A management planning method for airport baggage

Abstract

The invention relates to the technical field of management planning, and discloses a management planning method for airport luggage. The method comprises the following steps: obtaining perspective views of several airport luggage, and dividing airport luggage boxes into feature categories according to the perspective views; generating N transportation management plans for airport luggage of N flights within a first preset time period according to feature classification initialization; collecting wear information of airport luggage of N flights within the first preset time period; evaluating corresponding transportation management plans based on the wear information to obtain luggage wear coefficients of the transportation management plans, and training a transportation management model based on the N flight transportation management plans and the corresponding luggage wear coefficients; and obtaining a target management plan for a target flight within a second preset time period based on the transportation management model and an optimization algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of management planning, and more specifically, to a management planning method for airport luggage. Background Art

[0002] With the rapid development of the aviation transportation industry, the amount of luggage transported at airports has continued to increase. The traditional way to manage luggage at airports is to cover the outside of the luggage with plastic inflatable films of varying thicknesses to reduce surface scratches, dents, and even damage caused by luggage collisions during transportation. However, the existing airport luggage management method lacks refined management of luggage transportation. The existing luggage transportation management is mainly based on a unified process and fails to be dynamically optimized according to the type of luggage. This results in insufficient protection for some fragile luggage, while unnecessary high-cost protection measures are taken for some luggage that is not easily damaged, increasing the cost of luggage management. Summary of the invention

[0003] The present invention provides a management and planning method for airport luggage, which solves the technical problems raised in the background technology.

[0004] The present invention provides a method for airport baggage management planning, comprising:

[0005] Step 1, obtaining perspective views of several airport luggages, and classifying the airport luggages into feature categories according to the perspective views; wherein the feature categories include: rigid shell category and flexible shell category; the flexible shell category includes: flexible shell edge rigid category and flexible shell edge flexible category;

[0006] Step 2, for the airport baggage of N flights within the first preset time period, N transportation management plans are generated according to the feature classification initialization; the transportation management plan indicates that the flexible airport baggage of the flexible shell edge is used to protect the remaining airport baggage of the feature category during transportation;

[0007] Step 3, collecting the wear information of the airport luggage of N flights within the first preset time period; the wear information includes: the number of scratches added to the airport luggage of each flight, the area of ​​each added scratch, the number of airport luggage and the satisfaction score of the passengers on the corresponding flights for the luggage check-in;

[0008] Step 4: Evaluate the corresponding transport management scheme based on the wear information to obtain the baggage wear coefficient of the transport management scheme, and train a transport management model based on the N voyage transport management schemes and the corresponding baggage wear coefficients; the output of the transport management model represents the predicted baggage wear coefficient of any transport management scheme within the second preset time period;

[0009] Step 6: within the second preset time period, obtain a target management plan for the target voyage based on the transportation management model and the optimization algorithm.

[0010] Furthermore, the airport luggage is classified into feature categories according to the perspective view, including:

[0011] Step 21, performing an X-ray scan on the airport luggage through a security inspection machine to obtain a perspective view of the airport luggage, and pre-processing the perspective view to obtain a standard view;

[0012] Step 22, if the outermost edge in the standard graph is blue, and the remaining edges are all inside the outermost edge, the corresponding airport luggage is classified as a rigid shell class, otherwise it is classified as a flexible shell class;

[0013] Step 23, for airport luggage with flexible shells, the standard graph is divided into a central area and K edge sub-areas, and the edge softness of the airport luggage is calculated. The calculation formula of the edge softness is as follows:

[0014] ;

[0015] ;

[0016] in, Indicates the softness of the edge of airport luggage. represents the number of edge sub-regions, represents the i-th edge sub-region, represents the maximum distance between the ith edge sub-region and the central region, represents the number of edge pixels in the ith edge sub-region, represents the first weight parameter, represents the number of yellow pixels in the i-th edge sub-region, represents the second weight parameter, represents the number of green pixels in the i-th edge sub-region, represents the third weight parameter, represents the number of blue pixels in the i-th edge sub-region, represents the central rigidity adjustment function, represents the adjustment factor, , Indicates the number of blue pixels in the center area, Indicates the number of yellow pixels in the center area, ,and ;

[0017] Step 24: if the edge softness is greater than or equal to the preset softness threshold, the corresponding airport luggage is classified as a flexible shell edge soft class; if the edge softness is less than the preset softness threshold, the corresponding airport luggage is classified as a flexible shell edge rigid class.

[0018] Furthermore, the perspective image is preprocessed, including:

[0019] The perspective images include: a primary perspective image and a secondary perspective image; wherein the primary perspective image represents the first perspective image taken when the airport luggage is placed in the security inspection machine, and the secondary perspective image represents the second perspective image taken when the airport luggage is bumped once in the security inspection machine; both the primary perspective image and the secondary perspective image are images including characteristic edges of three colors: yellow, green and blue;

[0020] An edge threshold is set. If the number of pixels corresponding to the feature edge is greater than the edge threshold, the corresponding feature edge is shielded; if the number of pixels corresponding to the feature edge is less than or equal to the edge threshold, the corresponding feature edge is retained; a first preprocessed image and a second preprocessed image corresponding to the main perspective image and the secondary perspective image are obtained; a first difference image of the first preprocessed image and the second preprocessed image is calculated; and a second difference image of the main perspective image and the first difference image is calculated, and the second difference image is used as the corresponding standard image of the airport baggage.

[0021] Furthermore, N transportation management solutions are generated based on the feature classification initialization, including:

[0022] Based on the random generation of a transportation management plan for baggage at several airports of a flight that meets the constraints, the transportation management plan is as follows:

[0023] Plastic inflatable films include: large, medium and small, with corresponding edge softness , and ;

[0024] The constraint conditions include: if the two sides of any rigid-shell airport luggage are flexible-shell edges, then plastic inflatable film shall not be used; if any side of any rigid-shell airport luggage is not flexible-shell edges, then plastic inflatable film shall be used;

[0025] If the two sides of any flexible outer shell edge rigid luggage are flexible outer shell edge flexible luggage, no plastic inflatable film shall be provided; if any side of any flexible outer shell edge rigid luggage is not flexible outer shell edge flexible luggage, plastic inflatable film shall be provided;

[0026] Transport Management Solutions The encoding is: ;in, Indicates the number of Airport luggage, , Indicates the order in which luggage is placed at the airport. Indicates the number of Each airport luggage will be randomly covered with any large, medium or small plastic inflatable film.

[0027] Furthermore, the calculation formula of the luggage wear coefficient of the transportation management solution is as follows:

[0028] ;in, represents the baggage wear coefficient of the transportation management plan for the i-th voyage, represents the number of airport luggage of the flexible class with flexible shell edge on the ith flight, represents the total amount of airport baggage for the ith flight, express The index of represents the number of new scratches on luggage at the ath airport in the i-th flight, represents the area of ​​the newly added scratches on the luggage at the ath airport in the i-th flight, It represents the passenger satisfaction score corresponding to the baggage at the ath airport in the i-th flight.

[0029] Furthermore, the transport management model includes:

[0030] The codes corresponding to N transport management schemes are used as training samples, and the baggage wear coefficients of the corresponding transport management schemes are used as sample labels to train a transport management model;

[0031] The transportation management model includes: input layer, hidden layer and output layer;

[0032] The input layer is used to input the code corresponding to the transportation management plan;

[0033] Hidden layer, based on mapping the encoding to obtain the hidden state of the encoding;

[0034] The output layer is configured with a classifier, and the classification space of the classifier represents the predicted baggage wear coefficient of the transportation management solution corresponding to the hidden state;

[0035] Among them, the hidden layer is constructed based on a one-dimensional convolutional neural network with the ReLU activation function; the classifier is a fully connected classifier; based on the mean square error of the predicted luggage wear coefficient and the corresponding sample label, the weight parameters and bias parameters of the hidden layer are updated and optimized through the back propagation algorithm.

[0036] Furthermore, the target management plan for the target voyage is obtained, including:

[0037] Step 71, obtaining characteristic classification of airport baggage of the target flight within a second preset time period;

[0038] Step 72, based on the characteristic classification of the target voyage, initialize and generate several transportation management plans that meet the constraint conditions, and the codes corresponding to the transportation management plans;

[0039] Step 73, obtaining the predicted luggage wear coefficient of each transportation management scheme based on the transportation management model, and obtaining the cost of the plastic inflatable film installed in the transportation management scheme, so as to obtain the fitness function value of each transportation management scheme;

[0040] Step 74, based on the fitness function value of each transportation management plan, sort the transportation management plans from small to large to obtain a feature sorting; retain a preset number of transportation management plans from the front to the back of the feature sorting, and update the remaining transportation management plans by random self-crossing;

[0041] Step 75, repeat step 74 for a preset number of times, output the transportation management plan with the smallest fitness function value, and use the transportation management plan with the smallest fitness function value as the target management plan for the target voyage.

[0042] Furthermore, the calculation formula of the fitness function is as follows:

[0043] ;

[0044] in, represents the fitness function value of the j-th transportation management plan for the target voyage, represents the predicted baggage wear coefficient of the j-th transportation management plan for the target voyage, represents the cost of the plastic inflatable film for the j-th transportation management plan of the target voyage, represents the first fitness weight, represents the second fitness weight, and All are not 0, , .

[0045] The beneficial effect of the present invention is that by constructing an intelligent transportation management system, combining luggage feature classification, luggage wear coefficient prediction and dynamic optimization algorithm, the refinement of airport luggage transportation management is significantly improved. Compared with the traditional transportation management method that relies on physical protection means, the present invention takes the optimization of the management level as the core, uses data-driven methods to dynamically generate transportation plans that meet the constraints, and provides a scientific decision-making basis by comprehensively evaluating the luggage protection effect and resource usage costs. Finally, the transportation management plan achieves a balance between luggage protection, resource conservation and environmental benefits, reducing operating costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1It is a flow chart of a method for managing and planning airport baggage of the present invention. DETAILED DESCRIPTION

[0047] The subject matter described herein will now be discussed with reference to example embodiments. It should be understood that the discussion of these embodiments is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the contents of this specification. Each example may omit, replace or add various processes or components as needed. In addition, the features described relative to some examples may also be combined in other examples.

[0048] like Figure 1 As shown, a management planning method for airport baggage includes:

[0049] Step 1, obtaining perspective views of several airport luggages, and classifying the airport luggages into feature categories according to the perspective views; wherein the feature categories include: rigid shell category and flexible shell category; the flexible shell category includes: flexible shell edge rigid category and flexible shell edge flexible category;

[0050] Step 2, for the airport baggage of N flights within the first preset time period, N transportation management plans are generated according to the feature classification initialization; the transportation management plan indicates that the flexible airport baggage of the flexible shell edge is used to protect the remaining airport baggage of the feature category during transportation;

[0051] Step 3, collecting the wear information of the airport luggage of N flights within the first preset time period; the wear information includes: the number of scratches added to the airport luggage of each flight, the area of ​​each added scratch, the number of airport luggage and the satisfaction score of the passengers on the corresponding flights for the luggage check-in;

[0052] Step 4: Evaluate the corresponding transport management scheme based on the wear information to obtain the baggage wear coefficient of the transport management scheme, and train a transport management model based on the N voyage transport management schemes and the corresponding baggage wear coefficients; the output of the transport management model represents the predicted baggage wear coefficient of any transport management scheme within the second preset time period;

[0053] Step 6: within the second preset time period, obtain a target management plan for the target voyage based on the transportation management model and the optimization algorithm.

[0054] In one embodiment of the present invention, airport luggage is classified into feature categories according to perspective, including:

[0055] Step 21, performing an X-ray scan on the airport luggage through a security inspection machine to obtain a perspective view of the airport luggage, and pre-processing the perspective view to obtain a standard view;

[0056] Step 22, if the outermost edge in the standard graph is blue, and the remaining edges are all inside the outermost edge, the corresponding airport luggage is classified as a rigid shell class, otherwise it is classified as a flexible shell class;

[0057] Step 23, for airport luggage with flexible shells, the standard graph is divided into a central area and K edge sub-areas, and the edge softness of the airport luggage is calculated. The calculation formula of the edge softness is as follows:

[0058] ;

[0059] ;

[0060] in, Indicates the softness of the edge of airport luggage. represents the number of edge sub-regions, represents the ith edge sub-region, represents the maximum distance between the ith edge sub-region and the central region, represents the number of edge pixels in the ith edge sub-region, represents the first weight parameter, represents the number of yellow pixels in the i-th edge sub-region, represents the second weight parameter, represents the number of green pixels in the i-th edge sub-region, represents the third weight parameter, represents the number of blue pixels in the i-th edge sub-region, represents the central rigidity adjustment function, represents the adjustment factor, , Indicates the number of blue pixels in the center area, Indicates the number of yellow pixels in the center area, ,and ;

[0061] Step 24: if the edge softness is greater than or equal to the preset softness threshold, the corresponding airport luggage is classified as a flexible shell edge soft class; if the edge softness is less than the preset softness threshold, the corresponding airport luggage is classified as a flexible shell edge rigid class.

[0062] It should be noted that both soft packages and hard packages are delivered to a predetermined location for installation via a distribution transmission device (such as a conveyor belt).

[0063] Airport luggage can usually be divided into soft bags (travel bags) and hard bags (suitcases), and the corresponding categories are flexible shell and rigid shell. The management process of airport luggage mainly depends on the accurate identification and classification of luggage characteristics and its transportation environment. In the present invention, the classification of airport luggage is based on the X-ray perspective of the security inspection machine. The X-ray perspective shows the difference in absorbance due to the different density of the object, which is respectively expressed as blue (high density, corresponding to rigid objects), green (medium density, corresponding to harder flexible objects) and yellow (low density, corresponding to flexible objects). On this basis, for soft baggage (flexible shell luggage), if the edge softness in its perspective is greater than or equal to the preset softness threshold (edge ​​softness is a quantitative indicator based on pixel features), this type of luggage is further divided into a flexible shell edge soft class. This type of luggage can be used as a buffer luggage due to its high softness to reduce the collision and wear of rigid luggage (hard bag luggage). During transportation, the flexible shell edge soft class luggage is preferentially arranged around the rigid shell luggage (hard bag luggage) to avoid direct collision between rigid luggage. If the edge softness of the flexible shell luggage is less than the preset softness threshold, the luggage is classified as the flexible shell edge rigid type. This type of luggage has poor softness and is not suitable as a cushioning luggage. It requires additional protection measures, such as using a plastic inflatable film for shock absorption. When the number of flexible shell edge flexible luggage in the transportation plan is insufficient, the two sides of the rigid shell luggage may lack effective cushioning protection. In order to avoid scratches, dents or damage on the luggage surface, a dynamic compensation strategy of the prior art is adopted: by covering the outer side of the luggage with a plastic inflatable film for shock absorption to fill the cushioning gap. The number and specifications (large, medium, small) of the inflatable film are dynamically adjusted according to the actual position and edge softness characteristics of different luggage in the transportation plan.

[0064] In one embodiment of the present invention, preprocessing the perspective image includes:

[0065] The perspective images include: a primary perspective image and a secondary perspective image; wherein the primary perspective image represents the first perspective image taken when the airport luggage is placed in the security inspection machine, and the secondary perspective image represents the second perspective image taken when the airport luggage is bumped once in the security inspection machine; both the primary perspective image and the secondary perspective image are images including characteristic edges of three colors: yellow, green and blue;

[0066] An edge threshold is set. If the number of pixels corresponding to the feature edge is greater than the edge threshold, the corresponding feature edge is shielded; if the number of pixels corresponding to the feature edge is less than or equal to the edge threshold, the corresponding feature edge is retained; a first preprocessed image and a second preprocessed image corresponding to the main perspective image and the secondary perspective image are obtained; a first difference image of the first preprocessed image and the second preprocessed image is calculated; and a second difference image of the main perspective image and the first difference image is calculated, and the second difference image is used as the corresponding standard image of the airport baggage.

[0067] In the process of managing airport baggage transportation, the feature classification of baggage is one of the core steps to optimize the transportation plan. However, when processing the X-ray perspective of airport baggage, the existing classification technology has low classification efficiency and limited accuracy due to the characteristics of the baggage itself (such as the wrinkled edges of soft bags). These redundant or tiny edges increase unnecessary computational complexity, resulting in slow classification speed. By designing a bumping mechanism in the security inspection machine, the airport baggage will vibrate to a certain extent during the X-ray perspective process, so as to obtain more accurate feature extraction results. By designing a slope that bends up and down on the internal conveyor belt of the security inspection machine, the slope is used to make the airport baggage vibrate to a certain extent when passing through the security inspection. The vibration causes the loose parts of the baggage surface (such as the wrinkled edges of soft bags) to deform, resulting in changes in these irrelevant edge information in the X-ray perspective. Through the difference in deformation before and after vibration, irrelevant edge features can be accurately removed, thereby accelerating the classification process. The irrelevant edge information is removed by differential processing, reducing the amount of calculation in the classification process. The interference caused by the wrinkled edges of soft baggage is filtered out, the core density distribution characteristics of the baggage are retained, and the accuracy of the classification results is significantly improved. The turbulence mechanism and differential processing method of the present invention solve the problems of low efficiency and high interference in the traditional airport luggage sorting process, and at the same time improve the intelligent level of luggage protection management through accurate feature classification.

[0068] In one embodiment of the present invention, N transportation management solutions are generated based on feature classification initialization, including:

[0069] Based on the random generation of a transportation management plan for baggage at several airports of a flight that meets the constraints, the transportation management plan is as follows:

[0070] Plastic inflatable films include: large, medium and small, with corresponding edge softness , and ;

[0071] The constraint conditions include: if the two sides of any rigid-shell airport luggage are flexible-shell edges, then plastic inflatable film shall not be used; if any side of any rigid-shell airport luggage is not flexible-shell edges, then plastic inflatable film shall be used;

[0072] If the two sides of any flexible outer shell edge rigid luggage are flexible outer shell edge flexible luggage, no plastic inflatable film shall be provided; if any side of any flexible outer shell edge rigid luggage is not flexible outer shell edge flexible luggage, plastic inflatable film shall be provided;

[0073] Transport Management Solutions The encoding is: ;in, Indicates the number of Airport luggage, , Indicates the order in which luggage is placed at the airport. Indicates the number of Each airport luggage will be randomly covered with any large, medium or small plastic inflatable film.

[0074] Specifically, through the reasonable combination of flexible luggage and plastic film, the protection needs of different types of luggage can be met. Rigid shell luggage avoids scratches or damage caused by direct collision. The softness of the edges of flexible shell luggage is fully utilized to achieve natural buffer protection. The amount of plastic inflatable film used is dynamically adjusted according to the classification and location of the luggage. Flexible shell edge soft luggage is used first in high-risk areas and does not require additional film protection. Plastic inflatable film is only used when necessary to avoid waste of resources. By automatically generating transportation management plans, the complexity of manual intervention is reduced. After the plan is optimized, the loading and protection steps of luggage are more efficient, which improves the overall efficiency of transportation. And ultimately provides training data support for the transportation management plan.

[0075] In one embodiment of the present invention, the calculation formula of the luggage wear coefficient of the transportation management solution is as follows:

[0076] ;in, represents the baggage wear coefficient of the transportation management plan for the i-th voyage, represents the number of airport luggage of flexible class with flexible shell edge for the ith flight, represents the total amount of airport baggage for the ith flight, express The index of represents the number of new scratches on luggage at the ath airport in the i-th flight, represents the area of ​​the newly added scratches on the luggage at the ath airport in the i-th flight, It represents the passenger satisfaction score corresponding to the baggage at the ath airport in the i-th flight.

[0077] Specifically, the luggage wear coefficient of the transportation management plan is used to measure the effectiveness of luggage protection during transportation. A scientific evaluation system was established by comprehensively considering the number of soft luggage on the edge of the flexible shell, the area of ​​new scratches on the luggage, the total area of ​​the luggage, and the passenger satisfaction score. Among them, the number of flexible luggage reflects the ability of buffering protection, the ratio of the area of ​​new scratches to the total area of ​​luggage reflects the actual wear, and passenger satisfaction further verifies the comprehensive performance of the transportation plan. Through the calculation of this coefficient, the advantages and disadvantages of the transportation management plan for each voyage can be accurately evaluated, which can not only help dynamically optimize the design of the plan, but also achieve a balance between protection effect and resource cost, ultimately improving transportation efficiency and reducing costs, while taking into account environmental protection and user experience. And ultimately provide training data support for the transportation management plan.

[0078] In one embodiment of the present invention, the transportation management model includes:

[0079] The codes corresponding to N transport management schemes are used as training samples, and the baggage wear coefficients of the corresponding transport management schemes are used as sample labels to train a transport management model;

[0080] The transportation management model includes: input layer, hidden layer and output layer;

[0081] The input layer is used to input the code corresponding to the transportation management plan;

[0082] Hidden layer, based on mapping the encoding to obtain the hidden state of the encoding;

[0083] The output layer is configured with a classifier, and the classification space of the classifier represents the predicted baggage wear coefficient of the transportation management solution corresponding to the hidden state;

[0084] Among them, the hidden layer is constructed based on a one-dimensional convolutional neural network with the ReLU activation function; the classifier is a fully connected classifier; based on the mean square error of the predicted luggage wear coefficient and the corresponding sample label, the weight parameters and bias parameters of the hidden layer are updated and optimized through the back propagation algorithm.

[0085] Specifically, by constructing a transportation management model based on a one-dimensional convolutional neural network, the encoding of the transportation management plan can be efficiently mapped to a hidden state, and the luggage wear coefficient can be accurately predicted through a fully connected classifier. The hidden layer uses a one-dimensional convolutional neural network with a ReLU activation function, which can efficiently extract nonlinear features in the transportation plan encoding and improve the model's ability to express complex data. The output layer optimizes the hidden layer weights and bias parameters through the mean square error to ensure the model's prediction accuracy and further improve the optimization ability of the transportation management plan. Compared with the traditional solution evaluation method that relies on manual experience, this model realizes the automation and efficiency of the transportation management solution evaluation, provides a scientific basis for the dynamic adjustment of the transportation plan, and greatly reduces the evaluation time cost, improving transportation efficiency and reliability.

[0086] In one embodiment of the present invention, obtaining a target management plan for a target voyage includes:

[0087] Step 71, obtaining characteristic classification of airport baggage of the target flight within a second preset time period;

[0088] Step 72, based on the characteristic classification of the target voyage, initialize and generate several transportation management plans that meet the constraint conditions, and the codes corresponding to the transportation management plans;

[0089] Step 73, obtaining the predicted luggage wear coefficient of each transportation management scheme based on the transportation management model, and obtaining the cost of the plastic inflatable film installed in the transportation management scheme, so as to obtain the fitness function value of each transportation management scheme;

[0090] Step 74, based on the fitness function value of each transportation management plan, sort the transportation management plans from small to large to obtain a feature sorting; retain a preset number of transportation management plans from the front to the back of the feature sorting, and update the remaining transportation management plans by random self-crossing;

[0091] Step 75, repeat step 74 for a preset number of times, output the transportation management plan with the smallest fitness function value, and use the transportation management plan with the smallest fitness function value as the target management plan for the target voyage.

[0092] Specifically, by optimizing and generating the transportation management plan for the target voyage, the problems of inaccurate protection plan, waste of resources and low optimization efficiency in the traditional baggage transportation management process are effectively solved. The optimization process of this embodiment is based on the iterative optimization of baggage feature classification, management model prediction and fitness function, and finally outputs an efficient transportation management plan suitable for the target voyage. First, the airport baggage of the target voyage is analyzed by feature classification, and the flexible shell and rigid shell characteristics of the baggage are accurately extracted from the feature classification data obtained in the second preset time period, and multiple transportation management plans and their corresponding codes are initialized and generated in combination with the constraints of the actual transportation scene. This process provides a data basis for the subsequent solution optimization, so that the management plan can accurately match the baggage characteristics and transportation needs of the target voyage. Secondly, the transportation management model is used to quickly calculate the predicted baggage wear coefficient and plastic inflatable film cost of each plan, and the fitness function is constructed to comprehensively evaluate the advantages and disadvantages of each plan. The transportation management model combines the efficient mapping ability of the one-dimensional convolutional neural network to the solution coding, and can complete the accurate prediction of all transportation plans in a short time, greatly improving the efficiency and accuracy of the transportation plan evaluation, and at the same time, the fitness function balances the relationship between the baggage protection effect and the resource use cost. Then, through the update mechanism based on the sorting of fitness function and random self-crossover, inferior solutions are continuously eliminated, high-quality solutions are retained, and new transportation management solutions are generated in the iterative process. This optimization mechanism gradually approaches the global optimal solution in multiple iterations, ensuring that the final output target management solution has the lowest fitness function value, which can not only minimize the luggage wear coefficient, but also significantly reduce the cost of using plastic inflatable film. Finally, the optimization process uses the transportation management solution with the smallest fitness function value as the final output of the target voyage, ensuring the comprehensive balance of the transportation management solution in terms of luggage protection effect, cost optimization and solution reliability. Compared with traditional methods, the optimization process of the present invention has the significant advantages of automation, efficiency and intelligence, and provides a systematic optimization solution for luggage transportation management, which can adapt to different voyages and luggage feature scenarios, improve transportation efficiency, reduce resource consumption, and meet environmental protection requirements.

[0093] In one embodiment of the present invention, the calculation formula of the fitness function is as follows:

[0094] ;

[0095] in, represents the fitness function value of the j-th transportation management plan for the target voyage, represents the predicted baggage wear coefficient of the j-th transportation management plan for the target voyage, represents the cost of the plastic inflatable film for the j-th transportation management plan of the target voyage, represents the first fitness weight, represents the second fitness weight, and All are not 0, , .

[0096] Specifically, a calculation method based on fitness function is proposed to comprehensively evaluate the advantages and disadvantages of transportation management schemes. The fitness function combines the predicted luggage wear coefficient and the use cost of plastic inflatable film, and dynamically weights it through weight coefficients to ensure that the evaluation results can accurately reflect the overall performance of the transportation management scheme, thereby providing a scientific basis for the optimization and selection of the scheme. The significant advantage of the fitness function is that it realizes the comprehensive optimization of multiple objectives. By considering the luggage wear coefficient and the cost of the inflatable film at the same time, the present invention effectively solves the contradiction between the protection effect and the waste of resources in the traditional method. The smaller the value of the fitness function, the stronger the protection ability of the scheme and the lower the cost, thereby guiding the optimization algorithm to select a better transportation management scheme. This method not only improves the scientificity and accuracy of the evaluation, but also provides a reliable basis for the actual deployment of the transportation management scheme, taking into account the needs of luggage protection, cost savings and environmental friendliness.

[0097] The above describes an embodiment of the present embodiment, but the present embodiment is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present embodiment, ordinary technicians in this field can also make many forms, all of which are protected by the present embodiment.

Claims

1. A method for airport baggage management planning, characterized in that: include: Step 1, obtaining perspective views of several airport luggages, and classifying the airport luggages into feature categories according to the perspective views; wherein the feature categories include: rigid shell category and flexible shell category; the flexible shell category includes: flexible shell edge rigid category and flexible shell edge flexible category; Step 2, for the airport baggage of N flights within the first preset time period, N transportation management plans are generated according to the feature classification initialization; the transportation management plan indicates that the flexible airport baggage of the flexible shell edge is used to protect the remaining airport baggage of the feature category during transportation; Step 3, collecting the wear information of the airport luggage of N flights within the first preset time period; the wear information includes: the number of scratches added to the airport luggage of each flight, the area of ​​each added scratch, the number of airport luggage and the satisfaction score of the passengers on the corresponding flights for the luggage check-in; Step 4: Evaluate the corresponding transport management scheme based on the wear information to obtain the baggage wear coefficient of the transport management scheme, and train a transport management model based on the N voyage transport management schemes and the corresponding baggage wear coefficients; the output of the transport management model represents the predicted baggage wear coefficient of any transport management scheme within the second preset time period; Step 5: within the second preset time period, based on the transportation management model and the optimization algorithm, obtain a target management plan for the target voyage.

2. The airport baggage management planning method according to claim 1, characterized in that: Airport luggage is categorized into characteristic categories based on perspective, including: Step 11, performing an X-ray scan on the airport luggage through a security inspection machine to obtain a perspective view of the airport luggage, and pre-processing the perspective view to obtain a standard view; Step 12: if the outermost edge in the standard graph is blue and the remaining edges are all inside the outermost edge, the corresponding airport luggage is classified as a rigid shell class, otherwise it is classified as a flexible shell class; Step 13: For airport luggage with flexible shells, the standard graph is divided into a central area and K edge sub-areas, and the edge softness of the airport luggage is calculated. The calculation formula of the edge softness is as follows: ； ； in, Indicates the softness of the edge of airport luggage. represents the number of edge sub-regions, represents the i-th edge sub-region, represents the maximum distance between the ith edge sub-region and the central region, represents the number of edge pixels in the ith edge sub-region, represents the first weight parameter, represents the number of yellow pixels in the i-th edge sub-region, represents the second weight parameter, represents the number of green pixels in the i-th edge sub-region, represents the third weight parameter, represents the number of blue pixels in the i-th edge sub-region, represents the central rigidity adjustment function, represents the adjustment factor, , Indicates the number of blue pixels in the center area, Indicates the number of yellow pixels in the center area, ,and ; Step 14: if the edge softness is greater than or equal to a preset softness threshold, the corresponding airport luggage is classified as a flexible shell edge softness class; if the edge softness is less than the preset softness threshold, the corresponding airport luggage is classified as a flexible shell edge rigid class.

3. The airport baggage management planning method according to claim 2, characterized in that: Preprocess the perspective image, including: The perspective images include: a primary perspective image and a secondary perspective image; wherein the primary perspective image represents the first perspective image taken when the airport luggage is placed in the security inspection machine, and the secondary perspective image represents the second perspective image taken when the airport luggage is bumped once in the security inspection machine; both the primary perspective image and the secondary perspective image are images including characteristic edges of three colors: yellow, green and blue; An edge threshold is set. If the number of pixels corresponding to the feature edge is greater than the edge threshold, the corresponding feature edge is shielded; if the number of pixels corresponding to the feature edge is less than or equal to the edge threshold, the corresponding feature edge is retained; a first preprocessed image and a second preprocessed image corresponding to the main perspective image and the secondary perspective image are obtained; a first difference image of the first preprocessed image and the second preprocessed image is calculated; and a second difference image of the main perspective image and the first difference image is calculated, and the second difference image is used as the corresponding standard image of the airport baggage.

4. The airport baggage management planning method according to claim 2, characterized in that: N transportation management solutions are generated based on feature classification initialization, including: Based on the random generation of a transportation management plan for baggage at several airports of a flight that meets the constraints, the transportation management plan is as follows: Plastic inflatable films include: large, medium and small, with corresponding edge softness , and ; The constraint conditions include: if the two sides of any rigid-shell airport luggage are flexible-shell edges, then plastic inflatable film shall not be used; if any side of any rigid-shell airport luggage is not flexible-shell edges, then plastic inflatable film shall be used; If the two sides of any flexible outer shell edge rigid luggage are flexible outer shell edge flexible luggage, no plastic inflatable film shall be provided; if any side of any flexible outer shell edge rigid luggage is not flexible outer shell edge flexible luggage, plastic inflatable film shall be provided; Transport Management Solutions The encoding is: ;in, Indicates the number of the wth voyage Airport luggage, , Indicates the order in which luggage is placed at the airport. Indicates the number of Each airport luggage will be randomly covered with any large, medium or small plastic inflatable film.

5. The airport baggage management planning method according to claim 4, characterized in that: Transport management model, including: The codes corresponding to N transport management schemes are used as training samples, and the baggage wear coefficients of the corresponding transport management schemes are used as sample labels to train a transport management model; The transportation management model includes: input layer, hidden layer and output layer; The input layer is used to input the code corresponding to the transportation management plan; Hidden layer, based on mapping the encoding to obtain the hidden state of the encoding; The output layer is configured with a classifier, and the classification space of the classifier represents the predicted baggage wear coefficient of the transportation management solution corresponding to the hidden state; Among them, the hidden layer is constructed based on a one-dimensional convolutional neural network with the ReLU activation function; the classifier is a fully connected classifier; based on the mean square error of the predicted luggage wear coefficient and the corresponding sample label, the weight parameters and bias parameters of the hidden layer are updated and optimized through the back propagation algorithm.

6. The airport baggage management planning method according to claim 5, characterized in that: Obtain the target management plan for the target voyage, including: Step 51, obtaining characteristic classification of airport baggage of the target flight within a second preset time period; Step 52, based on the characteristic classification of the target voyage, initialize and generate several transportation management plans that meet the constraint conditions, and the codes corresponding to the transportation management plans; Step 53, obtaining the predicted luggage wear coefficient of each transportation management scheme based on the transportation management model, and obtaining the cost of the plastic inflatable film installed in the transportation management scheme, so as to obtain the fitness function value of each transportation management scheme; Step 54, based on the fitness function value of each transportation management plan, sort the transportation management plans from small to large to obtain a feature sorting; and retain a preset number of transportation management plans from the feature sorting from front to back, and update the remaining transportation management plans by random self-crossing; Step 55, repeat step 54 for a preset number of times, output the transportation management plan with the smallest fitness function value, and use the transportation management plan with the smallest fitness function value as the target management plan for the target voyage.

7. The airport baggage management planning method according to claim 6, characterized in that: The calculation formula of the fitness function is as follows: ； in, represents the fitness function value of the j-th transportation management plan for the target voyage, represents the predicted baggage wear coefficient of the j-th transportation management plan for the target voyage, represents the cost of the plastic inflatable film for the j-th transportation management plan of the target voyage, represents the first fitness weight, represents the second fitness weight, and All are not 0, , .

Images