that the accuracy of the bundling force adjustment of the bundling machine can be effectively improved.

[0085] The embodiment of the present application also discloses a bundling force adjustment system of a bundling machine, comprising a processor and a memory, the memory storing computer program instructions, when the computer program instructions are executed by the processor, realizing the bundling force adjustment method of the bundling machine provided by the present application.

[0086] The above system also includes a communication bus and a communication interface and other components well known to those skilled in the art, the settings and functions of which are known in the art, and therefore will not be repeated here.

[0087] In the present application, the aforementioned memory can be any tangible medium containing or storing a program, which can be used or combined with an instruction execution system, device or apparatus.

[0088] The above are the preferred embodiments of the present application, not limited to the protection scope of the present application, therefore: any equivalent changes made according to the structure, shape, principle of the present application should be covered within the protection scope of the present application.

Claims

1. A method of adjusting a tightening force of a strapping machine, characterized by, The method comprises the following steps: Obtain the parameter items corresponding to each sample in the bundle tightness test set, the parameter items including volume, weight and friction; Obtain the volume index of the sample by the difference between the volume of the sample and the volumes of other samples, and obtain the weight index and the friction index of the sample; Obtain the volume weight of the sample according to the sum of the volume index, the weight index and the friction index of the sample, and obtain the weight weight and the friction weight of the sample; Obtain the weighted mean of each parameter item of the sample, and calculate the fluctuation index of the parameter item according to the weighted value and the weighted mean of each parameter item of the sample; Calculate the penalty parameter of the parameter item, and the penalty parameter is negatively correlated with the fluctuation index of the parameter item; Train a support vector regression model based on the bundle tightness of the sample, the weight of each parameter item and the penalty parameter, so as to realize the bundle tightness adjustment of the bundling machine.

2. The method of claim 1, wherein the tightening force of the banding machine is adjusted by, Before the step of obtaining the parameter items corresponding to each sample in the bundle tightness test set, the method further comprises the following steps: Pre-set the number of sample categories, and obtain the bundle tightness test set after standard normalization processing of the volume data, the weight data and the friction data of each sample.

3. The method of claim 1, wherein the tightening force of the banding machine is adjusted by, The step of obtaining the volume index of the sample by the difference between the volume of the sample and the volumes of other samples, and obtaining the weight index and the friction index of the sample comprises the following steps: Take the cumulative sum of the difference between the volume of one sample and the volumes of other samples as the volume index of the sample, take the cumulative sum of the difference between the weight of one sample and the weights of other samples as the weight index of the sample, and take the cumulative sum of the difference between the friction of one sample and the frictions of other samples as the friction index of the sample.

4. The method of claim 1, wherein the tightening force of the banding machine is adjusted by, The step of obtaining the volume weight of the sample according to the sum of the volume index, the weight index and the friction index of the sample, and obtaining the weight weight and the friction weight of the sample comprises the following steps: Normalize the ratio of the sum of the volume index, the weight index and the friction index to the volume weight of the sample, and obtain the weight weight and the friction weight of the sample; normalize the ratio of the sum of the volume index, the weight index and the friction index to the weight weight of the sample; and normalize the ratio of the sum of the volume index, the weight index and the friction index to the friction weight of the sample.

5. The method of claim 1, wherein the tightening force of the banding machine is adjusted by, The step of calculating the fluctuation index of the parameter item comprises the following steps: Take the absolute value of the difference between the weighted value and the weighted mean of each parameter item of the sample as the index of the exponential function with e as the base, and obtain the fluctuation index of the parameter item.

6. The method of claim 1, wherein the tightening force of the banding machine is adjusted by, The step of calculating the penalty parameter of the parameter item comprises the following steps: Take the difference between 1 and the fluctuation index of the parameter item as the penalty parameter of the parameter item.

7. The method of claim 1, wherein the tightening force of the banding machine is adjusted by, The step of training a support vector regression model based on the bundle tightness of the sample, the weight of each parameter item and the penalty parameter comprises the following steps: Weight the parameter items based on the weight of each parameter item of the sample to obtain weighted parameters, take each weighted parameter of each sample as a group of feature vectors, obtain a feature matrix of the bundle tightness test set, take the bundle tightness of each sample as a target vector, train an SVR model using the feature matrix, the target vector and the penalty parameter, and obtain a bundle tightness prediction model.

8. The method of claim 7, wherein the tightening force of the banding machine is adjusted by, The step of training a support vector regression model based on the bundle tightness of the sample, the weight of each parameter item and the penalty parameter to realize the bundle tightness adjustment of the bundling machine comprises the following steps: The parameters of the to-be-banded article are input into the bundling force prediction model to obtain the required bundling force of the to-be-banded article, and the bundling force of the banding machine is adjusted based on the difference between the actual bundling force of the banding machine and the required bundling force of the to-be-banded article.

9. The method of claim 8, wherein the step of adjusting the tightening force of the banding machine is performed by adjusting the speed of the banding machine. The bundling force of the banding machine is adjusted based on the difference between the actual bundling force of the banding machine and the required bundling force of the to-be-banded article, including: A control signal is generated based on the difference between the actual bundling force of the banding machine and the required bundling force of the to-be-banded article, and the tightening component adjusts the packaging belt based on the control signal to complete the bundling of the to-be-banded article.

10. A tightening force adjustment system of a banding machine, characterized by, Including: A processor and a memory, the memory storing computer program instructions which, when executed by the processor, implement the bundling force adjustment method of the banding machine according to any one of claims 1-9.

Images