Delivery network route optimization method and system

Abstract

The invention relates to a deliverable network route optimization method and system, and belongs to the technical field of communication operation, and the method comprises the steps: obtaining a deliverable network first route through employing an ant algorithm; key cross node parameters in the first route of the deliverable network are extracted, and semantics of the entity words in the context are obtained based on the key cross node parameters; and taking semantics as dynamic parameters, orderly carrying out federated learning according to the service association weights to obtain target semantics, and determining a deliverable network target route based on the target semantics. According to the method, the problem that an optimal detection agent deployment scheme is obtained according to a detection agent deployment strategy in the integrated delivery process is solved, and the detection agent deployment scheme is possibly locally optimal but not globally optimal.

Description

Technical Field

[0001] The present invention belongs to the field of communication operation technology, and in particular relates to a method and system for optimizing a delivery network route. Background Art

[0002] In the relevant technologies, in recent years, with the accumulation of big data and the rise of AI technology, the concept of intelligent operation and maintenance AIOps has gradually become a trend. However, it is a big challenge for enterprises to truly implement intelligence and achieve practical results. It is not simply chasing the "AIOps" trend, but encountering problems to be solved, and the basic data accumulation has begun to take shape, and the algorithm can basically meet the requirements, so we really take action to make some attempts. In order to minimize the overall communication overhead of resource detection, the deployment plan of the detection agent needs to determine how many detection agents are set in the underlying network and which underlying nodes are set as detection agents. Therefore, how to better implement the deployment plan of the detection agent has become an urgent problem to be solved. Summary of the invention

[0003] In view of the above deficiencies in the prior art, an object of the invention is to provide a method and system for optimizing a deliverable network route.

[0004] In a first aspect of the present invention, a method for optimizing a deliverable network route is proposed, comprising: S1, using an ant algorithm to obtain a first route of a deliverable network; S2, extracting key cross-node parameters in the first route of the deliverable network, and obtaining the semantics of entity words in a context based on the key cross-node parameters; S3, using the semantics as dynamic parameters to perform federated learning in order according to business association weights to obtain target semantics, and determining a target route of the deliverable network based on the target semantics.

[0005] Furthermore, an ant algorithm is used to obtain a first route of the deliverable network, including: determining an ant colony algorithm formula and initializing parameters of the ant colony algorithm formula; in the case of the first iteration, calculating a total distance of a path traveled by each ant and updating the pheromone of each generation of ants; while updating the pheromone of each generation of ants, continuing the iteration until the iteration ends, and when the iteration ends, selecting a path traveled by ants with the shortest total path distance; and using the path traveled by ants with the shortest total path distance as the first route of the deliverable network.

[0006] Furthermore, the ant colony algorithm formula includes:

[0007]

[0008] in, is the probability that ant k moves from city i to city j at time t; α is the pheromone heuristic factor, which reflects the effect of pheromone on ant path selection; β is the expected heuristic factor, which reflects the degree to which pheromone is valued when ants choose paths; τ ij (t) is the pheromone concentration on the path connecting city i and city j at time t; n ij (t) is the expected degree of ants moving from city i to city j, n ij (t) = 1 / dij, where dij is the distance between city i and city j; J k (i) is the city that ant k can currently choose; Update the pheromones of each generation of ants; among them, is the pheromone concentration released by the kth ant on the path connecting city i and city j, L k is the total distance of the path traveled by the kth ant, Q is the pheromone an ant has in its lifetime, Δτ ij is the sum of the pheromone concentrations released by ants on the path connecting city i and city j, m is the number of ants, and ρ represents the degree of pheromone dilution.

[0009] Furthermore, the semantics of the entity words in the context are obtained based on the key cross-node parameters, including: training the key cross-node parameters based on the ChatGPT model to obtain the semantics of the key cross-node parameters; and using the knowledge-enhanced KET model to analyze the semantics of the key cross-node parameters to obtain the semantics of the entity words in the context.

[0010] Further, according to the relevance coefficient rel of the entity word k =min-max(s k )*abs(cos(CR(X i ), c k ))c k , determine the association strength between the entity word and the context, where s k For entity word c k The confidence score of CR is the entity word c k The context representation corresponding to the sentence, X i The element currently being processed.

[0011] Furthermore, the semantics are used as dynamic parameters and federated learning is carried out in order according to the business association weights to obtain the target semantics, including: in the Nth training iteration of multiple terminals, based on the ChatGPT model, the semantics are used as dynamic parameters, and federated learning training is carried out according to the business association weights, and the training results are reported to the central server; the central server receives and aggregates the training results, and sends the global model and training configuration of the N+1th iteration to the terminal at the same time; the terminal trains the semantics based on the global model and the training configuration of the N+1th iteration to obtain the target semantics.

[0012] Further, determining a deliverable network target route based on the target semantics includes: acquiring a target database corresponding to semantics and routes; acquiring a route corresponding to the target semantics based on the target database, and using the route as the deliverable network target route.

[0013] In a second aspect of the present invention, a deliverable network route optimization system is proposed, including: an acquisition module, used to use an ant algorithm to acquire a first route of a deliverable network; a obtaining module, used to extract key cross-node parameters in the first route of the deliverable network, and obtain the semantics of entity words in the context based on the key cross-node parameters; a determination module, used to use the semantics as dynamic parameters to perform federated learning in order according to business association weights to obtain target semantics, and determine the target route of the deliverable network based on the target semantics.

[0014] According to a third aspect of the present invention, an electronic device is proposed, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any of the methods described in the first aspect of the present invention.

[0015] According to a fourth aspect of the present invention, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute any one of the methods according to the first aspect of the present invention.

[0016] The beneficial effects of the present invention are as follows:

[0017] The method and system for optimizing the route of a deliverable network of the present invention adopt an ant algorithm to obtain the first route of the deliverable network; extract key cross-node parameters in the first route of the deliverable network, and obtain the semantics of entity words in the context based on the key cross-node parameters; use the semantics as dynamic parameters to perform federated learning in order according to the business association weights to obtain the target semantics, and determine the target route of the deliverable network based on the target semantics. This method solves the problem of obtaining the optimal detection agent deployment solution from the detection agent deployment strategy in the integrated delivery process, and the possible local optimal rather than global optimal detection agent deployment solution. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are only used to illustrate specific embodiments and are not considered to limit the present invention. In the entire drawings, the same reference symbols represent the same components. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0019] Figure 1 is a flow chart of a method for optimizing a deliverable network route according to an embodiment of the present invention;

[0020] Figure 2 is a flow chart of a method for optimizing a deliverable network route according to a specific embodiment of the present invention;

[0021] Figure 3 is a schematic diagram of a federated learning process according to an embodiment of the present invention;

[0022] Figure 4 is a structural block diagram of a deliverable network route optimization system according to an embodiment of the present invention;

[0023] Figure 5 It is a structural block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work should fall within the scope of protection of the present invention.

[0025] Furthermore, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily obscuring the concepts disclosed in the present invention.

[0026] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the orientation or position relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inside", "outside", etc. is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention. In addition, the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. The terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0027] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of methods and systems consistent with some aspects of the present invention as detailed in the appended claims.

[0028] The present invention proposes a method, system and related equipment for optimizing a network route of deliverables. Specifically, the method, system and related equipment for optimizing a network route of deliverables according to an embodiment of the present invention are described below with reference to the accompanying drawings.

[0029] Figure 1 1 is a flow chart of a method for optimizing a delivery network route according to an embodiment of the present invention. It should be noted that the method for optimizing a delivery network route according to an embodiment of the present invention can be applied to a system for optimizing a delivery network route according to an embodiment of the present invention. The system for optimizing a delivery network route can be configured on an electronic device or in a server. This embodiment of the present application does not limit this.

[0030] like Figure 1 As shown, the deliverable network route optimization method includes:

[0031] S110, uses the ant algorithm to obtain the first route of the delivery network.

[0032] In an embodiment of the present invention, from the perspective of product operation, before integrated delivery, the deliverable network cost needs to be estimated. When the deliverable network is evaluated, an ant algorithm can be used to obtain the first route of the deliverable network.

[0033] In an embodiment of the present invention, an ant colony algorithm formula is determined, and parameters of the ant colony algorithm formula are initialized; in the case of the first iteration, the total distance of the path traveled by each ant is calculated, and the pheromone of each generation of ants is updated; while updating the pheromone of each generation of ants, iteration is continued until the iteration ends, and when the iteration ends, the path traveled by the ants with the shortest total path distance is selected; the path traveled by the ants with the shortest total path distance is used as the first route of the deliverable network.

[0034] Among them, the ant colony algorithm formula includes:

[0035]

[0036] in, is the probability that ant k moves from city i to city j at time t; α is the pheromone heuristic factor, which reflects the effect of pheromone on ant path selection; β is the expected heuristic factor, which reflects the degree to which pheromone is valued when ants choose paths; τ ij (t) is the pheromone concentration on the path connecting city i and city j at time t; n ij (t) is the expected degree of ants moving from city i to city j, n ij (t) = 1 / dij, where dij is the distance between city i and city j; J k (i) is the city that ant k can currently choose;

[0037] Among them, according to Update the pheromones of each generation of ants; among them, is the pheromone concentration released by the kth ant on the path connecting city i and city j, L k is the total distance of the path traveled by the kth ant, Q is the pheromone an ant has in its lifetime, Δτ ij is the sum of the pheromone concentrations released by ants on the path connecting city i and city j, m is the number of ants, and ρ represents the degree of pheromone dilution.

[0038] S120, extracting key cross-node parameters in the first route of the deliverable network, and obtaining the semantics of the entity word in the context based on the key cross-node parameters.

[0039] In an embodiment of the present invention, when the first route of the deliverable network is obtained, the parameters of the key intersections in the first route of the deliverable network can be extracted. For example, if A to B and Y to B both pass through node C, then C is a key intersection node, where the key intersection node parameters include but are not limited to delay, core business name, etc. When the key intersection node parameters in the first route of the deliverable network are extracted, the semantics of the entity word in the context can be obtained based on the key intersection node parameters based on model training. For specific implementation methods, please refer to the subsequent embodiments.

[0040] S130, using semantics as dynamic parameters to perform federated learning in order according to business association weights to obtain target semantics, and determining a deliverable network target route based on the target semantics.

[0041] According to the method for optimizing the route of a deliverable network in an embodiment of the present invention, an ant algorithm is used to obtain the first route of the deliverable network; key cross-node parameters in the first route of the deliverable network are extracted, and the semantics of the entity words in the context are obtained based on the key cross-node parameters; the semantics are used as dynamic parameters and federated learning is carried out in order according to the business association weights to obtain the target semantics, and the target route of the deliverable network is determined based on the target semantics. This method solves the problem of obtaining the optimal detection agent deployment solution from the detection agent deployment strategy in the integrated delivery process, and the possible local optimal rather than global optimal detection agent deployment solution.

[0042] In order to make it easier for those skilled in the art to understand the present invention, Figure 2 A method for optimizing a deliverable network route according to a specific embodiment of the present invention is as follows. Figure 2 As shown, the deliverable network route optimization method includes:

[0043] S210, using an ant algorithm to obtain the first route of the deliverable network.

[0044] In the embodiment of the present invention, the implementation of step S210 may refer to the implementation of step S110 described above, and the present invention will not elaborate on this any more.

[0045] S220, extracting key cross-node parameters in the first route of the deliverable network.

[0046] S230, obtaining the semantics of the entity word in the context based on the key cross node parameters.

[0047] In an embodiment of the present invention, key cross-node parameters are trained based on the ChatGPT model to obtain the semantics of the key cross-node parameters; the knowledge-enhanced KET model is used to analyze the semantics of the key cross-node parameters to obtain the semantics of the entity words in the context.

[0048] Among them, according to the relevance coefficient of the entity word rel k=min-max(s k )*abs(cos(CR(X i ), c k ))c k , determine the association strength between the entity word and the context, where s k For entity word c k The confidence score of CR is the entity word c k The context representation corresponding to the sentence, X i The element currently being processed.

[0049] Among them, abs is the absolute value function and cos is the cosine distance function.

[0050] In an embodiment of the present invention, when the semantics of an entity word in a context is obtained, the semantics can be matched with a semantic database to obtain attribute information of the semantics, wherein the attribute information includes whether the semantics is related to the business and the association weight between the semantics and the business.

[0051] S240, using semantics as dynamic parameters to perform federated learning in order according to business-related weights to obtain target semantics.

[0052] In an embodiment of the present invention, in the Nth training iteration of multiple terminals, based on the ChatGPT model, semantics are used as dynamic parameters, and federated learning training is performed according to the business association weights, and the training results are reported to the central server; the central server receives and aggregates the training results, and sends the global model and training configuration of the N+1th iteration to the terminal; the terminal trains the semantics based on the global model and training configuration of the N+1th iteration to obtain the target semantics.

[0053] Among them, the training results include but are not limited to DNN gradients, etc.

[0054] That is, the first training iteration is performed on multiple terminals to initialize federated learning. Each terminal uses semantics as dynamic parameters based on the ChatGPT model and trains according to the business-related weights, and then reports the training results. The central server receives and aggregates these training results; the central server sends the global model and training configuration of the next iteration to the terminal, and the terminal continues training based on the new global model and training configuration until the set number of iterations is reached or other stop conditions are met. After the training is completed, the target semantics is obtained.

[0055] In the Nth training iteration, the terminal performs local training using locally collected images / videos based on the global model downloaded from the central server.

[0056] Among them, the terminal reports the training results to the central server through the 5G line channel.

[0057] In one embodiment of the present invention, in order to fully utilize the training resources on the terminal and minimize the training delay, Figure 3 The training sequence shown requires the reporting of the training results of the N-1th iteration and the global model distribution and training parameter configuration of the N+1th iteration to be completed during the training of the Nth iteration. Of course, in the actual training process, it is also possible to consider using a relaxed sequence at the cost of sacrificing the convergence speed of the training, that is, performing non-real-time federated learning.

[0058] S250, determining a deliverable network target route based on the target semantics.

[0059] In an embodiment of the present invention, a target database corresponding to semantics and routes is obtained; based on the target database, a route corresponding to the target semantics is obtained, and the route is used as a deliverable network target route.

[0060] According to the delivery network route optimization method of the embodiment of the present invention, the ant algorithm can simulate the path selection behavior of ants in the foraging process, and use pheromones and heuristic factors to quickly find a better first route in a complex delivery network. This method can handle large-scale networks and has good robustness and global search capabilities; after obtaining the first route, by extracting key cross-node parameters (such as delay, core business name, etc.), it can identify nodes that are critical to network performance or business continuity; using the ChatGPT model and the knowledge-enhanced KET model to perform semantic analysis on key node parameters, it is possible to deeply understand the meaning and role of these nodes in the context, and provide an accurate information basis for subsequent optimization; matching semantics with semantic databases can obtain semantic attribute information, including relevance and relevance weights with businesses. This helps to train semantics in an orderly manner according to business relevance weights in the subsequent federated learning process, improving the pertinence and efficiency of learning; the federated learning mechanism allows distributed training of models on multiple terminals while protecting user data privacy. Through multiple iterations of training, the model can be gradually optimized to obtain more accurate target semantics; using the target semantics as a dynamic parameter and combining it with the business-related weights for federated learning can achieve dynamic optimization of the network route of the deliverables. This dynamism helps to adapt to changes in the network and business needs, and improve the flexibility and adaptability of the route; through high-speed communication channels such as 5G, real-time reporting of training results and rapid distribution of global models can be achieved, further shortening the training delay and improving the real-time performance of the system; this solution reduces the time cost of manual intervention and decision-making through automated route optimization and semantic analysis, and improves overall operational efficiency; at the same time, through precise semantic matching and federated learning, unnecessary resource waste can be avoided and operating costs can be reduced; it achieves intelligent, dynamic and efficient optimization of the network route of the deliverables, and improves overall operational efficiency and cost-effectiveness.

[0061] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0062] According to one aspect of an embodiment of the present invention, a delivery network route optimization system is also provided. Figure 4 is a schematic diagram of a deliverable network route optimization system according to an embodiment of the present invention; Figure 4 As shown, including:

[0063] An acquisition module 410 is used to acquire a first route of a deliverable network using an ant algorithm;

[0064] The obtaining module 420 is used to extract key cross-node parameters in the first route of the deliverable network, and obtain the semantics of the entity word in the context based on the key cross-node parameters;

[0065] The determination module 430 is used to perform federated learning in order according to the business association weights using the semantics as dynamic parameters to obtain target semantics, and determine a deliverable network target route based on the target semantics.

[0066] According to the delivery network route optimization system of the embodiment of the present invention, the ant algorithm is used to obtain the first route of the delivery network; the key cross node parameters in the first route of the delivery network are extracted, and the semantics of the entity words in the context are obtained based on the key cross node parameters; the semantics are used as dynamic parameters and federated learning is carried out in order according to the business association weights to obtain the target semantics, and the target route of the delivery network is determined based on the target semantics. This solves the problem that the detection agent deployment strategy in the integrated delivery process obtains the optimal detection agent deployment solution, which may be a local optimal but not a global optimal detection agent deployment solution.

[0067] Optionally, the acquisition module 410 is specifically used to determine the ant colony algorithm formula and initialize the parameters of the ant colony algorithm formula; in the case of the first iteration, calculate the total distance of the path traveled by each ant and update the pheromone of each generation of ants; while updating the pheromone of each generation of ants, continue to iterate until the iteration ends, and when the iteration ends, select the path traveled by the ants with the shortest total path distance; and use the path traveled by the ants with the shortest total path distance as the first route of the deliverable network.

[0068] Optionally, the ant colony algorithm formula includes: in, is the probability that ant k moves from city i to city j at time t; α is the pheromone heuristic factor, which reflects the effect of pheromone on ant path selection; β is the expected heuristic factor, which reflects the degree to which pheromone is valued when ants choose paths; τ ij (t) is the pheromone concentration on the path connecting city i and city j at time t; n ij (t) is the expected degree of ants moving from city i to city j, n ij (t) = 1 / dij, where dij is the distance between city i and city j; J k (i) is the city that ant k can currently choose; Update the pheromones of each generation of ants; among them, is the pheromone concentration released by the kth ant on the path connecting city i and city j, L k is the total distance of the path traveled by the kth ant, Q is the pheromone an ant has in its lifetime, Δτ ij is the sum of the pheromone concentrations released by ants on the path connecting city i and city j, m is the number of ants, and ρ represents the degree of pheromone dilution.

[0069] Optionally, a module 420 is obtained, which is specifically used to train the key cross-node parameters based on the ChatGPT model to obtain the semantics of the key cross-node parameters; and use the knowledge-enhanced KET model to analyze the semantics of the key cross-node parameters to obtain the semantics of the entity word in the context.

[0070] Optionally, according to the relevance coefficient rel of the entity word k =min-max(s k )*abs(cos(CR(X i ), c k ))c k , determine the association strength between the entity word and the context, where s k For entity word c k The confidence score of CR is the entity word ck The context representation corresponding to the sentence, X i The element currently being processed.

[0071] Optionally, the determination module 430 is specifically used to, in the Nth training iteration of multiple terminals, based on the ChatGPT model, use the semantics as dynamic parameters, perform federated learning training according to the business association weights, and report the training results to the central server; the central server receives and aggregates the training results, and at the same time sends the global model and training configuration of the N+1th iteration to the terminal; the terminal trains the semantics based on the global model and the training configuration of the N+1th iteration to obtain the target semantics.

[0072] Optionally, the determination module 430 is specifically configured to obtain a target database corresponding to semantics and routes; based on the target database, obtain a route corresponding to the target semantics, and use the route as a deliverable network target route.

[0073] According to one aspect of an embodiment of the present invention, an electronic device is provided.

[0074] Figure 5 Schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Figure 5 As shown, the electronic device may include one or more ( Figure 5 Only one is shown in the figure) processor 102 (processor 102 may include but is not limited to a microprocessor (Microprocessor Unit, referred to as MPU) or a programmable logic device (Programmable logic device, referred to as PLD)) and a memory 104 for storing data. In an exemplary embodiment, the electronic device may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 5 The structure shown is for illustration only and does not limit the structure of the above terminal device. Figure 5 More or fewer components as shown, or with Figure 5 Equivalent functions or comparisons shown Figure 5 A different configuration with more features is shown.

[0075] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the deliverable network route optimization method in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the terminal device via a network. Examples of the above network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0076] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the switching device. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0077] The present invention provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable the computer to execute a delivery network route optimization method.

[0078] The applicant of the present invention has made a detailed explanation and description of the implementation examples of the present invention in conjunction with the drawings in the specification. However, those skilled in the art should understand that the above implementation examples are only preferred implementation schemes of the present invention, and the detailed description is only to help readers better understand the spirit of the present invention, but not to limit the scope of protection of the present invention. On the contrary, any improvements or modifications based on the inventive spirit of the present invention should fall within the scope of protection of the present invention.

[0079] In addition, each functional unit in each embodiment of the present invention may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0080] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the embodiments of the present invention, rather than to limit them. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions recorded in the above embodiments, or replace some of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. Any changes or replacements that can be easily thought of by those skilled in the art within the technical scope disclosed in the present invention should be covered within the protection scope of the present invention.

Claims

1. A method for optimizing a deliverable network route, characterized in that: include: S1, uses the ant algorithm to obtain the first route of the deliverable network; S2, extracting key cross-node parameters in the first route of the deliverable network, and obtaining the semantics of the entity word in the context based on the key cross-node parameters; S3, using the semantics as dynamic parameters to perform federated learning in order according to the business association weights to obtain target semantics, and determining a deliverable network target route based on the target semantics.

2. The method for optimizing the delivery network route according to claim 1, characterized in that: Use the ant algorithm to obtain the first route of the deliverable network, including: Determine an ant colony algorithm formula, and initialize parameters of the ant colony algorithm formula; In the first iteration, the total distance of the path traveled by each ant is calculated, and the pheromone of each generation of ants is updated; While updating the pheromones of each generation of ants, the iteration continues until the iteration ends. At the end of the iteration, the path taken by the ants with the shortest total path distance is selected; The path taken by the ant with the shortest total path distance is taken as the first route of the delivery network.

3. The method for optimizing the delivery network route according to claim 1, characterized in that: Ant colony algorithm formula, including: in, is the probability that ant k moves from city i to city j at time t; α is the pheromone heuristic factor, which reflects the effect of pheromone on ant path selection; β is the expected heuristic factor, which reflects the degree to which pheromone is valued when ants choose paths; τ ij (t) is the pheromone concentration on the path connecting city i and city j at time t; n ij (t) is the expected degree of ants moving from city i to city j, n ij (t) = 1 / dij, where dij is the distance between city i and city j; J k (i) is the city that ant k can currently choose; Among them, according to Update pheromones for each generation of ants; in, is the pheromone concentration released by the kth ant on the path connecting city i and city j, L k is the total distance of the path traveled by the kth ant, Q is the pheromone an ant has in its lifetime, Δτ ij is the sum of the pheromone concentrations released by ants on the path connecting city i and city j, m is the number of ants, and ρ represents the degree of pheromone dilution.

4. The method for optimizing the delivery network route according to claim 1, characterized in that: The semantics of the entity word in the context is obtained based on the key cross node parameters, including: Training the key cross-node parameters based on the ChatGPT model to obtain the semantics of the key cross-node parameters; The knowledge-enhanced KET model is used to analyze the semantics of the key cross-node parameters to obtain the semantics of the entity word in the context.

5. The method for optimizing the delivery network route according to claim 1, characterized in that: According to the relevance coefficient rel of the entity word k =min-max(s k )*abs(cos(CR(X i ), c k ))c k , determine the association strength between the entity word and the context, where s k For entity word c k The confidence score of CR is the entity word c k The context representation corresponding to the sentence, X i The element currently being processed.

6. The method for optimizing a deliverable network route according to claim 1, characterized in that: The semantics are used as dynamic parameters and federated learning is performed in order according to the business association weights to obtain the target semantics, including: In the Nth training iteration of multiple terminals, based on the ChatGPT model, the semantics are used as dynamic parameters, and federated learning training is performed according to the business association weights, and the training results are reported to the central server; The central server receives and aggregates the training results, and sends the global model and training configuration of the N+1th iteration to the terminal; The terminal trains the semantics based on the global model of the N+1th iteration and the training configuration to obtain the target semantics.

7. The method for optimizing a deliverable network route according to claim 1, characterized in that: Determining a deliverable network target route based on the target semantics includes: Obtain a target database corresponding to semantics and routes; Based on the target database, a route corresponding to the target semantics is obtained, and the route is used as the deliverable network target route.

8. A deliverable network route optimization system, characterized in that: include: An acquisition module, used for acquiring the first route of the deliverable network by using an ant algorithm; An obtaining module, used to extract key cross-node parameters in the first route of the deliverable network, and obtain the semantics of entity words in the context based on the key cross-node parameters; The determination module is used to perform federated learning in order according to the business association weights using the semantics as dynamic parameters to obtain target semantics, and determine the deliverable network target route based on the target semantics.

9. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 7.

Images