Dynamic progressive path planning method based on EPO algorithm

Abstract

The invention relates to the technical field of logistics path planning, in particular to a dynamic progressive path planning method based on an EPO algorithm, and the method comprises the steps: collecting customer demands, and carrying out the matching of corresponding cargo attributes; sequentially constructing target functions corresponding to the transportation cost, the time cost, the carrier carbon emission cost and the carrier service quality cost, respectively determining weight parameters, and integrating and establishing a multi-target cost function; respectively determining a transportation state value and a storage state value of a logistics node in a cargo transportation process, and integrating to generate a logistics node dynamic judgment mechanism; and performing variation improvement on the EPO algorithm based on a differential variation method, outputting an optimal carrier combination number, a transportation mode and a logistics node number in a cargo transportation process, and obtaining an optimal path planning result. The EPO algorithm after differential variation optimization can dynamically adjust the optimization direction according to the change of the transportation environment, and has stronger adaptability and stability in a complex logistics network with a huge number of logistics nodes.

Description

Technical Field

[0001] The present invention relates to the technical field of logistics path planning, and particularly relates to a dynamic progressive path planning method based on the EPO algorithm. Background Art

[0002] With the rapid development of global trade and the increasing complexity of the logistics network, multimodal transport has attracted wide attention as an efficient cargo transport mode; multimodal transport realizes the efficient and economical transport of goods through the effective connection of various transport modes such as road, rail, and water transport. In the multimodal transport system, the planning of the transport path and the dynamic adaptability to the transport environment directly affect the overall efficiency and service quality of the supply chain.

[0003] However, in most of the existing path planning algorithms, they are mostly considered based on a static environment or calculate the path according to preset information. In the real complex and changeable dynamic environment, it is easy for the original decision to fail due to environmental changes and fall into the local optimum of path planning, unable to have the ability to perceive and dynamically judge the real-time state of nodes during the transport process, and unable to make timely adjustments, resulting in problems such as path failure or resource misallocation; and if a node failure occurs, the vehicle is very likely to waste a lot of time waiting for the node to resume normal operation or re-route; in addition, for the dynamic environment, the existing method usually adopts an optimization scheme that combines the EPO (Emperor Penguin Optimization Algorithm) algorithm with the traditional graph search algorithm, but when facing a complex multimodal transport network, it is easy to fall into the local optimum and lacks a global dynamic coordination mechanism. Especially in the multimodal transport scenario, the state of the nodes, that is, environmental states such as congestion degree, service accessibility, and resource vacancy, have a decisive impact on the feasibility and efficiency of the overall path. If the changing trend of the nodes cannot be accurately judged and the path strategy cannot be dynamically adjusted, it will directly affect the stability and response speed of the transport chain. Summary of the Invention

[0004] In order to solve the technical problems that the existing path planning algorithms are difficult to adapt to the dynamic environment and are easy to fall into the local optimum in the complex and changeable logistics transport environment, and cannot accurately judge the changing trend of nodes and affect the feasibility and efficiency of the overall path planning, the purpose of the present invention is to provide a dynamic progressive path planning method based on the EPO algorithm, and the specific technical solution adopted is as follows:

[0005] Obtain logistics nodes based on cargo transport, collect customer requirements, and match corresponding cargo attributes;

[0006] According to the customer requirements and the corresponding goods attributes, the objective functions corresponding to the transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost are constructed in sequence, and the weight parameters are determined respectively, and a multi-objective cost function is integrated and established;

[0007] The transportation status value and warehousing status value of the logistics nodes in the process of goods transportation are determined respectively, and a dynamic judgment mechanism for logistics nodes is integrated and generated;

[0008] Through the multi-objective cost function and the dynamic judgment mechanism of logistics nodes, the EPO algorithm is mutated and improved based on the differential mutation method, and the optimal carrier combination number, transportation mode, and logistics node number in the process of goods transportation are output, and the optimal path planning result is obtained.

[0009] Preferably, according to the customer requirements and the corresponding goods attributes, the objective functions corresponding to the transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost are constructed in sequence, including:

[0010] Construct the objective function of transportation cost, and the corresponding calculation formula is:

[0011]

[0012] Among them, C (x) represents the objective function of transportation cost; M represents the total number of transportation modes in goods transportation; N k represents the total number of logistics nodes; β i represents all reachable subsequent nodes of the i-th logistics node in the transportation process; d ij represents the actual transportation distance between the i-th logistics node and the j-th logistics node; represents the unit distance cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the transfer cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode;

[0013] Construct the objective function of time cost, and the corresponding calculation formula is:

[0014]

[0015] Among them, T (x) represents the objective function of time cost; represents the transportation time of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the transportation time corresponding weight parameter; represents the transfer time of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; Indicates the transfer time The corresponding weight parameter; Indicates the time window cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode;

[0016] Based on the use of fuel vehicles and / or new energy vehicles for road transportation during the goods transportation process, construct the objective function of the carrier's carbon emission cost, and the corresponding calculation formula is:

[0017]

[0018] Among them, R (x) Indicates the objective function of the carrier's carbon emission cost; Indicates the carbon emission cost of transporting from the i-th logistics node to the j-th logistics node by fuel vehicle; Indicates the carbon emission cost of transporting from the i-th logistics node to the j-th logistics node by new energy vehicle;

[0019] Construct the objective function of the carrier's service quality cost, and the corresponding calculation formula is:

[0020]

[0021] Among them, Q (x) Indicates the objective function of the carrier's service quality cost; Indicates the quantity of goods that need to be preferentially transported from the i-th logistics node to the j-th logistics node by the k-th transportation mode; q total Indicates the quantity of goods that need to be preferentially transported for all logistics nodes; Indicates the cold chain goods transported from the i-th logistics node to the j-th logistics node by the k-th transportation mode; T cold-maxl Indicates the cold chain goods for all logistics nodes; Indicates the quantity of orders that can be promptly responded to when transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; q allowed Indicates the quantity of orders for all logistics nodes; Indicates the transportation accuracy rate of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; w5, w6, w7, w8 respectively represent the service quality of preferentially transporting goods, the service quality of transporting cold chain goods, the service quality of order response rate, and the service quality of transportation accuracy rate.

[0022] Preferably, determine the weight parameters respectively, and integrate and establish a multi-objective cost function, and the corresponding calculation formula is:

[0023] F(X) = w1C (x) + w2T (x) + w3R (x) + w4Q(x)

[0024] Among them, F(X) represents the multi-objective cost function; w1, w2, w3, and w4 respectively represent the corresponding weight parameters determined according to customer requirements; C (x) , T (x) , R (x) , and Q (x) respectively represent the objective functions of transportation cost, time cost, carbon emission cost of the carrier, and service quality cost of the carrier.

[0025] Preferably, the transportation status value and storage status value of the logistics nodes during the goods transportation process are determined respectively, and a logistics node dynamic judgment mechanism is integrated, including:

[0026] Based on the goods attributes, the reliability of different transportation modes, the maintenance probability of the logistics nodes, and the influence degree of environmental factors on the transportation of the logistics nodes are determined in sequence, and the transportation status value of the logistics nodes when different transportation modes are adopted is obtained;

[0027] The storage points at the logistics nodes are divided into a healthy state and a failure state, the cluster centers are obtained respectively, the storage status of the storage points is determined, the cluster centers are updated, and the distance between the storage points and the updated cluster centers is calculated to obtain the storage status value of the logistics nodes;

[0028] The node dynamic judgment mechanism is generated by combining the transportation status value and storage status value of the logistics nodes.

[0029] Preferably, based on the goods attributes, the reliability of different transportation modes, the maintenance probability of the logistics nodes, and the influence degree of environmental factors on the transportation of the logistics nodes are determined in sequence, and the transportation status value of the logistics nodes when different transportation modes are adopted is obtained, including:

[0030] Determine the reliability of different transportation modes, and the corresponding calculation formula is:

[0031]

[0032] Among them, represents the reliability of the i-th logistics node when the k-th transportation mode is adopted; C i,k represents the load upper limit of the k-th transportation mode at the i-th logistics node; L i,k represents the current load of the k-th transportation mode at the i-th logistics node;

[0033] Determine the maintenance probability of the logistics nodes, and the corresponding calculation formula is:

[0034]

[0035] Among them, Denote the maintenance probability of the logistics equipment in the \(i\)-th logistics node when adopting the \(k\)-th transportation mode; \(N\) represents the number of logistics equipment in the \(i\)-th logistics node; Denote the remaining life of the logistics equipment in the \(i\)-th logistics node; Denote the total life of the logistics equipment in the \(i\)-th logistics node;

[0036] Determine the influence degree of environmental factors on the transportation of the logistics node. The environmental factors include weather influence and traffic congestion, and the corresponding calculation formula is:

[0037]

[0038] where Denote the influence degree of environmental factors on the transportation of the \(i\)-th logistics node when adopting the \(k\)-th transportation mode; \(T\) i,k Denote the traffic congestion index at the \(i\)-th logistics node when adopting the \(k\)-th transportation mode; \(W\) i,k Denote the weather influence index at the \(i\)-th logistics node when adopting the \(k\)-th transportation mode; \(\alpha\) and \(\delta\) respectively represent the weights of the traffic congestion index and the weather influence index;

[0039] Obtain the transportation status value of the logistics node when different transportation modes are adopted by the logistics node. The corresponding calculation formula is:

[0040]

[0041] where Denote the transportation status value when adopting the \(k\)-th transportation mode at the \(i\)-th logistics node.

[0042] Preferably, divide the storage points at the logistics node into a healthy state and a failure state, respectively obtain the cluster centers, determine the storage state of the storage points, update the cluster centers, calculate the distance between the storage points and the updated cluster centers, and obtain the storage status value of the logistics node, including:

[0043] Obtain the feature vector based on the storage points at any logistics node, denoted as \(X\) i =\(\{h_1, h_2, h_3, \ldots, h\) n \(\}\), \(n\) represents the total number of storage points at the \(i\)-th logistics node. Divide them into a healthy state and a failure state according to the feature vector, and respectively obtain the cluster centers, denoted as

[0044] Determine the storage state of any storage point. The corresponding calculation formula is:

[0045]

[0046] where \(h\) s Denote the storage state of the \(s\)-th storage point; \(S\)i,s represents the current storage volume of the s-th storage point at the i-th logistics node; represents the maximum storage volume of the s-th storage point at the i-th logistics node;

[0047] Initialize and update the cluster center, and the corresponding calculation formula is:

[0048]

[0049] where, represents the updated cluster center for the healthy state or failure state; J i,(0→n+1) represents the transportation status value corresponding to the change in the update process of the i-th logistics node; h heealthy / failed represents the numerical value corresponding to the healthy state or failure state of the storage point;

[0050] Calculate the distance between the storage point and the updated cluster center, and the corresponding calculation formula is:

[0051]

[0052] where, D healthy / failed represents the distance between the storage point and the updated cluster center for the healthy state or failure state;

[0053] Obtain the storage status value of the logistics node, and the corresponding calculation formula is:

[0054]

[0055] where, represents the storage status value of the i-th logistics node.

[0056] Preferably, a node dynamic judgment mechanism is generated by combining the transportation status value and the storage status value of the logistics node, including:

[0057] Calculate the health degree of the logistics node, and the corresponding calculation formula is:

[0058]

[0059] where, J i represents the current health degree of the i-th logistics node; represents the transportation status value of the i-th logistics node using the k-th transportation method; represents the storage status value of the i-th logistics node;

[0060] Set a screening threshold. When the health degree is higher than the screening threshold, it indicates that the current logistics node is in a normal state; when the health degree is lower than the screening threshold, it indicates that the current logistics node is in an abnormal state.

[0061] Preferably, through a multi-objective cost function and a logistics node dynamic judgment mechanism, the EPO algorithm is mutated and improved based on the differential mutation method, and the optimal carrier combination number, transportation mode, and logistics node number during the cargo transportation process are output to obtain the optimal path planning result, including:

[0062] The EPO algorithm is mutated and improved using the differential mutation method. Based on the logistics node dynamic judgment mechanism, it is judged whether a new exploration direction is obtained. If so, a new exploration direction is determined in combination with the multi-objective cost function, and the distance between the position of the transportation vehicle used during the transportation process and the final logistics node of the corresponding transportation path is calculated;

[0063] Iteration is performed according to the new exploration direction and distance until the iteration is successful, and the optimal carrier combination number, transportation mode, and logistics node number during the cargo transportation process are output to obtain the optimal path planning result.

[0064] Preferably, the EPO algorithm is mutated and improved using the differential mutation method. Based on the logistics node dynamic judgment mechanism, it is judged whether a new exploration direction is obtained. If so, a new exploration direction is determined in combination with the multi-objective cost function, and the distance between the position of the transportation vehicle used during the transportation process and the final logistics node of the corresponding transportation path is calculated, including:

[0065] If the current logistics node is abnormal, it means that a new exploration direction needs to be obtained, and a new exploration direction is determined in combination with the multi-objective cost function. The corresponding calculation formula is:

[0066]

[0067] where σ represents the new exploration direction determined based on the current abnormal logistics node; F(X) represents the multi-objective cost function; ξ represents the random exploration weight; T t represents the temperature change function of the current logistics node; r represents the rate of controlling temperature decay; represents the distance between the current abnormal logistics node and the candidate logistics node; T represents the time spent obtaining the new exploration direction; max Gen represents the maximum number of iterations; gen represents the number of iterations for determining the temperature change of the current logistics node; x represents the adjustable parameter for controlling the maximum number of iterations;

[0068] The trend of the transportation vehicle used during the transportation process moving towards the optimal logistics node is determined, and the distance between the position of the transportation vehicle used during the transportation process and the final logistics node of the corresponding transportation path is calculated. The corresponding calculation formula is:

[0069]

[0070] where, represents the distance between the location of the currently adopted means of transportation during transportation and the final logistics node of the corresponding transportation route; S represents the tendency of the currently adopted means of transportation to move towards the optimal logistics node during transportation; represents a parameter for avoiding conflicts between the currently adopted means of transportation and the next logistics node; represents the location of the final logistics node of the transportation route; represents the location of the currently adopted means of transportation during transportation; N unvisted 、N visted respectively represent the logistics nodes not passed by the currently adopted means of transportation and the logistics nodes that have been passed during transportation; respectively represent the locations of the logistics nodes not passed and the logistics nodes that have been passed; f and l respectively represent the exploration and extension parameters during the iteration process of the EPO algorithm; M represents the control parameter for avoiding conflicts between the logistics nodes passed during transportation; T1 represents the temperature change function of the EPO algorithm; represents the location of the current logistics node; represents the locations of other nodes during transportation; Rand represents a random number uniformly distributed on [0,1].

[0071] Preferably, iterate according to the new exploration direction and distance until the iteration is successful, and output the optimal carrier combination number, transportation mode, and the logistics node numbers during the cargo transportation process to obtain the optimal path planning result, including:

[0072] Improve the EPO algorithm by mutation according to the new exploration direction and distance to determine the EPO model, and the corresponding calculation formula is:

[0073]

[0074] Among them, represents the EPO model; J represents the health degree of the logistics node before exploration; represents the logistics node that best meets the customer needs among the logistics nodes adjacent to the current logistics node for judging the multi-objective function; represents the location of the newly planned logistics node after determining the new exploration direction;

[0075] Iterate based on the EPO model. If the iteration is successful, output the optimal carrier combination number, transportation mode, and the logistics node numbers during the cargo transportation process to determine the optimal path planning result; if the iteration fails, ignore the objective function with the lowest weight parameter based on the multi-objective function, and re-iterate using the EPO model until the iteration is successful to obtain the optimal path planning result.

[0076] The present invention has the following beneficial effects:

[0077] This application uses a multi-objective cost function. That is, on the premise of meeting the timeliness and reliability of cargo transportation, it adapts to the changes in different transportation scenarios to meet the diverse needs of customers. Combining with the dynamic evaluation mechanism of logistics node status, it real-time monitors dynamic environmental factors such as traffic congestion, equipment failures, and weather changes, enabling the transportation route to be adaptively adjusted, effectively avoiding failed nodes, and ensuring the continuity and stability of cargo transportation. That is, by introducing a dynamic path adjustment strategy, the transportation route can adapt to environmental changes during the iteration process, improving the flexibility and optimization ability of path planning, and achieving the overall optimization of transportation costs, time efficiency, and service quality. Cooperating with the global planning and direction-finding capabilities of the EPO algorithm, it uses differential mutation to optimize the original EPO algorithm, enabling the path planning model to dynamically adjust the optimization direction according to changes in the transportation environment, improving the optimization efficiency, and realizing the improvement of the intelligent decision-making ability of the path optimization algorithm. It has stronger adaptability and stability in complex logistics networks with a large number of logistics nodes. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0079] Figure 1 It is a flowchart of the steps of a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention;

[0080] Figure 2 It is an implementation flowchart of a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention;

[0081] Figure 3 It is a comparison chart of transportation costs between a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention and different algorithms;

[0082] Figure 4 It is a comparison chart of operation times between a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention and different algorithms;

[0083] Figure 5 It is a comparison chart of convergence times between a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention and different algorithms;

[0084] Figure 6A comparison chart of the adaptability of a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention with different algorithms using different numbers of logistics nodes. Detailed implementation manners

[0085] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details a dynamic progressive path planning method based on the EPO algorithm proposed by the present invention, its specific implementation manners, structures, features and effects as follows. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0086] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0087] The following specifically describes in conjunction with the accompanying drawings the specific solution of a dynamic progressive path planning method based on the EPO algorithm provided by the present invention.

[0088] Please refer to Figure 1 and Figure 2 , which respectively show the step flow chart and implementation flow chart of a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention. The method includes:

[0089] Step S1: Obtain logistics nodes based on goods transportation, collect customer requirements, and match corresponding goods attributes;

[0090] Step S2: According to customer requirements and corresponding goods attributes, sequentially construct objective functions corresponding to transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost, and respectively determine weight parameters, and integrate and establish a multi-objective cost function;

[0091] Step S3: Respectively determine the transportation status value and warehousing status value of the logistics nodes during the goods transportation process, and integrate and generate a logistics node dynamic judgment mechanism;

[0092] Step S4: Through the multi-objective cost function and the logistics node dynamic judgment mechanism, based on the differential mutation method, mutate and improve the EPO algorithm, and output the optimal carrier combination number, transportation mode, and the logistics node number during the goods transportation process to obtain the optimal path planning result.

[0093] For better illustration, the EPO algorithm is a new type of swarm intelligence algorithm that simulates the behavior of emperor penguins gathering to generate heat to maintain body temperature in the cold winter. Each emperor penguin individual will move towards the direction with higher temperature in the group and find the optimal position at the highest temperature. During the movement, each individual adjusts according to the distance between itself and the optimal position and continuously updates the optimal position, so as to achieve the optimal movement and position update of individuals within the group.

[0094] As an alternative implementation, in this embodiment, in step S1, logistics nodes are obtained based on cargo transportation, and 100 logistics nodes in a certain region of the country are analyzed. The regional scope roughly covers from the west to the east of the country, including inland and coastal cities. 40 logistics nodes are selected respectively, and they are incremented to 100 logistics nodes in sequence according to the iterative process of the algorithm, that is, a dynamic progressive path planning is carried out according to the number of logistics nodes. Part of the logistics nodes are selected for training and gradually incremented to the maximum number of logistics nodes in the entire logistics network, forming an intermodal logistics network, which can comprehensively reflect the characteristics and needs of logistics nodes in different regions, making the path planning more in line with the actual situation; and the dynamic progressive method can improve the efficiency and accuracy of path planning, enabling the algorithm provided in this application to adjust the path planning in real time during each iteration, weigh the status information between different logistics nodes, and be more adaptable to the dynamic changes of the logistics network, ensuring that the path planning scheme obtained after continuous adjustment and optimization is more comprehensive, with stronger flexibility and scalability, and thus obtaining the optimal cost and decision-making time.

[0095] It can be explained that customer requirements refer to the personalized requirements for realizing the corresponding transportation operation of goods, such as the transportation mode selected by the customer, the type of transportation vehicle, and the expected service quality provided; cargo attributes include the physical attributes of the goods, that is, the size, weight, fragility, and shelf life of the goods, etc.; transportation requirements, such as temperature and humidity control, etc.; environmental factors, that is, whether the goods are susceptible to moisture or vibration. Matching customer requirements with cargo attributes can achieve the precise transportation of goods from the starting logistics node to the target logistics node, ensuring the safe and efficient transportation of goods.

[0096] Furthermore, in step S2, according to customer requirements and the corresponding cargo attributes, objective functions corresponding to transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost are constructed in sequence, including:

[0097] Construct the objective function of transportation cost, and the corresponding calculation formula is:

[0098]

[0099] Among them, C (x) represents the objective function of transportation cost; M represents the total number of transportation modes in cargo transportation; N krepresents the total number of logistics nodes; β i represents all the reachable subsequent nodes of the i-th logistics node during transportation; d ij represents the actual transportation distance between the i-th logistics node and the j-th logistics node; represents the unit distance cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the transfer cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode.

[0100] Optionally, in this embodiment, M represents the total number of transportation modes in cargo transportation. Based on the multimodal transportation logistics network formed in the country, it corresponds to railway transportation, road transportation, and air transportation, that is, M = 3.

[0101] Construct the objective function of time cost, and the corresponding calculation formula is:

[0102]

[0103] where, T (x) represents the objective function of time cost; represents the transportation time of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the transportation time corresponding weight parameter; represents the transfer time of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the transfer time corresponding weight parameter; represents the time window cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode.

[0104] Specifically, according to customer requirements, some goods may have a limited transportation period, and they must arrive within the time range specified by the customer. If the goods arrive early, the customer needs to bear more warehousing costs, that is, the costs generated by storing the goods in the warehouse for a period of time due to early arrival, including warehouse rent, maintenance costs, and management costs, etc.; if the goods arrive late, the customer needs to bear more shortage costs, that is, the losses caused by the inability to meet customer needs due to late arrival of the goods, which may include the loss of sales opportunities, the decline of customer satisfaction, and possible compensation costs, etc.; therefore, determine the constraint conditions of the objective function of time cost, comprehensively consider the two costs, and find the best transportation time to minimize the cost. The corresponding calculation formula is:

[0105]

[0106]

[0107] Among them, represents the normalized time value of the transportation time and the transshipment time The range of values is between [0, 1]; represents the time weight obtained by mapping the time value The range of values is between [0.5, 1]; represents the time window cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the unit time storage cost generated by arriving early when transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the unit time shortage cost generated by arriving late when transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the start time of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the transportation time of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the end time of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode.

[0108] It can be understood that while using fuel vehicles, new energy vehicles will also be introduced for freight transportation. Due to the different costs between fuel and charging, there is a cost difference between fuel consumption and charging consumption.

[0109] Based on the use of fuel vehicles and / or new energy vehicles for road transportation during the cargo transportation process, the objective function of the carrier's carbon emission cost is constructed, and the corresponding calculation formula is:

[0110]

[0111] Among them, R (x) represents the objective function of the carrier's carbon emission cost; represents the carbon emission cost of transporting from the i-th logistics node to the j-th logistics node by fuel vehicle; represents the carbon emission cost of transporting from the i-th logistics node to the j-th logistics node by new energy vehicle.

[0112] Understandably, according to different customer requirements, the service quality of carriers for special goods also varies. That is, if customers require guaranteed priority transportation, cold-chain transportation, urgent processing, or need order goods with a higher accuracy rate, carriers need to provide additional protection or urgent transportation for such special goods, resulting in different costs. Correspondingly, the higher the service quality, the higher the cost paid by the customer for the entrusted carrier. Therefore, corresponding weights need to be set for different goods to ensure the accuracy of the constructed objective function.

[0113] Construct the objective function of the carrier service quality cost, and the corresponding calculation formula is:

[0114]

[0115] Among them, Q (x) represents the objective function of the carrier service quality cost; represents the quantity of goods that need priority transportation when using the k-th transportation mode from the i-th logistics node to the j-th logistics node; q total represents the quantity of goods that need priority transportation for all logistics nodes; represents the cold-chain goods when using the k-th transportation mode from the i-th logistics node to the j-th logistics node; T cold-maxl represents the cold-chain goods for all logistics nodes; represents the order volume that can be promptly responded to when using the k-th transportation mode from the i-th logistics node to the j-th logistics node; q allowed represents the order volume for all logistics nodes; represents the transportation accuracy rate when using the k-th transportation mode from the i-th logistics node to the j-th logistics node; w5, w6, w7, and w8 respectively represent the service quality of priority transportation goods, the service quality of transporting cold-chain goods, the service quality of order response rate, and the service quality of transportation accuracy rate.

[0116] Furthermore, in step S2, the weight parameters are determined respectively, and a multi-objective cost function is integrated and established. The corresponding calculation formula is:

[0117] F(X) = w1C (x) + w2T (x) + w3R (x) + w4Q (x)

[0118] Among them, F(X) represents the multi-objective cost function; w1, w2, w3, and w4 respectively represent the corresponding weight parameters determined according to customer requirements; C (x) , T (x) , R (x) and Q (x)Objective functions representing transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost respectively.

[0119] Make an explanation. Determine the weight parameters corresponding to each objective function in combination with specific business scenarios, that is, integrate the transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost, consider multiple factors to obtain a complete multi-objective cost function, comprehensively reflect various costs and benefits in the transportation process, and ensure the rationality and flexibility of optimizing path decisions under different customer demands and different transportation conditions to achieve the maximization of transportation efficiency.

[0120] It can be understood that for different transportation modes such as railway transportation, road transportation, and air transportation, the dynamic environmental factors they face are also different. That is, when considering different transportation modes comprehensively during the transportation of goods, the failure status of logistics nodes caused by traffic congestion, equipment failures, warehouse overstocking, weather reasons, etc. in the initial planned path should be considered, so as to be able to make another more optimal path plan in real time when the logistics nodes fail, and ensure that the goods can complete the transportation task smoothly and efficiently.

[0121] Furthermore, in step S3, it includes:

[0122] Step S31: Based on the goods attributes, sequentially determine the reliability of different transportation modes, the maintenance probability for logistics nodes, and the impact degree of environmental factors on the transportation of logistics nodes, to obtain the transportation status values of logistics nodes when using different transportation modes.

[0123] It can be explained that the transportation status value of a logistics node refers to the railway status, road congestion degree, and port status corresponding to railway transportation, road transportation, and air transportation.

[0124] Furthermore, in step S31, it includes:

[0125] Step S311: Determine the reliability of different transportation modes, and the corresponding calculation formula is:

[0126]

[0127] Among them, represents the reliability of the i-th logistics node when using the k-th transportation mode; C i,k represents the load upper limit of the k-th transportation mode at the i-th logistics node; L i,k represents the current load of the k-th transportation mode at the i-th logistics node;

[0128] Step S312: Determine the maintenance probability for logistics nodes, and the corresponding calculation formula is:

[0129]

[0130] Among them, represents the maintenance probability of the logistics equipment in the i-th logistics node when adopting the k-th transportation mode; N represents the number of logistics equipment in the i-th logistics node; represents the remaining life of the logistics equipment in the i-th logistics node; represents the total life of the logistics equipment in the i-th logistics node;

[0131] Step S313: Determine the influence degree of environmental factors on the transportation of the logistics node. The environmental factors include weather influence and traffic congestion, and the corresponding calculation formula is:

[0132]

[0133] Among them, represents the influence degree of environmental factors on the transportation of the i-th logistics node when adopting the k-th transportation mode; T i,k represents the traffic congestion index at the i-th logistics node when adopting the k-th transportation mode; W i,k represents the weather influence index at the i-th logistics node when adopting the k-th transportation mode; α and δ respectively represent the weights of the traffic congestion index and the weather influence index;

[0134] Step S314: Obtain the transportation state value of the logistics node when different transportation modes are adopted by the logistics node, and the corresponding calculation formula is:

[0135]

[0136] Among them, represents the transportation state value of the i-th logistics node when adopting the k-th transportation mode.

[0137] Step S32: Divide the storage points at the logistics node into healthy state and failure state, respectively obtain the cluster centers, determine the storage state of the storage points, update the cluster centers, calculate the distance between the storage points and the updated cluster centers, and obtain the storage state value of the logistics node.

[0138] It can be explained that the storage state value refers to whether the storage points in the logistics node corresponding to the required goods transportation are overstocked, so as to judge whether to replace the storage points preset based on the current logistics node; it can be understood that for the logistics nodes in the multi-modal logistics network, a logistics node does not correspond to only one storage point, and it contains multiple storage points at different locations. Therefore, the clustering algorithm is used to judge whether there is a possibility of temporarily replacing the storage points.

[0139] Furthermore, in step S32, it includes:

[0140] Step S321: Obtain the feature vector based on the storage points at any logistics node, denoted as Xi = {h1, h2, h3, …, h n}, where n represents the total number of storage points at the i-th logistics node. It is divided into a healthy state and a failure state according to the eigenvector, and the cluster centers are obtained respectively, denoted as

[0141] Step S322: Determine the storage state of any storage point. The corresponding calculation formula is:

[0142]

[0143] where h s represents the storage state of the s-th storage point; S i,s represents the current storage volume of the s-th storage point at the i-th logistics node; represents the maximum storage volume of the s-th storage point at the i-th logistics node.

[0144] It should be noted that the storage state of the storage point refers to the current storage state in the logistics node when the goods transportation has not started yet. If it is overstocked, the storage state value corresponding to the storage point is 1; otherwise, if it is not overstocked, it corresponds to 0.

[0145] Step S323: Initialize and update the cluster center. The corresponding calculation formula is:

[0146]

[0147] where represents the updated cluster center of the healthy state or failure state; J i,(0→n+1) represents the value of the transportation state change corresponding to the i-th logistics node during the update process; h healthy / failed represents the value corresponding to the healthy state or failure state of the storage point;

[0148] Step S324: Calculate the distance between the storage point and the updated cluster center. The corresponding calculation formula is:

[0149]

[0150] where D healthy / failed represents the distance between the storage point and the updated cluster center of the healthy state or failure state.

[0151] It should be noted that the weighted Euclidean distance is used to calculate the distance between the storage point and the updated cluster center. It can assign corresponding weights to each dimension feature based on the importance of different dimension features of the storage point, making the distance calculation more accurate and enabling more effective management of the storage point.

[0152] Step S325: Obtain the warehousing status value of the logistics node. The corresponding calculation formula is:

[0153]

[0154] where represents the warehousing status value of the i-th logistics node.

[0155] It should be noted that the warehousing status value of the logistics node is calculated through distance normalization, that is, the health status or failure status of the logistics node is accurately measured and normalized to a unified scale, so as to track the warehousing status in real time, ensure that the inventory level of the warehousing point matches the customer demand, and reduce the risks of overstocking and out-of-stock.

[0156] Step S33: Generate a node dynamic judgment mechanism by combining the transportation status value and the warehousing status value of the logistics node.

[0157] Specifically, in step S33, it includes:

[0158] Step S331: Calculate the health degree of the logistics node. The corresponding calculation formula is:

[0159]

[0160] where J i represents the current health degree of the i-th logistics node; represents the transportation status value of the i-th logistics node using the k-th transportation method; represents the warehousing status value of the i-th logistics node;

[0161] Step S332: Set a screening threshold. When the health degree is higher than the screening threshold, it indicates that the current logistics node is in a normal state; when the health degree is lower than the screening threshold, it indicates that the current logistics node is in an abnormal state.

[0162] Preferably, in this embodiment, the screening threshold is set to J th = 0.0785.

[0163] It can be explained that J i represents the current health degree of the i-th logistics node. The higher this value is than the screening threshold, the more normal the corresponding i-th logistics node is; on the contrary, when this value is lower than the screening threshold, the logistics node is abnormal at this time. In addition, for the transportation status value and the warehousing status value if any of the two status values is lower than the screening threshold J th, it is determined that the status of the current corresponding logistics node is abnormal. This setting is used to prevent the omission or misjudgment of the health status of the logistics node, so as to more accurately confirm the status of the current logistics node. Among them, the normal status of the logistics node means that during the goods transportation process, the transportation and warehousing functions of the logistics node are in good condition, can meet the logistics needs, and ensure the logistics efficiency. On the contrary, the abnormal status of the logistics node refers to environmental factors such as transportation delays and substandard warehousing conditions.

[0164] Further, in step S4, it includes:

[0165] Step S41: Use the differential mutation method to mutate and improve the EPO algorithm. Based on the logistics node dynamic judgment mechanism, judge whether a new exploration direction is obtained. If so, combine the multi-objective cost function to determine the new exploration direction, and calculate the distance between the position of the transportation vehicle used during the transportation process and the final logistics node of the corresponding transportation path;

[0166] Step S42: Iterate according to the new exploration direction and distance until the iteration is successful, and output the optimal carrier combination number, transportation mode, and logistics node number during the goods transportation process to obtain the optimal path planning result.

[0167] For better illustration, use the differential mutation method to mutate and improve the EPO algorithm, that is, by introducing the difference information between individuals in the population to guide the mutation operation, improve the search ability and convergence speed of the algorithm. It uses the difference information between individuals in the population to generate new candidate solutions, so as to increase the diversity of the population, avoid the algorithm falling into the local optimal solution prematurely, and improve the efficiency and quality of global search. That is, the improved EPO algorithm can dynamically adjust the optimization direction according to environmental changes at different stages, dynamically avoid failed logistics nodes, and real-time adjust the path planning scheme, so as to obtain the overall cost-optimal goods transportation path planning scheme.

[0168] Further, in step S41, it includes:

[0169] Step S411: If the current logistics node is abnormal, it means that a new exploration direction needs to be obtained. Combine the multi-objective cost function to determine the new exploration direction. The corresponding calculation formula is:

[0170]

[0171] Among them, σ represents the new exploration direction determined based on the current abnormal logistics node; F(X) represents the multi-objective cost function; ξ represents the random exploration weight; T t represents the temperature change function of the current logistics node; r represents the rate of controlling temperature decay; It represents the distance between the current abnormal logistics node and the candidate logistics node; T represents the time taken to obtain a new exploration direction; max Gen represents the maximum number of iterations; gen represents the number of iterations for determining the temperature change of the current logistics node; x represents an adjustable parameter for controlling the maximum number of iterations.

[0172] Make an explanation. It represents the gradient direction of the optimal multi-objective cost function. It indicates that the current logistics node is abnormal, that is, the logistics node fails, and exploration is carried out among the subsequent logistics nodes in the transportation process. Its temperature decays as the distance increases. Among them, ξ represents the random exploration weight, which is used to control the exploration in non-optimal directions; T represents the time taken to obtain a new exploration direction. When the exploration radius in searching for new logistics nodes is greater than the radius of the formed multimodal logistics network, T = 0; when the exploration radius is less than or equal to the radius of the multimodal logistics network, T = 1.

[0173] Step S412: Determine the trend of the transportation vehicle used in the transportation process to move towards the optimal logistics node, and calculate the distance between the position of the transportation vehicle used in the transportation process and the final logistics node of the corresponding transportation path. The corresponding calculation formula is:

[0174]

[0175] Among them, It represents the distance between the position of the transportation vehicle currently used in the transportation process and the final logistics node of the corresponding transportation path; S represents the trend of the transportation vehicle currently used in the transportation process to move towards the optimal logistics node; It represents a parameter used to avoid conflicts between the currently used transportation vehicle and the next logistics node; It represents the position of the final logistics node of the transportation path; It represents the position of the transportation vehicle currently used in the transportation process; N unvisted 、N visted respectively represent the logistics nodes that the transportation vehicle currently used in the transportation process has not passed through and the logistics nodes that have already been passed through; respectively represent the positions of the logistics nodes that have not been passed through and the logistics nodes that have already been passed through; f and l respectively represent the exploration and extension parameters in the iterative process of the EPO algorithm; M represents the control parameter for avoiding conflicts between the logistics nodes passed through during the transportation process; T1 represents the temperature change function of the EPO algorithm; It represents the position of the current logistics node; It represents the position of other nodes in the transportation process; Rand represents a random number uniformly distributed on [0,1].

[0176] It is explained that S represents the tendency of the currently adopted means of transportation to move towards the optimal logistics node during transportation. It can reduce the attraction influence brought by the logistics nodes that have been passed through during transportation, prevent exploring towards the logistics nodes that have already been passed through when searching for the optimal logistics node, and avoid falling into a local optimal solution, that is, strengthen the attraction of all unpassed logistics nodes. Preferably, It represents a parameter used to avoid conflicts between the currently adopted means of transportation and the next logistics node. In this embodiment, its value ranges are [2, 3] and [1.5, 2] respectively.

[0177] Further, in step S42, it includes:

[0178] Step S421: Mutate and improve the EPO algorithm according to the new exploration direction and distance to determine the EPO model. The corresponding calculation formula is:

[0179]

[0180] Among them, represents the EPO model; J represents the health degree of the logistics node before exploration; represents the logistics node that best meets the customer needs in the multi-objective function among the logistics nodes adjacent to the current logistics node; represents the position of the newly planned logistics node after determining the new exploration direction;

[0181] Step S422: Iterate based on the EPO model. If the iteration is successful, output the optimal carrier combination number, transportation mode, and the logistics node numbers during the goods transportation process to determine the optimal path planning result; if the iteration fails, ignore the objective function with the lowest weight parameter based on the multi-objective function and re-iterate using the EPO model until the optimal path planning result is obtained successfully.

[0182] It should be noted that when the EPO model is adopted, during the convergence process of the algorithm, invalid logistics nodes can be avoided in a timely manner, and a new transportation route with the lowest relative cost can be planned among the remaining logistics nodes. Specifically, based on customer requirements, an initial carrier combination is preset, and any transportation mode is selected. The number of each logistics node during the goods transportation process is recorded. Based on the EPO model, the optimal route planning result is found from the multimodal transportation logistics network, that is, during the iteration process, an optimal solution that meets all constraint conditions is found, namely the optimal route planning result, and the carrier combination number, transportation mode, and the number of logistics nodes during the goods transportation process corresponding to the optimal solution are output. Correspondingly, if the optimal solution cannot be found during the iteration process, that is, the iteration fails, the objective function with the lowest weight parameter in the multi-objective cost function is ignored, the constraint conditions are simplified, the logistics nodes are updated to focus on the objective function that has a greater impact on the overall route planning, and then the EPO model is used for iteration until an optimal route planning result is obtained.

[0183] For better illustration, the route planning results output by a dynamic progressive route planning method based on the EPO algorithm proposed in this application are compared with the unimproved EPO algorithm and the traditional genetic algorithm model (GA, Genetic Algorithm) to obtain the costs, operation times, and convergence times of different algorithms.

[0184] Please refer to Figure 3 , which shows a comparison chart of transportation costs between a dynamic progressive route planning method based on the EPO algorithm provided in an embodiment of the present invention and different algorithms. It should be noted that for the solution of the global cost by the GA algorithm, the overall trend does not change with the increase in the number of logistics nodes, while the advantages of the route planning method proposed in this application and the unimproved EPO algorithm lie in global optimization and global convergence, that is, as the number of available logistics nodes in the logistics network increases, both algorithms can perform information trade-offs among different logistics nodes, fully consider various factors to optimize the route planning, and effectively reduce the overall transportation cost. However, the route planning method proposed in this application saves approximately 2.1% - 6.3% compared with the unimproved EPO algorithm in terms of cost, and saves 9.2% - 16.8% compared with the GA algorithm as the number of logistics nodes increases. Therefore, the route planning method proposed in this application can more accurately balance route selection and global resource allocation, has stronger adaptability and optimization ability in a more complex logistics network environment, and improves the transportation efficiency and economy of the supply chain.

[0185] Please combine Figure 4 and Figure 5, which respectively shows the comparison diagrams of the operation time and the convergence time between a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention and different algorithms; it can be explained that during the cargo transportation process in a logistics network with a large number of logistics nodes and a complex and changeable environment, the convergence speeds of the three algorithms are similar, while the operation time of the path planning method proposed in this application is better than that of the GA algorithm and the EPO algorithm. It has higher flexibility during the exploration and convergence process of the search space, enabling it to achieve a better path selection while ensuring the operation efficiency; in terms of operation time, due to the introduction of an adaptive path optimization strategy in the path planning method proposed in this application, the initial calculation amount is large, but as the number of unpassed logistics nodes becomes smaller and smaller, the operation speed in the later stage is significantly improved, with an average reduction of 17.8% in operation time; in terms of convergence performance, the path planning method proposed in this application optimizes the temperature control and the individual movement strategy, enabling it to ensure the global search ability while ensuring that the convergence speed is not too slow; that is, compared with the GA and the unimproved EPO algorithms, the path planning method proposed in this application can maintain a faster operation speed and a better optimization result in a large-scale logistics network, fully verifying its advantages in complex logistics path planning.

[0186] Please refer to Figure 6 , which shows the adaptability comparison diagram of a dynamic progressive path planning method based on the EPO algorithm provided by an embodiment of the present invention and different algorithms with different numbers of logistics nodes; Python is used to calculate the adaptability function of the three algorithms in scenarios with different numbers of transportation logistics nodes. The higher the adaptation value, the better the performance of the corresponding algorithm. Among them, Figures a, b, c, and d respectively correspond to the adaptability curves with 40, 60, 80, and 100 logistics nodes. The horizontal axis represents the iteration times of the algorithm, and the vertical axis represents the adaptation value. GA is the traditional genetic algorithm, EPO is the unimproved emperor penguin algorithm, and DPA is a dynamic progressive path planning method based on the EPO algorithm provided in this application; different algorithms and different numbers of logistics nodes show different advantages and convergence speeds.

[0187] Specifically, when the number of logistics nodes is 40, both the GA algorithm and the EPO algorithm have faster global search and convergence speeds, and the stabilized fitness function values are both higher than 0.7. However, for the path planning method proposed in this application, when performing convergence operations and finding the heat direction, since more judgment factors need to be considered, the number of iterations in the early stage is relatively large, resulting in a slightly lagging optimization rate, and the overall decision-making quality lags behind by 14.2%-18.7%. When the number of logistics nodes is 60, as the number of logistics nodes increases, uncertain events during transportation occur more frequently. Perhaps due to the failure of some logistics nodes, the algorithm may fall into a local optimum or be unable to break out of the judgment loop. At this time, the convergence speeds of the three algorithms are similar, but the fitness function value of the path planning method proposed in this application is higher, with an optimization of about 12.3%-26.5%, indicating better global optimization ability. When the number of logistics nodes continues to increase, that is, when the corresponding numbers are 80 and 100, due to the larger number of logistics nodes, the logistics network becomes more complex and the changing factors also increase relatively. The path planning method proposed in this application needs to continuously iterate to calculate new optimal solutions, and the explored heat direction is also constantly changing. Therefore, for the early stage of transportation when there are still many remaining logistics nodes, the convergence speed and fitness function value of the algorithm are worse than those of the other GA and unimproved EPO algorithms. However, after determining the heat convergence direction in the later stage of the operation, the convergence speed gradually increases, and the fitness function value is 11.6%-22.4% better than that of the GA and unimproved EPO algorithms. It should be noted that the path planning method proposed in this application can effectively avoid the local optimum trap through its dynamic progressive strategy when facing high-dimensional complex networks and dynamic environments, demonstrating stronger global optimization ability. That is, during the goods transportation process in a logistics network with complex and changing environments, the method provided in this application has stronger adaptability and global optimization ability, and can effectively reduce transportation costs and shorten decision-making time.

[0188] Understandably, this application adopts a multi-objective cost function, that is, on the premise of meeting the timeliness and reliability of cargo transportation, it adapts to the changes in different transportation scenarios to meet the diverse needs of customers; combined with the dynamic status evaluation mechanism of logistics nodes, it monitors dynamic environmental factors such as traffic congestion, equipment failure, and weather changes in real time, so that the transportation path can be adaptively adjusted, effectively avoid failed nodes, and ensure the continuity and stability of cargo transportation, that is, introduces a dynamic path adjustment strategy, so that the transportation path can adapt to environmental changes during the iteration process, improve the flexibility and optimization ability of path planning, and realize the coordinated optimization of transportation cost, time efficiency and service quality; cooperate with the global planning and direction-finding capabilities of the EPO algorithm, and use differential mutation to optimize the original EPO algorithm, so that the path planning model can dynamically adjust the optimization direction according to changes in the transportation environment, improve the optimization efficiency, and realize the improvement of the intelligent decision-making ability of the path optimization algorithm, and has stronger adaptability and stability for complex logistics networks with a huge number of logistics nodes.

[0189] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0190] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.

Claims

1. A dynamic progressive path planning method based on the EPO algorithm, characterized in that, The method includes: Obtaining logistics nodes based on goods transportation, collecting customer demands, and matching corresponding goods attributes; According to the customer demands and the corresponding goods attributes, successively constructing objective functions for transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost, respectively determining weight parameters, and integrating and establishing a multi-objective cost function; Respectively determining the transportation status value and storage status value of the logistics nodes during the goods transportation process, and integrating and generating a dynamic judgment mechanism for logistics nodes; Through the multi-objective cost function and the dynamic judgment mechanism of logistics nodes, based on the differential mutation method, mutate and improve the EPO algorithm, and output the optimal carrier combination number, transportation mode, and the logistics node number during the goods transportation process to obtain the optimal path planning result.

2. The dynamic progressive path planning method based on the EPO algorithm according to claim 1, characterized in that According to the customer demands and the corresponding goods attributes, successively constructing objective functions for transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost, including: Constructing the objective function of transportation cost, and the corresponding calculation formula is: Among them, C (x) represents the objective function of transportation cost; M represents the total number of transportation modes in cargo transportation; N k represents the total number of logistics nodes; β i represents all reachable subsequent nodes of the i-th logistics node during transportation; d ij represents the actual transportation distance between the i-th logistics node and the j-th logistics node; represents the unit distance cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; represents the transshipment cost of transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; Constructing the objective function of time cost, and the corresponding calculation formula is: Among them, T (x) represents the objective function of time cost; represents the transportation time of using the k-th transportation mode from the i-th logistics node to the j-th logistics node; represents the transportation time corresponding weight parameter; represents the transshipment time of using the k-th transportation mode from the i-th logistics node to the j-th logistics node; represents the transshipment time corresponding weight parameter; represents the time window cost of using the k-th transportation mode from the i-th logistics node to the j-th logistics node; Based on the goods transportation process, using fuel vehicles and / or new energy vehicles for road transportation, constructing the objective function of carrier carbon emission cost, and the corresponding calculation formula is: Among them, R (x) represents the objective function of the carrier's carbon emission cost; represents the carbon emission cost of transporting from the i-th logistics node to the j-th logistics node by fuel vehicle; represents the carbon emission cost of transporting from the i-th logistics node to the j-th logistics node by new energy vehicle; Constructing the objective function of carrier service quality cost, and the corresponding calculation formula is: Among them, Q (x) represents the objective function of the carrier service quality cost; represents the quantity of goods that need to be preferentially transported from the i-th logistics node to the j-th logistics node by the k-th transportation mode; q total represents the quantity of goods that need to be preferentially transported for all logistics nodes; represents the cold-chain goods transported from the i-th logistics node to the j-th logistics node by the k-th transportation mode; T cold-maxl represents the cold-chain goods for all logistics nodes; represents the quantity of orders that can be promptly responded to when transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; q allowed represents the quantity of orders for all logistics nodes; represents the transportation accuracy rate when transporting from the i-th logistics node to the j-th logistics node by the k-th transportation mode; w5, w6, w7, and w8 respectively represent the service quality of preferentially transporting goods, the service quality of transporting cold-chain goods, the service quality of order response rate, and the service quality of transportation accuracy rate.

3. The dynamic progressive path planning method based on the EPO algorithm according to claim 2, characterized in that Respectively determining weight parameters, and integrating and establishing a multi-objective cost function, and the corresponding calculation formula is: F(X) = w1C (x) + w2T (x) + w3R (x) + w4Q (x) Among them, F(X) represents the multi-objective cost function; w1, w2, w3, and w4 respectively represent the corresponding weight parameters determined according to customer requirements; C (x) , T (x) , R (x) , and Q (x) respectively represent the objective functions of transportation cost, time cost, carrier carbon emission cost, and carrier service quality cost.

4. A dynamic progressive path planning method based on the EPO algorithm according to claim 1, characterized in that Respectively determining the transportation status value and storage status value of the logistics nodes during the goods transportation process, and integrating and generating a dynamic judgment mechanism for logistics nodes, including: Based on the goods attributes, successively determining the reliability of different transportation modes, the maintenance probability for logistics nodes, and the influence degree of environmental factors on the transportation of logistics nodes, to obtain the transportation status value of the logistics nodes when different transportation modes are adopted; Dividing the storage points at the logistics nodes into a healthy state and a failure state, respectively obtaining the cluster centers, determining the storage status of the storage points, updating the cluster centers, and calculating the distance between the storage points and the updated cluster centers to obtain the storage status value of the logistics nodes; Combining the transportation status value and storage status value of the logistics nodes to generate a node dynamic judgment mechanism.

5. The dynamic progressive path planning method based on the EPO algorithm according to claim 4, characterized in that Based on the goods attributes, successively determining the reliability of different transportation modes, the maintenance probability for logistics nodes, and the influence degree of environmental factors on the transportation of logistics nodes, to obtain the transportation status value of the logistics nodes when different transportation modes are adopted, including: Determining the reliability of different transportation modes, and the corresponding calculation formula is: Among them, represents the reliability of the i-th logistics node when adopting the k-th transportation mode; C i,k represents the load upper limit of the k-th transportation mode at the i-th logistics node; L i,k represents the current load of the k-th transportation mode at the i-th logistics node; Determining the maintenance probability for logistics nodes, and the corresponding calculation formula is: Among them, represents the maintenance probability of the logistics equipment in the \(i\)-th logistics node when adopting the \(k\)-th transportation mode; \(N\) represents the number of logistics equipment in the \(i\)-th logistics node; represents the remaining life of the logistics equipment in the \(i\)-th logistics node; represents the total life of the logistics equipment in the \(i\)-th logistics node; Determining the influence degree of environmental factors on the transportation of logistics nodes, where the environmental factors include weather influence and traffic congestion, and the corresponding calculation formula is: Among them, represents the influence degree of environmental factors on the transportation of the i-th logistics node when adopting the k-th transportation mode; T i,k represents the traffic congestion index at the i-th logistics node when adopting the k-th transportation mode; W i,k represents the weather influence index at the i-th logistics node when adopting the k-th transportation mode; α and δ respectively represent the weights of the traffic congestion index and the weather influence index; Obtaining the transportation status value of the logistics nodes when different transportation modes are adopted, and the corresponding calculation formula is: Among them, represents the transportation status value of the k-th transportation mode at the i-th logistics node.

6. The dynamic progressive path planning method based on the EPO algorithm according to claim 4, characterized in that Dividing the storage points at the logistics nodes into a healthy state and a failure state, respectively obtaining the cluster centers, determining the storage status of the storage points, updating the cluster centers, and calculating the distance between the storage points and the updated cluster centers to obtain the storage status value of the logistics nodes, including: Obtain a feature vector based on the storage points at any logistics node, denoted as X i ={h1, h2, h3, …, h n}, where n represents the total number of storage points at the i-th logistics node. It is divided into a healthy state and a failure state according to the feature vector, and the cluster centers are obtained respectively, denoted as Determine the storage status of any storage point, and the corresponding calculation formula is: Among them, h s represents the storage status of the s-th storage point; S i,s represents the current storage volume of the s-th storage point at the i-th logistics node; represents the maximum storage volume of the s-th storage point at the i-th logistics node; Initialize and update the cluster center, and the corresponding calculation formula is: Among them, represents the cluster center after the update of the health state or failure state; J i,(0→n+1) represents the transportation state value corresponding to the change in the update process of the i-th logistics node; h healthy / failed represents the value corresponding to the health state or failure state of the storage point; Calculate the distance between the storage point and the updated cluster center, and the corresponding calculation formula is: Among them, D healthy / failed represents the distance between the storage point and the cluster center after the health status or failure status is updated; Obtain the storage status value of the logistics node, and the corresponding calculation formula is: Among them, represents the warehousing status value of the i-th logistics node.

7. A dynamic progressive path planning method based on the EPO algorithm according to claim 4, characterized in that Generate a node dynamic judgment mechanism by combining the transportation status value and the storage status value of the logistics node, including: Calculate the health degree of the logistics node, and the corresponding calculation formula is: Among them, J i represents the current health level of the i-th logistics node; represents the transportation status value of the k-th transportation mode at the i-th logistics node; represents the warehousing status value of the i-th logistics node; Set a screening threshold. When the health degree is higher than the screening threshold, it means that the current logistics node is in a normal state; when the health degree is lower than the screening threshold, it means that the current logistics node is in an abnormal state.

8. A dynamic progressive path planning method based on the EPO algorithm according to claim 1, characterized in that Through the multi-objective cost function and the logistics node dynamic judgment mechanism, based on the differential mutation method, mutate and improve the EPO algorithm, and output the optimal carrier combination number, transportation mode, and the logistics node number in the process of goods transportation to obtain the optimal path planning result, including: Use the differential mutation method to mutate and improve the EPO algorithm. Based on the logistics node dynamic judgment mechanism, judge whether a new exploration direction is obtained. If so, combine the multi-objective cost function to determine the new exploration direction, and calculate the distance between the position of the transportation vehicle used in the transportation process and the final logistics node of the corresponding transportation path; Iterate according to the new exploration direction and distance until the iteration is successful, and output the optimal carrier combination number, transportation mode, and the logistics node number in the process of goods transportation to obtain the optimal path planning result.

9. A dynamic progressive path planning method based on the EPO algorithm according to claim 8, characterized in that Use the differential mutation method to mutate and improve the EPO algorithm. Based on the logistics node dynamic judgment mechanism, judge whether a new exploration direction is obtained. If so, combine the multi-objective cost function to determine the new exploration direction, and calculate the distance between the position of the transportation vehicle used in the transportation process and the final logistics node of the corresponding transportation path, including: If the current logistics node is abnormal, it means that a new exploration direction needs to be obtained. Combine the multi-objective cost function to determine the new exploration direction, and the corresponding calculation formula is: Among them, σ represents a new exploration direction determined based on the current abnormal logistics node; F(X) represents a multi-objective cost function; ξ represents a random exploration weight; T t represents the temperature change function of the current logistics node; r represents the rate of controlling temperature decay; represents the distance between the current abnormal logistics node and the candidate logistics node; T represents the time taken to obtain a new exploration direction; max Gen represents the maximum number of iterations; gen represents the number of iterations for determining the temperature change of the current logistics node; x represents an adjustable parameter for controlling the maximum number of iterations; Determine the trend of the transportation vehicle used in the transportation process to move towards the optimal logistics node, and calculate the distance between the position of the transportation vehicle used in the transportation process and the final logistics node of the corresponding transportation path, and the corresponding calculation formula is: Among them, represents the distance between the location of the currently adopted means of transportation during transportation and the final logistics node of the corresponding transportation route; S represents the tendency of the currently adopted means of transportation to move towards the optimal logistics node during transportation; represents a parameter used to avoid conflicts between the currently adopted means of transportation and the next logistics node; represents the location of the final logistics node of the transportation route; represents the location of the currently adopted means of transportation during transportation; N unvisted 、N visted respectively represent the logistics nodes not passed by the currently adopted means of transportation and the logistics nodes that have been passed during transportation; respectively represent the locations of the logistics nodes not passed and the logistics nodes that have been passed; f and l respectively represent the exploration and extension parameters in the iterative process of the EPO algorithm; M represents the control parameter for avoiding conflicts between the logistics nodes passed during transportation; T1 represents the temperature change function of the EPO algorithm; represents the location of the current logistics node; represents the locations of other nodes during transportation; Rand represents a random number uniformly distributed on [0,1].

10. A dynamic progressive path planning method based on the EPO algorithm according to claim 9, characterized in that Iterate according to the new exploration direction and distance until the iteration is successful, and output the optimal carrier combination number, transportation mode, and the logistics node number in the process of goods transportation to obtain the optimal path planning result, including: Mutate and improve the EPO algorithm according to the new exploration direction and distance, and determine the EPO model, and the corresponding calculation formula is: Among them, represents the EPO model; J represents the health of the logistics node before exploration; represents the logistics node that best meets the customer needs in the multi-objective function among the logistics nodes adjacent to the current logistics node; represents the position of the newly planned logistics node after determining the new exploration direction; Iterate based on the EPO model. If the iteration is successful, output the optimal carrier combination number, transportation mode, and the logistics node number in the process of goods transportation to determine the optimal path planning result; if the iteration fails, ignore the objective function with the lowest weight parameter based on the multi-objective function, and use the EPO model to re-iterate until the iteration is successful to obtain the optimal path planning result.

Images