Automatic intelligent stacking method for agricultural products

Abstract

The invention discloses an automatic intelligent stacking method for agricultural products, which relates to the technical field of intelligent stacking, and comprises the following steps: collecting intelligent stacking data including agricultural product data, agricultural product box data, box transfer data, whole-layer close-packed data and cardboard data, and utilizing the agricultural product data to carry out intelligent stacking; calculating the total weight of the agricultural products and the rotting index of the agricultural products, calculating the air circulation index of the agricultural product box through the data of the agricultural product box in combination with a weight summation method, and constructing a stacking precision evaluation model in combination with the rotting index of the agricultural products, the air circulation index of the agricultural product box and a clamping plate wear coefficient; according to the method, the data acquisition technology, the model construction technology and the intelligent stacking precision control technology are closely combined with the modern information technology, the accurate adjustment of the stacking precision of the agricultural products is achieved, and the intelligent degree in the automatic stacking process of the agricultural products is remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent palletizing, and in particular to an automated intelligent palletizing method for agricultural products. Background Art

[0002] In the process of agricultural modernization, the production scale of agricultural products has continued to expand, and the disadvantages of traditional manual palletizing methods have become increasingly prominent. Manual palletizing is inefficient and labor-intensive. Long-term high-intensity palletizing work can easily cause physical fatigue among workers, leading to health problems. At the same time, labor costs continue to rise, further squeezing the profit margins of agricultural production. In addition, manual palletizing cannot ensure the consistency of palletizing quality. Different workers have different ways of stacking agricultural products, which may lead to poor palletizing stability and easy collapse during transportation and storage, damaging agricultural products. With the rapid development of the Internet of Things, artificial intelligence, and machine algorithm technology, new opportunities have been provided for the transformation of agricultural production. The mature application of automated intelligent palletizing technology has significant advantages in efficiency, precision, and stability. The development of automated intelligent palletizing methods suitable for the characteristics of agricultural production has become a key demand for improving agricultural production efficiency, reducing costs, and ensuring the quality of agricultural products.

[0003] Although the existing technology has made great progress in the direction of intelligent palletizing of agricultural products, there are still some problems that need to be optimized. The existing process of palletizing agricultural products mostly relies on manual labor. During the palletizing process, it is impossible to monitor the deterioration of agricultural products in real time, and it is difficult to improve the accuracy of palletizing. Therefore, how to analyze the deterioration of agricultural products, adjust the palletizing process of agricultural products in real time, and improve the palletizing accuracy of agricultural products is the problem we want to solve. Now we propose an automated intelligent palletizing method for agricultural products. Summary of the Invention

[0004] To achieve the above objectives, the present invention is implemented through the following technical solutions: an automated intelligent palletizing method for agricultural products, comprising the following steps:

[0005] Step 1: Collect intelligent palletizing data including agricultural product data, agricultural product box data, box transfer data, whole layer close-packing data and pallet data, providing a data foundation for the implementation of subsequent steps;

[0006] Step 2: Using agricultural product data, calculate the total weight of agricultural products and the agricultural product spoilage index. Since agricultural products are placed in a closed environment in an agricultural product box, there is a possibility of agricultural product spoilage, which in turn affects the subsequent storage effect of agricultural products. In addition, due to the different materials of agricultural product boxes, for paper agricultural product boxes, when agricultural products are spoiled, the paper agricultural product boxes may soften, which may cause the stacked agricultural products to collapse. By calculating the agricultural product spoilage index, the spoilage of agricultural products during the palletizing process can be monitored, providing technical support for improving the accuracy of agricultural product palletizing.

[0007] Step 3: Using the agricultural product box data and the weighted summation method, calculate the air circulation index of the agricultural product box. This air circulation index is derived from the basis of agricultural product spoilage. If the air circulation of the agricultural product box is poor, it will accelerate the spoilage of some agricultural products. The air circulation index further solves the problem of agricultural product spoilage.

[0008] Step 4: Based on the pallet data, calculate the pallet wear assessment data, then build a pallet wear assessment model and obtain the pallet wear coefficient. Since the existing technology lacks a quantitative indicator for pallet wear, the pallet wear coefficient is introduced to monitor the wear of the pallet and repair or replace it in a timely manner;

[0009] Step 5: Combine the spoilage index of agricultural products, the air circulation index of agricultural product boxes, and the pallet wear coefficient to build a palletizing accuracy assessment model, thereby adjusting the palletizing accuracy of agricultural products and reducing the error of subsequent intelligent palletizing;

[0010] Step 6. Based on the results of the agricultural product palletizing accuracy adjustment, combined with the total weight of the agricultural products, agricultural product box data, box transfer data and whole-layer dense arrangement data, the agricultural products are intelligently palletized.

[0011] A further improvement of the technical solution of the present invention is that in step 1, the process of collecting intelligent palletizing data includes:

[0012] Deploy different data collection equipment, combined with data entry technology, to collect agricultural product data, agricultural product box data, box transfer data, whole-layer dense data, and pallet data. The data collection equipment includes electronic platform scales, texture analyzers, fruit and vegetable respirometers, pH meters, temperature sensors, humidity sensors, oxygen detectors, carbon dioxide detectors, ethylene detectors, wind speed sensors, ultrasonic flow meters, angle sensors, laser range sensors, and anti-slip coefficient measuring instruments.

[0013] The agricultural product data includes the real-time weight, hardness value, elasticity value, carbon dioxide emission per unit time, and real-time pH value of the agricultural product; the agricultural product box data includes the number, temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed, air flow, and maximum load of the agricultural product box; the box transfer data is the real-time placement angle of the agricultural product box during the box transfer process; the whole-layer close-packing data includes the real-time gap between the agricultural product boxes during the whole-layer close-packing process; the pallet data includes the real-time length, real-time width, real-time thickness, real-time height, real-time weight, anti-slip coefficient, and number of uses of the pallet;

[0014] Specifically, an electronic platform scale is used to collect the real-time weight of agricultural products and pallets; a texture analyzer is used to collect the hardness and elasticity values of agricultural products; a fruit and vegetable respirometer and a pH meter are used to collect the carbon dioxide release and pH value of agricultural products per unit time, respectively; a temperature sensor, a humidity sensor, an oxygen detector, a carbon dioxide detector, an ethylene detector, a wind speed sensor and an ultrasonic flow meter are used to collect the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box, respectively; combined with data entry technology, the maximum load and number of the agricultural product box are obtained from the agricultural product box specification table; an angle sensor is used to collect the real-time placement angle of the agricultural product box in the box transfer link; a laser ranging sensor is used to collect the real-time gap between the agricultural product boxes, the real-time length, real-time width, real-time thickness and real-time height of the pallet in the whole layer close arrangement link; an anti-slip coefficient measuring instrument is used to collect the anti-slip coefficient of the pallet, and the number of times the pallet is used is obtained from the pallet barcode through data entry technology;

[0015] Perform data cleaning and data standardization on the collected intelligent palletizing data, assign timestamps to agricultural product data, agricultural product box data, box transfer data, whole-layer close-packed data, and pallet data, and adjust the assigned timestamps to synchronize the collection time of agricultural product data, agricultural product box data, box transfer data, whole-layer close-packed data, and pallet data;

[0016] The agricultural product data is integrated to generate an intelligent palletizing dataset, which is divided into a training set and a test set, with a ratio of 7:3.

[0017] A further improvement of the technical solution of the present invention is that in step 2, the calculation process of the total weight of the agricultural products includes:

[0018] Based on the real-time weight of agricultural products, the sum of the real-time weights of each agricultural product is calculated to obtain the total weight of the agricultural products at each moment;

[0019] Select time points t1 and t2, and then obtain the weight of the agricultural products corresponding to time points t1 and t2 respectively. Then calculate the change rate of the agricultural product weight from t1 to t2, and integrate the change rate of the agricultural product weight from t1 to t2 into the intelligent palletizing data set. The calculation process includes:

[0020]

[0021] Among them, G a is the rate of change of agricultural product weight from t1 to t2, G a1 and G a2 are the weights of agricultural products corresponding to time points t1 and t2 respectively.

[0022] A further improvement of the technical solution of the present invention is that in step 2, the calculation process of the agricultural product spoilage index includes:

[0023] Extract the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value, and weight change rate of agricultural products from t1 to t2 in the intelligent palletizing data set;

[0024] A multivariate linear regression algorithm is used, with training set data as input and agricultural product spoilage index as output. The linear relationship between the hardness, elasticity, carbon dioxide release per unit time, real-time pH value, and the rate of change of agricultural product weight from t1 to t2 and the agricultural product spoilage index is learned to train an agricultural product spoilage assessment model.

[0025] The test set data is input into the agricultural product spoilage assessment model to evaluate the performance of the agricultural product spoilage assessment model. The regression coefficient and intercept term of the agricultural product spoilage assessment model are adjusted to optimize the agricultural product spoilage assessment model, and then the final agricultural product spoilage assessment model is obtained. The expression of the agricultural product spoilage assessment model is:

[0026] H=α1h1+α2h2+α3h3+α4h4+α5G a

[0027] Among them, H is the agricultural product spoilage index, α1, α2, α3, α4 and α5 are the regression coefficients of the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the change rate of the agricultural product weight from t1 to t2, h1, h2, h3, h4 and G a They are the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the rate of change of the weight of the agricultural product from t1 to t2;

[0028] Combined with the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the rate of change of agricultural product weight from t1 to t2, the corresponding agricultural product spoilage index is output.

[0029] A further improvement of the technical solution of the present invention is that in step 3, the calculation process of the air circulation index of the agricultural product box includes:

[0030] According to the air circulation characteristics, weights are assigned to the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box;

[0031] Based on the assigned weights, combined with the weight summation method, the air ventilation index of the agricultural product box is calculated. The calculation process is as follows:

[0032] Y=w1y1+w2y2+w3y3+w4y4+w5y5+w6y6+w7y7

[0033] Among them, Y is the air circulation index of the agricultural product box, w1, w2, w3, w4, w5, w6 and w7 are the weights of the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box, respectively, and y1, y2, y3, y4, y5, y6 and y7 are the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box, respectively.

[0034] A further improvement of the technical solution of the present invention is that in step 4, the calculation process of the pallet wear assessment data includes:

[0035] The pallet wear assessment data includes the length change rate, width change rate, thickness change rate, height change rate and weight change rate of the pallet from t1 to t2;

[0036] Select time points t1 and t2, calculate the length change rate, width change rate, thickness change rate, height change rate, and weight change rate of the pallet from t1 to t2, and then obtain the pallet wear assessment data. The pallet wear assessment data is then integrated into the intelligent palletizing data set. The calculation process includes:

[0037]

[0038] Among them, A, B, C, D and G b are the length change rate, width change rate, thickness change rate and height change rate of the pallet from t1 to t2 respectively; A t1 and A t2 are the lengths of the pallet at time points t1 and t2 respectively; B t1 and B t2are the width of the pallet at time t1 and time t2 respectively; C t1 and C t2 are the thickness of the pallet at time t1 and time t2 respectively; D t1 and D t2 are the heights of the pallet at time points t1 and t2, G b1 and G b2 are the weights of the pallet at time points t1 and t2 respectively.

[0039] A further improvement of the technical solution of the present invention is that in step 4, the process of constructing a pallet wear assessment model and obtaining the pallet wear coefficient includes:

[0040] The pallet wear assessment data, anti-slip coefficient, and number of uses of the pallet are extracted from the intelligent palletizing dataset. The training set data is combined with the neural network algorithm. The pallet wear assessment data, anti-slip coefficient, and number of uses of the pallet are used as input, and the pallet wear coefficient is used as output. The nonlinear relationship between the pallet wear assessment data, the anti-slip coefficient, the number of uses, and the pallet wear coefficient is learned to train the pallet wear assessment model.

[0041] Input the test set data into the pallet wear assessment model, adjust the pallet wear assessment model parameters, optimize the pallet wear assessment model, deploy the pallet wear assessment model into the system, and combine the current pallet wear assessment data as well as the pallet's anti-slip coefficient and number of uses to output the corresponding pallet wear coefficient.

[0042] A further improvement of the technical solution of the present invention is that in step 5, the process of constructing the palletizing accuracy evaluation model includes:

[0043] The agricultural product spoilage index, the air circulation index of the agricultural product box, and the pallet wear coefficient were integrated to generate a palletizing accuracy evaluation dataset. The palletizing accuracy evaluation dataset was then divided into a training set and a test set, with a ratio of 8:2.

[0044] Combining training set data with a multivariate linear regression algorithm, the agricultural product spoilage index, the air circulation index of the agricultural product box, and the pallet wear coefficient are used as input, and the agricultural product palletizing accuracy is used as output. The linear relationship between the agricultural product spoilage index, the air circulation index of the agricultural product box, the pallet wear coefficient, and the agricultural product palletizing accuracy is learned to train a palletizing accuracy evaluation model.

[0045] Input the test set data into the palletizing accuracy evaluation model, adjust the regression coefficient and intercept term of the palletizing accuracy evaluation model, optimize the performance of the palletizing accuracy evaluation model, and obtain the final palletizing accuracy evaluation model. The expression of the palletizing accuracy evaluation model is:

[0046] T=β1H+β2Y+β3M

[0047] Among them, T is the palletizing accuracy of agricultural products, β1, β2 and β3 are the regression coefficients of the agricultural product spoilage index, the air circulation index of the agricultural product box and the pallet wear coefficient respectively, H, Y and M are the agricultural product spoilage index, the air circulation index of the agricultural product box and the pallet wear coefficient respectively;

[0048] Combined with the current agricultural product decay index, the air circulation index of the agricultural product box and the pallet wear coefficient, the corresponding agricultural product palletizing accuracy is output.

[0049] A further improvement of the technical solution of the present invention is that in step 5, the process of adjusting the stacking accuracy of agricultural products includes:

[0050] When the palletizing accuracy of agricultural products is lower than 0.3, it indicates that the agricultural products are abnormally spoiled. A spoilage index threshold for agricultural products is set. When the spoilage index of the corresponding agricultural products is lower than the spoilage index threshold, the temperature, humidity, oxygen concentration, and carbon dioxide concentration of the palletizing environment are adjusted to slow down the spoilage trend of the agricultural products. Agricultural products with spoilage index higher than the spoilage index threshold are removed.

[0051] When the palletizing accuracy of agricultural products is between 0.3 and 0.6, it indicates that there is abnormal air circulation in the agricultural product box. A threshold value of the air circulation index of the agricultural product box is set. When the air circulation index of the corresponding agricultural product box is lower than the threshold value of the air circulation index of the agricultural product box, the changes in the air circulation index of the agricultural product box are continuously monitored to optimize the air circulation in the agricultural product box. For agricultural product boxes with an air circulation index higher than the threshold value of the agricultural product box, more air holes are added to the corresponding agricultural product box and the gaps between the agricultural product boxes are increased.

[0052] When the palletizing accuracy of agricultural products is greater than 0.6, it indicates that the pallet has abnormal wear. A pallet wear coefficient threshold is set. When the pallet wear coefficient is lower than the pallet wear coefficient threshold, the change of the pallet wear coefficient is continuously monitored. When the pallet wear coefficient is higher than the pallet wear coefficient threshold, the worn parts of the pallet are checked, and minor wear is repaired. For severe wear, the pallet is replaced.

[0053] A further improvement of the technical solution of the present invention is that in step 6, based on the agricultural product palletizing accuracy adjustment result, combined with the agricultural product box data, box transfer data and whole layer dense arrangement data, the process of intelligent palletizing of agricultural products includes:

[0054] S1. Compare the total mass of the agricultural products with the maximum load of the agricultural product box, adjust the total mass of the agricultural products to a state less than the maximum load of the agricultural product box, and transport them to the merging main line via the roller conveyor line;

[0055] S2. The belt line transports the agricultural product boxes to the weighing machine and labeling machine. The weighing machine checks the weight of each agricultural product box. The labeling machine assigns a label to the agricultural product box with the box number. The edge roller line adjusts the position of the agricultural product boxes to ensure that the agricultural product boxes are neatly arranged. The agricultural product boxes are then scanned and unqualified agricultural product boxes are rejected based on the scanning results.

[0056] S3. Based on the palletizing method of agricultural products, the placement angle threshold of the agricultural product boxes is set. Combined with the real-time placement angle of the agricultural product boxes, the placement of qualified agricultural product boxes is evaluated. Then, the agricultural product boxes are rotated by the rotating cylinder to ensure that the agricultural product boxes are placed in the correct direction, providing a certain guarantee for improving the palletizing accuracy of agricultural products;

[0057] S4. Using the side roller line, the agricultural product boxes that have passed the box transfer process are arranged on the power roller line in the agricultural product palletizing method, preparing for the subsequent full-layer close stacking, ensuring that the agricultural product boxes are arranged tightly and orderly;

[0058] S5. Push the arranged agricultural product boxes to the roller line through the continuous box pusher. Set the gap threshold of the agricultural product boxes and adjust the gap accordingly based on the gap between the agricultural product boxes. This will achieve a dense arrangement of the agricultural product boxes in a whole layer, combining individual agricultural product boxes into a complete layer, and improving the efficiency and stability of palletizing.

[0059] S6. Use the whole layer pushing line to push the densely packed agricultural product boxes to the palletizing forming line. This step transports the whole layer of agricultural products to the key position for palletizing, making the final preparation for palletizing.

[0060] S7, transport the empty pallet to the palletizing elevator through the pallet conveyor line, scan the empty pallet barcode, and record the pallet information. The pallet information is the pallet data, which is convenient for subsequent management and tracking;

[0061] S8. Control the opening and closing of the palletizing forming line to place the entire layer of agricultural product boxes on the empty pallet to complete the palletizing operation. Then, bind the agricultural product box code with the pallet barcode to achieve information association between the agricultural product box and the pallet;

[0062] S9. After palletizing, the pallet transmits the bound agricultural product box code to the client system through the programmable logic controller to complete the information exchange. The pallet carrying the whole layer of agricultural product boxes is transported to the wrapping machine through the pallet conveyor to wrap the palletized products with protective film. Finally, the agricultural products after film wrapping are transported to the warehouse for storage.

[0063] The beneficial effects of the present invention are as follows: compared with the traditional automatic intelligent palletizing method for agricultural products, the data acquisition technology, model building technology and intelligent palletizing precision control technology in the method of the present invention are closely integrated with modern information technology, accurately capturing agricultural product data, agricultural product box data, transfer box data, whole layer dense data and pallet data, and then obtaining the agricultural product spoilage index, agricultural product box air circulation index and pallet wear coefficient. Combined with the multivariate linear regression algorithm, the palletizing precision of the agricultural product is obtained, thereby achieving precise adjustment of the palletizing precision of the agricultural product and realizing effective monitoring of the palletizing process of the agricultural product. The problem that the existing palletizing process of agricultural products mostly relies on manual labor and cannot monitor the spoilage of agricultural products in real time during the palletizing process and is difficult to improve the palletizing precision is solved. The method of the present invention ensures that the dynamic monitoring standard of the automatic intelligent palletizing of agricultural products can be refined within a more precise range, so that the monitored data becomes a more accurate indicator under the same conditions. The development and application of this method significantly enhances the intelligence level of the automatic palletizing process of agricultural products. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0065] Figure 1 The present invention is a flow chart of an automated intelligent palletizing method for agricultural products. DETAILED DESCRIPTION

[0066] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0067] like Figure 1 As shown, the present invention provides an automated intelligent palletizing method for agricultural products, which comprises the following steps:

[0068] Step 1: Collect intelligent palletizing data including agricultural product data, agricultural product box data, box transfer data, whole layer close-packing data and pallet data, providing a data foundation for the implementation of subsequent steps;

[0069] Step 2: Using agricultural product data, calculate the total weight of agricultural products and the agricultural product spoilage index. Since agricultural products are placed in a closed environment in an agricultural product box, there is a possibility of agricultural product spoilage, which in turn affects the subsequent storage effect of agricultural products. In addition, due to the different materials of agricultural product boxes, for paper agricultural product boxes, when agricultural products are spoiled, the paper agricultural product boxes may soften, which may cause the stacked agricultural products to collapse. By calculating the agricultural product spoilage index, the spoilage of agricultural products during the palletizing process can be monitored, providing technical support for improving the accuracy of agricultural product palletizing.

[0070] Step 3: Using the agricultural product box data and the weighted summation method, calculate the air circulation index of the agricultural product box. This air circulation index is derived from the basis of agricultural product spoilage. If the air circulation of the agricultural product box is poor, it will accelerate the spoilage of some agricultural products. The air circulation index further solves the problem of agricultural product spoilage.

[0071] Step 4: Based on the pallet data, calculate the pallet wear assessment data, then build a pallet wear assessment model and obtain the pallet wear coefficient. Since the existing technology lacks a quantitative indicator for pallet wear, the pallet wear coefficient is introduced to monitor the wear of the pallet and repair or replace it in a timely manner;

[0072] Step 5: Combine the spoilage index of agricultural products, the air circulation index of agricultural product boxes, and the pallet wear coefficient to build a palletizing accuracy assessment model, thereby adjusting the palletizing accuracy of agricultural products and reducing the error of subsequent intelligent palletizing;

[0073] Step 6. Based on the results of the agricultural product palletizing accuracy adjustment, combined with the total weight of the agricultural products, agricultural product box data, box transfer data and whole-layer dense arrangement data, the agricultural products are intelligently palletized.

[0074] In step 1, the collection process of intelligent palletizing data includes:

[0075] Deploy various data collection devices, combined with data entry technology, to collect agricultural product data, agricultural product box data, box transfer data, whole-layer dense data, and pallet data. The collection devices include electronic platform scales, texture analyzers, fruit and vegetable respirometers, pH meters, temperature sensors, humidity sensors, oxygen detectors, carbon dioxide detectors, ethylene detectors, wind speed sensors, ultrasonic flow meters, angle sensors, laser rangefinders, and anti-slip coefficient measuring instruments.

[0076] Agricultural product data includes the real-time weight, hardness, elasticity, carbon dioxide emission per unit time, and pH value of the agricultural products; agricultural product box data includes the box number, temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed, air flow, and maximum load; transfer box data includes the real-time placement angle of the agricultural product boxes during the transfer process; the whole-layer close-packing data includes the real-time gap between the agricultural product boxes during the whole-layer close-packing process; and pallet data includes the real-time length, width, thickness, height, weight, anti-slip coefficient, and number of uses of the pallet.

[0077] Specifically, an electronic platform scale is used to collect the real-time weight of agricultural products and pallets; a texture analyzer is used to collect the hardness and elasticity values of agricultural products; a fruit and vegetable respirometer and a pH meter are used to collect the carbon dioxide release and pH value of agricultural products per unit time, respectively; a temperature sensor, a humidity sensor, an oxygen detector, a carbon dioxide detector, an ethylene detector, a wind speed sensor and an ultrasonic flow meter are used to collect the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box, respectively; combined with data entry technology, the maximum load and number of the agricultural product box are obtained from the agricultural product box specification table; an angle sensor is used to collect the real-time placement angle of the agricultural product box in the box transfer link; a laser ranging sensor is used to collect the real-time gap between the agricultural product boxes, the real-time length, real-time width, real-time thickness and real-time height of the pallet in the whole layer close arrangement link; an anti-slip coefficient measuring instrument is used to collect the anti-slip coefficient of the pallet, and the number of times the pallet is used is obtained from the pallet barcode through data entry technology;

[0078] Perform data cleaning and data standardization on the collected intelligent palletizing data, assign timestamps to agricultural product data, agricultural product box data, box transfer data, whole-layer close-packed data, and pallet data, and adjust the assigned timestamps to synchronize the collection time of agricultural product data, agricultural product box data, box transfer data, whole-layer close-packed data, and pallet data;

[0079] The agricultural product data is integrated to generate an intelligent palletizing dataset, which is divided into a training set and a test set, with a ratio of 7:3.

[0080] In step 2, the calculation process of the total weight of agricultural products includes:

[0081] Based on the real-time weight of agricultural products, the sum of the real-time weights of each agricultural product is calculated to obtain the total weight of the agricultural products at each moment;

[0082] Select time points t1 and t2, and then obtain the weight of the agricultural products corresponding to time points t1 and t2 respectively. Then calculate the change rate of the agricultural product weight from t1 to t2, and integrate the change rate of the agricultural product weight from t1 to t2 into the intelligent palletizing data set. The calculation process includes:

[0083]

[0084] Among them, G a is the rate of change of agricultural product weight from t1 to t2, G a1 and G a2 are the weights of agricultural products corresponding to time points t1 and t2 respectively.

[0085] In step 2, the calculation process of the agricultural product spoilage index includes:

[0086] Extract the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value, and weight change rate of agricultural products from t1 to t2 in the intelligent palletizing data set;

[0087] A multivariate linear regression algorithm is used, with training set data as input and agricultural product spoilage index as output. The linear relationship between the hardness, elasticity, carbon dioxide release per unit time, real-time pH value, and the rate of change of agricultural product weight from t1 to t2 and the agricultural product spoilage index is learned to train an agricultural product spoilage assessment model.

[0088] The test set data is input into the agricultural product spoilage assessment model to evaluate the performance of the agricultural product spoilage assessment model. The regression coefficient and intercept term of the agricultural product spoilage assessment model are adjusted to optimize the agricultural product spoilage assessment model, and then the final agricultural product spoilage assessment model is obtained. The expression of the agricultural product spoilage assessment model is:

[0089] H=α1h1+α2h2+α3h3+α4h4+α5G a

[0090] Among them, H is the agricultural product spoilage index, α1, α2, α3, α4 and α5 are the regression coefficients of the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the change rate of the agricultural product weight from t1 to t2, h1, h2, h3, h4 and G a They are the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the rate of change of the weight of the agricultural product from t1 to t2;

[0091] Combined with the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the rate of change of agricultural product weight from t1 to t2, the corresponding agricultural product spoilage index is output.

[0092] In step 3, the calculation process of the air circulation index of the agricultural product box includes:

[0093] According to the air circulation characteristics, weights are assigned to the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box;

[0094] Based on the assigned weights, combined with the weight summation method, the air ventilation index of the agricultural product box is calculated. The calculation process is as follows:

[0095] Y=w1y1+w2y2+w3y3+w4y4+w5y5+w6y6+w7y7

[0096] Among them, Y is the air circulation index of the agricultural product box, w1, w2, w3, w4, w5, w6 and w7 are the weights of the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box, respectively, and y1, y2, y3, y4, y5, y6 and y7 are the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box, respectively.

[0097] In step 4, the calculation process of the pallet wear assessment data includes:

[0098] The pallet wear assessment data includes the length change rate, width change rate, thickness change rate, height change rate and weight change rate of the pallet from t1 to t2;

[0099] Select time points t1 and t2, calculate the length change rate, width change rate, thickness change rate, height change rate, and weight change rate of the pallet from t1 to t2, and then obtain the pallet wear assessment data. The pallet wear assessment data is then integrated into the intelligent palletizing data set. The calculation process includes:

[0100]

[0101] Among them, A, B, C, D and G b are the length change rate, width change rate, thickness change rate and height change rate of the pallet from t1 to t2 respectively; A t1 and A t2 are the lengths of the pallet at time points t1 and t2 respectively; B t1 and B t2 are the width of the pallet at time t1 and time t2 respectively; C t1 and C t2are the thickness of the pallet at time t1 and time t2 respectively; D t1 and D t2 are the heights of the pallet at time points t1 and t2, G b1 and G b2 are the weights of the pallet at time points t1 and t2 respectively.

[0102] In step 4, the process of building a pallet wear assessment model and obtaining the pallet wear coefficient includes:

[0103] The pallet wear assessment data, anti-slip coefficient, and number of uses of the pallet are extracted from the intelligent palletizing dataset. The training set data is combined with the neural network algorithm. The pallet wear assessment data, anti-slip coefficient, and number of uses of the pallet are used as input, and the pallet wear coefficient is used as output. The nonlinear relationship between the pallet wear assessment data, the anti-slip coefficient, the number of uses, and the pallet wear coefficient is learned to train the pallet wear assessment model.

[0104] Input the test set data into the pallet wear assessment model, adjust the pallet wear assessment model parameters, optimize the pallet wear assessment model, deploy the pallet wear assessment model into the system, and combine the current pallet wear assessment data as well as the pallet's anti-slip coefficient and number of uses to output the corresponding pallet wear coefficient.

[0105] In step 5, the construction process of the palletizing accuracy evaluation model includes:

[0106] The agricultural product spoilage index, the air circulation index of the agricultural product box, and the pallet wear coefficient were integrated to generate a palletizing accuracy evaluation dataset. The palletizing accuracy evaluation dataset was then divided into a training set and a test set, with a ratio of 8:2.

[0107] Combining training set data with a multivariate linear regression algorithm, the agricultural product spoilage index, the air circulation index of the agricultural product box, and the pallet wear coefficient are used as input, and the agricultural product palletizing accuracy is used as output. The linear relationship between the agricultural product spoilage index, the air circulation index of the agricultural product box, the pallet wear coefficient, and the agricultural product palletizing accuracy is learned to train a palletizing accuracy evaluation model.

[0108] Input the test set data into the palletizing accuracy evaluation model, adjust the regression coefficient and intercept term of the palletizing accuracy evaluation model, optimize the performance of the palletizing accuracy evaluation model, and obtain the final palletizing accuracy evaluation model. The expression of the palletizing accuracy evaluation model is:

[0109] T=β1H+β2Y+β3M

[0110] Among them, T is the palletizing accuracy of agricultural products, β1, β2 and β3 are the regression coefficients of the agricultural product spoilage index, the air circulation index of the agricultural product box and the pallet wear coefficient respectively, H, Y and M are the agricultural product spoilage index, the air circulation index of the agricultural product box and the pallet wear coefficient respectively;

[0111] Combined with the current agricultural product decay index, the air circulation index of the agricultural product box and the pallet wear coefficient, the corresponding agricultural product palletizing accuracy is output.

[0112] In step 5, the process of adjusting the palletizing accuracy of agricultural products includes:

[0113] When the palletizing accuracy of agricultural products is lower than 0.3, it indicates that the agricultural products are abnormally spoiled. A spoilage index threshold for agricultural products is set. When the spoilage index of the corresponding agricultural products is lower than the spoilage index threshold, the temperature, humidity, oxygen concentration, and carbon dioxide concentration of the palletizing environment are adjusted to slow down the spoilage trend of the agricultural products. Agricultural products with spoilage index higher than the spoilage index threshold are removed.

[0114] When the palletizing accuracy of agricultural products is between 0.3 and 0.6, it indicates that there is abnormal air circulation in the agricultural product box. A threshold value of the air circulation index of the agricultural product box is set. When the air circulation index of the corresponding agricultural product box is lower than the threshold value of the air circulation index of the agricultural product box, the changes in the air circulation index of the agricultural product box are continuously monitored to optimize the air circulation in the agricultural product box. For agricultural product boxes with an air circulation index higher than the threshold value of the agricultural product box, more air holes are added to the corresponding agricultural product box and the gaps between the agricultural product boxes are increased.

[0115] When the palletizing accuracy of agricultural products is greater than 0.6, it indicates that the pallet has abnormal wear. A pallet wear coefficient threshold is set. When the pallet wear coefficient is lower than the pallet wear coefficient threshold, the change of the pallet wear coefficient is continuously monitored. When the pallet wear coefficient is higher than the pallet wear coefficient threshold, the worn parts of the pallet are checked, and minor wear is repaired. For severe wear, the pallet is replaced.

[0116] In step 6, based on the results of the agricultural product palletizing accuracy adjustment, combined with the agricultural product box data, box transfer data, and whole-layer dense data, the process of intelligent palletizing of agricultural products includes:

[0117] S1. Compare the total mass of the agricultural products with the maximum load of the agricultural product box, adjust the total mass of the agricultural products to a state less than the maximum load of the agricultural product box, and transport them to the merging main line via the roller conveyor line;

[0118] S2. The belt line transports the agricultural product boxes to the weighing machine and labeling machine. The weighing machine checks the weight of each agricultural product box. The labeling machine assigns a label to the agricultural product box with the box number. The edge roller line adjusts the position of the agricultural product boxes to ensure that the agricultural product boxes are neatly arranged. The agricultural product boxes are then scanned and unqualified agricultural product boxes are rejected based on the scanning results.

[0119] S3. Based on the palletizing method of agricultural products, the placement angle threshold of the agricultural product boxes is set. Combined with the real-time placement angle of the agricultural product boxes, the placement of qualified agricultural product boxes is evaluated. Then, the agricultural product boxes are rotated by the rotating cylinder to ensure that the agricultural product boxes are placed in the correct direction, providing a certain guarantee for improving the palletizing accuracy of agricultural products;

[0120] S4. Using the side roller line, the agricultural product boxes that have passed the box transfer process are arranged on the power roller line in the agricultural product palletizing method, preparing for the subsequent full-layer close stacking, ensuring that the agricultural product boxes are arranged tightly and orderly;

[0121] S5. Push the arranged agricultural product boxes to the roller line through the continuous box pusher. Set the gap threshold of the agricultural product boxes and adjust the gap accordingly based on the gap between the agricultural product boxes. This will achieve a dense arrangement of the agricultural product boxes in a whole layer, combining individual agricultural product boxes into a complete layer, and improving the efficiency and stability of palletizing.

[0122] S6. Use the whole layer pushing line to push the densely packed agricultural product boxes to the palletizing forming line. This step transports the whole layer of agricultural products to the key position for palletizing, making the final preparation for palletizing.

[0123] S7, transport the empty pallet to the palletizing elevator through the pallet conveyor line, scan the empty pallet barcode, and record the pallet information. The pallet information is the pallet data, which is convenient for subsequent management and tracking;

[0124] S8. Control the opening and closing of the palletizing forming line to place the entire layer of agricultural product boxes on the empty pallet to complete the palletizing operation. Then, bind the agricultural product box code with the pallet barcode to achieve information association between the agricultural product box and the pallet;

[0125] S9. After palletizing, the pallet transmits the bound agricultural product box code to the client system through the programmable logic controller to complete the information exchange. The pallet carrying the whole layer of agricultural product boxes is transported to the wrapping machine through the pallet conveyor to wrap the palletized products with protective film. Finally, the agricultural products after film wrapping are transported to the warehouse for storage.

[0126] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. An automated intelligent palletizing method for agricultural products, characterized by: The following steps are involved: Step 1: Collect intelligent palletizing data including agricultural product data, agricultural product box data, box transfer data, whole layer close-packing data, and pallet data; Step 2: Using agricultural product data, calculate the total weight of agricultural products and the agricultural product spoilage index; Step 3: Calculate the air circulation index of the agricultural product box using the agricultural product box data and the weighted summation method; Step 4: Based on the pallet data, calculate the pallet wear assessment data, then build a pallet wear assessment model and obtain the pallet wear coefficient; Step 5: Combine the agricultural product decay index, the air circulation index of the agricultural product box, and the pallet wear coefficient to construct a palletizing accuracy evaluation model, and then adjust the agricultural product palletizing accuracy; Step 6. Based on the results of the agricultural product palletizing accuracy adjustment, combined with the total weight of the agricultural products, agricultural product box data, box transfer data and whole-layer dense arrangement data, the agricultural products are intelligently palletized.

2. The method for automated intelligent palletizing of agricultural products according to claim 1, characterized in that: In step 1, the process of collecting intelligent palletizing data includes: Deploy different data collection equipment, combined with data entry technology, to collect agricultural product data, agricultural product box data, box transfer data, whole-layer dense data, and pallet data. The data collection equipment includes electronic platform scales, texture analyzers, fruit and vegetable respirometers, pH meters, temperature sensors, humidity sensors, oxygen detectors, carbon dioxide detectors, ethylene detectors, wind speed sensors, ultrasonic flow meters, angle sensors, laser range sensors, and anti-slip coefficient measuring instruments. The agricultural product data includes the real-time weight, hardness value, elasticity value, carbon dioxide emission per unit time, and real-time pH value of the agricultural product; the agricultural product box data includes the number, temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed, air flow, and maximum load of the agricultural product box; the box transfer data is the real-time placement angle of the agricultural product box during the box transfer process; the whole-layer close-packing data includes the real-time gap between the agricultural product boxes during the whole-layer close-packing process; the pallet data includes the real-time length, real-time width, real-time thickness, real-time height, real-time weight, anti-slip coefficient, and number of uses of the pallet; Perform data cleaning and data standardization on the collected intelligent palletizing data, assign timestamps to agricultural product data, agricultural product box data, box transfer data, whole-layer close-packed data, and pallet data, and adjust the assigned timestamps to synchronize the collection time of agricultural product data, agricultural product box data, box transfer data, whole-layer close-packed data, and pallet data; Agricultural product data are integrated to generate an intelligent palletizing dataset, which is then divided into a training set and a test set.

3. The method for automated intelligent palletizing of agricultural products according to claim 2, characterized in that: In step 2, the calculation process of the total weight of agricultural products includes: Based on the real-time weight of agricultural products, the sum of the real-time weights of each agricultural product is calculated to obtain the total weight of the agricultural products at each moment; Select time points t1 and t2, and then obtain the weight of the agricultural products corresponding to time points t1 and t2 respectively, and then calculate the change rate of the agricultural product weight from t1 to t2, and integrate the change rate of the agricultural product weight from t1 to t2 into the intelligent palletizing dataset.

4. The method for automated intelligent palletizing of agricultural products according to claim 3, characterized in that: In step 2, the calculation process of the agricultural product spoilage index includes: Extract the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value, and weight change rate of agricultural products from t1 to t2 in the intelligent palletizing data set; A multivariate linear regression algorithm is used, with training set data as input and agricultural product spoilage index as output. The linear relationship between the hardness, elasticity, carbon dioxide release per unit time, real-time pH value, and the rate of change of agricultural product weight from t1 to t2 and the agricultural product spoilage index is learned to train an agricultural product spoilage assessment model. The test set data is input into the agricultural product spoilage assessment model to evaluate the performance of the agricultural product spoilage assessment model. The regression coefficient and intercept term of the agricultural product spoilage assessment model are adjusted to optimize the agricultural product spoilage assessment model, thereby obtaining the final agricultural product spoilage assessment model. The expression of the agricultural product spoilage assessment model is as follows: H=α1h1+α2h2+α3h3+α4h4+α5G a Among them, H is the agricultural product spoilage index, α1, α2, α3, α4 and α5 are the regression coefficients of the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the change rate of the agricultural product weight from t1 to t2, h1, h2, h3, h4 and G a They are the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the rate of change of the weight of the agricultural product from t1 to t2; Combined with the hardness value, elasticity value, carbon dioxide release per unit time, real-time pH value and the rate of change of agricultural product weight from t1 to t2, the corresponding agricultural product spoilage index is output.

5. The method for automated intelligent palletizing of agricultural products according to claim 4, characterized in that: In step 3, the calculation process of the air circulation index of the agricultural product box includes: According to the air circulation characteristics, weights are assigned to the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box; Based on the assigned weights, combined with the weight summation method, the air ventilation index of the agricultural product box is calculated. The calculation process is as follows: Y=w1y1+w2y2+w3y3+w4y4+w5y5+w6y6+w7y7 Among them, Y is the air circulation index of the agricultural product box, w1, w2, w3, w4, w5, w6 and w7 are the weights of the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box, respectively, and y1, y2, y3, y4, y5, y6 and y7 are the temperature, humidity, oxygen concentration, carbon dioxide concentration, ethylene concentration, wind speed and air flow of the agricultural product box, respectively.

6. The method for automated intelligent palletizing of agricultural products according to claim 5, characterized in that: In step 4, the calculation process of the pallet wear assessment data includes: The pallet wear assessment data includes the length change rate, width change rate, thickness change rate, height change rate and weight change rate of the pallet from t1 to t2; Select time points t1 and t2, calculate the length change rate, width change rate, thickness change rate, height change rate and weight change rate of the pallet from t1 to t2, and then obtain the pallet wear assessment data, and integrate the pallet wear assessment data into the intelligent palletizing data set.

7. The method for automated intelligent palletizing of agricultural products according to claim 6, characterized in that: In step 4, the process of constructing a pallet wear assessment model and obtaining a pallet wear coefficient includes: The pallet wear assessment data, anti-slip coefficient, and number of uses of the pallet are extracted from the intelligent palletizing dataset. The training set data is combined with the neural network algorithm. The pallet wear assessment data, anti-slip coefficient, and number of uses of the pallet are used as input, and the pallet wear coefficient is used as output. The nonlinear relationship between the pallet wear assessment data, the anti-slip coefficient, the number of uses, and the pallet wear coefficient is learned to train the pallet wear assessment model. Input the test set data into the pallet wear assessment model, adjust the pallet wear assessment model parameters, optimize the pallet wear assessment model, deploy the pallet wear assessment model into the system, and combine the current pallet wear assessment data as well as the pallet's anti-slip coefficient and number of uses to output the corresponding pallet wear coefficient.

8. The method for automated intelligent palletizing of agricultural products according to claim 7, characterized in that: In step 5, the construction process of the palletizing accuracy evaluation model includes: The agricultural product spoilage index, the air circulation index of the agricultural product box, and the pallet wear coefficient are integrated to generate a palletizing accuracy evaluation dataset, which is then divided into a training set and a test set. Combining training set data with a multivariate linear regression algorithm, the agricultural product spoilage index, the air circulation index of the agricultural product box, and the pallet wear coefficient are used as input, and the agricultural product palletizing accuracy is used as output. The linear relationship between the agricultural product spoilage index, the air circulation index of the agricultural product box, the pallet wear coefficient, and the agricultural product palletizing accuracy is learned to train a palletizing accuracy evaluation model. Input the test set data into the palletizing accuracy evaluation model, adjust the regression coefficient and intercept term of the palletizing accuracy evaluation model, optimize the performance of the palletizing accuracy evaluation model, and obtain the final palletizing accuracy evaluation model. The expression of the palletizing accuracy evaluation model is as follows: T=β1H+β2Y+β3M Among them, T is the palletizing accuracy of agricultural products, β1, β2 and β3 are the regression coefficients of the agricultural product spoilage index, the air circulation index of the agricultural product box and the pallet wear coefficient respectively, H, Y and M are the agricultural product spoilage index, the air circulation index of the agricultural product box and the pallet wear coefficient respectively; Combined with the current agricultural product decay index, the air circulation index of the agricultural product box and the pallet wear coefficient, the corresponding agricultural product palletizing accuracy is output.

9. The method for automated intelligent palletizing of agricultural products according to claim 8, characterized in that: In step 5, the process of adjusting the stacking accuracy of agricultural products includes: When the palletizing accuracy of agricultural products is lower than 0.3, it indicates that the agricultural products are abnormally spoiled. A spoilage index threshold for agricultural products is set. When the spoilage index of the corresponding agricultural products is lower than the spoilage index threshold, the temperature, humidity, oxygen concentration, and carbon dioxide concentration of the palletizing environment are adjusted to slow down the spoilage trend of the agricultural products. Agricultural products with spoilage index higher than the spoilage index threshold are removed. When the palletizing accuracy of agricultural products is between 0.3 and 0.6, it indicates that there is abnormal air circulation in the agricultural product box. A threshold value of the air circulation index of the agricultural product box is set. When the air circulation index of the corresponding agricultural product box is lower than the threshold value of the air circulation index of the agricultural product box, the changes in the air circulation index of the agricultural product box are continuously monitored to optimize the air circulation in the agricultural product box. For agricultural product boxes with an air circulation index higher than the threshold value of the agricultural product box, more air holes are added to the corresponding agricultural product box and the gaps between the agricultural product boxes are increased. When the palletizing accuracy of agricultural products is greater than 0.6, it indicates that the pallet has abnormal wear. A pallet wear coefficient threshold is set. When the pallet wear coefficient is lower than the pallet wear coefficient threshold, the change of the pallet wear coefficient is continuously monitored. When the pallet wear coefficient is higher than the pallet wear coefficient threshold, the worn parts of the pallet are checked, and minor wear is repaired. For severe wear, the pallet is replaced.

10. The method for automated intelligent palletizing of agricultural products according to claim 9, characterized in that: In step 6, based on the agricultural product palletizing accuracy adjustment result, combined with the agricultural product box data, box transfer data and whole layer dense arrangement data, the process of intelligent palletizing of agricultural products includes: S1. Compare the total mass of the agricultural products with the maximum load of the agricultural product box, adjust the total mass of the agricultural products to a state less than the maximum load of the agricultural product box, and transport them to the merging main line via the roller conveyor line; S2. The belt line transports the agricultural product boxes to the weighing machine and labeling machine. The weighing machine checks the weight of each agricultural product box. The labeling machine assigns a label to the agricultural product box with the box number. The edge roller line adjusts the position of the agricultural product box, and then completes the agricultural product box code scanning work. Based on the code scanning results, unqualified agricultural product boxes are rejected; S3. Based on the agricultural product palletizing method, a placement angle threshold for the agricultural product boxes is set. The placement of qualified agricultural product boxes is evaluated based on the real-time placement angle of the agricultural product boxes, and the agricultural product boxes are then transferred using a transfer cylinder. S4. Arrange the agricultural product boxes after the box transfer process on the power roller line in the agricultural product palletizing manner using the side roller line; S5. Push the arranged agricultural product boxes to the roller line through a continuous box pusher, set a gap threshold for the agricultural product boxes, and adjust the gaps between the boxes accordingly based on the gaps between the boxes, thereby achieving dense arrangement of the boxes in a whole layer. S6. Use the whole layer pushing line to push the densely packed agricultural product boxes to the palletizing line; S7, conveying the empty pallet to the palletizing elevator through the pallet conveyor line, scanning the empty pallet barcode, and recording the pallet information, which is the pallet data; S8. Control the opening and closing of the palletizing line, place the entire layer of agricultural product boxes on the empty pallet, complete the palletizing operation, and then bind the agricultural product box code with the pallet barcode; S9. After palletizing, the pallet transmits the bound agricultural product box code to the client system through the programmable logic controller to complete the information exchange. The pallet carrying the whole layer of agricultural product boxes is transported to the wrapping machine through the pallet conveyor to wrap the palletized products with protective film. Finally, the agricultural products after film wrapping are transported to the warehouse for storage.

Images