Intelligent Goods Delivery Route Optimization Method

Abstract

This invention discloses an intelligent goods delivery route optimization method, relating to the field of intelligent logistics and delivery technology. The method comprises the following components: S1, metaverse space construction; S2, user role creation and access; S3, immersive collaborative route planning; and S4, route optimization algorithm and data-driven approach. By constructing a logistics and delivery metaverse space, this invention achieves real-time synchronization between the real world and the virtual space, including synchronized updates of order information, vehicle location, and traffic conditions. This helps logistics dispatchers perform more accurate route planning in the virtual environment. Delivery personnel can also view and provide feedback on route planning schemes through virtual roles before executing tasks, thereby achieving multi-party collaborative immersive route planning. This method can significantly improve the efficiency and accuracy of goods delivery, reducing delivery delays or errors caused by information asymmetry or poor communication.

Description

Technical Field

[0001] This invention relates to the field of intelligent logistics and distribution technology, specifically to an intelligent cargo delivery route optimization method. Background Technology

[0002] With the booming development of e-commerce and the rapid expansion of the logistics industry, the efficiency and accuracy of goods delivery have become key indicators for measuring the quality of logistics services. In the traditional logistics delivery model, due to the untimely and inaccurate transmission of information and the limitations of communication channels, logistics dispatchers often find it difficult to make optimal delivery route planning, resulting in extended delivery time, increased costs, and decreased customer satisfaction.

[0003] In traditional technologies, logistics delivery route planning mainly relies on dispatchers' experience and limited data analysis tools. This method has obvious limitations. First, due to the lack of real-time and comprehensive data support, dispatchers find it difficult to accurately grasp information such as vehicle location, traffic conditions, and dynamic changes in orders, resulting in inaccurate route planning. Second, traditional communication methods carry the risk of delays and misunderstandings. Feedback from delivery personnel and customers during the delivery process is difficult to convey to dispatchers in a timely and accurate manner, affecting the timeliness and effectiveness of route adjustments.

[0004] In summary, traditional logistics and distribution route planning technologies suffer from drawbacks such as untimely and inaccurate information transmission and limited communication channels, leading to low delivery efficiency, increased costs, and decreased customer satisfaction. Therefore, the development of intelligent cargo delivery route optimization methods is particularly important. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide an intelligent goods delivery route optimization method. It can realize real-time synchronization between the real world and the virtual space by constructing a logistics delivery metaverse space, including comprehensive updates and sharing of data such as order information, vehicle location, and traffic conditions. On this basis, advanced algorithms and data analysis technologies are used to accurately plan and dynamically adjust delivery routes, thereby significantly improving the efficiency and accuracy of goods delivery.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an intelligent goods delivery route optimization method, the specific steps of which are as follows:

[0007] S1. Metaverse Space Construction: Construct a logistics and distribution metaverse space. In this space, virtual warehouses, roads, and delivery vehicle elements are built according to actual logistics scenarios and synchronized with real-world logistics data in real time, including order information, vehicle location, and traffic conditions.

[0008] S2. User Role Creation and Access: Virtual roles are created for logistics dispatchers, delivery personnel, and customers, and corresponding terminal devices are provided to access the metaverse space. Users can move and interact freely in the metaverse space in the form of virtual roles through the terminal devices.

[0009] S3, Immersive Collaborative Route Planning: In the metaverse space, logistics dispatchers formulate preliminary cargo delivery route planning schemes based on order information and route planning algorithms, and display the expected driving routes of delivery vehicles in a virtual scene;

[0010] Before carrying out delivery tasks, delivery personnel enter the metaverse space and view the cargo delivery route planning scheme formulated by the logistics dispatcher through a virtual character. At the same time, they can combine their own experience to ask questions or make suggestions about the route in the virtual scene and communicate with the logistics dispatcher in real time via voice, text or gesture.

[0011] Customers can also view the delivery progress and estimated arrival time of goods in the metaverse space. If there are special needs, they can negotiate with logistics dispatchers and deliverymen through virtual characters. Based on feedback from multiple parties, the logistics dispatcher will adjust the goods delivery route planning scheme in real time in the metaverse space and display the effect of the adjustment in a visual way for all parties to confirm.

[0012] S4. Path Optimization Algorithm and Data-Driven Approach: During the operation of the metaverse space, continuously collect interaction data from all parties, operational data in virtual scenarios, and real-time logistics data from the real world. Utilize machine learning algorithms to analyze and learn from this data, continuously optimize the path planning algorithm, and make the cargo delivery path planning scheme more in line with actual needs, thereby improving delivery efficiency.

[0013] Furthermore, the construction of the logistics distribution metaverse space employs a fusion technology of 3D reconstruction and digital twins, specifically including:

[0014] Using drones equipped with LiDAR to perform omnidirectional scanning of actual logistics scenarios and acquire point cloud data. ,in Indicating the first point in the point cloud data The three-dimensional coordinates of the points , This represents the total number of point cloud data.

[0015] Use a high-definition camera to capture scene image data , For the first The scene data is aligned with point cloud data from different perspectives using a point cloud registration algorithm. Based on a multi-view stereo matching algorithm, a high-precision 3D model is constructed by combining point cloud data and image data. Digital twin technology is introduced to establish a real-time mapping relationship between the real logistics scene and the virtual meta-universe space.

[0016] For warehouse goods entry and exit information and real-time road traffic flow data, data is transmitted at time intervals via a data interface. To ensure the security and immutability of data transmission, synchronous updates are performed using blockchain technology for encrypted storage and verification. Each data block contains the hash value of the previous data block, the hash value of the current data, and a timestamp, thus constructing a blockchain data chain. ,in Represents the first in the blockchain One data block, , This represents the total number of data blocks.

[0017] Furthermore, when the delivery personnel and logistics dispatchers engage in real-time voice, text, or gesture-based interactive communication, an interaction optimization mechanism based on semantic understanding and sentiment analysis is employed, specifically:

[0018] Natural language processing (NLP) technology is used to segment, tag, and syntactically analyze the voice or text information input by delivery personnel, converting the text into word vector representations. ,in For the first Vector representation of each word , This represents the total number of word vectors.

[0019] A sentiment analysis model is constructed using an improved recurrent neural network structure, incorporating an attention mechanism to weight word vectors and calculate sentiment scores. The calculation formula is: ,in Attention weights are calculated by determining the similarity between word vectors and the global semantic vector. For word vectors Feature extraction function, For activation functions;

[0020] Logistics dispatchers judge delivery personnel's feedback attitude based on their emotional scores and combine this with semantic understanding results to make targeted adjustments to the cargo delivery route planning scheme. At the same time, visual emotional feedback icons are set in the interactive interface to display the emotional state of delivery personnel in real time, helping dispatchers make decisions more quickly and accurately.

[0021] Furthermore, the path planning algorithm employs a multi-objective path planning algorithm based on improved particle swarm optimization, which incorporates delivery time... Delivery costs Carbon emissions As the optimization objective, construct the objective function. ,in , , These are the weighting coefficients, and The method for determining the weighting coefficients is as follows: By analyzing historical delivery data, a judgment matrix is ​​constructed using the analytic hierarchy process (AHP). Logistics experts are invited to score the relative importance of three objectives: delivery time, delivery cost, and carbon emissions. The weighting coefficients are then calculated. The particle's position represents the delivery path, and its velocity represents the direction and step size of path adjustments. Dynamic inertia weights are also introduced. Its calculation formula is ,in and These are the maximum and minimum values ​​of the inertia weight, respectively. This represents the current iteration number. To maximize the number of iterations, an elite learning strategy is adopted, selecting the best particle in each iteration as the elite particle to guide other particles to learn its excellent features, thereby accelerating algorithm convergence and improving the efficiency and quality of path planning.

[0022] Furthermore, when the customer negotiates with the logistics dispatcher and delivery person in the metaverse space, it includes:

[0023] A demand response mechanism based on preference mining is adopted, which constructs a customer preference model by using customer operational behavior data in the metaverse space and historical order data;

[0024] Association rule mining algorithms are used to uncover the relationships between customer needs and delivery elements, and to calculate customer preferences for different delivery options. The calculation formula is:

[0025]

[0026] in, For the first The weights of the associated factors were obtained by training on historical data using regression analysis methods in machine learning. For the first The response values ​​of each related factor;

[0027] Based on customer preferences, logistics dispatchers prioritize generating delivery route adjustment plans that align with those preferences, and then showcase the advantages and expected results of these plans to customers within the metaverse space, thereby increasing customer satisfaction and acceptance of the delivery solutions.

[0028] Furthermore, the real-time adjustment of the cargo delivery route planning scheme in the metaverse space includes:

[0029] A local optimization algorithm based on a fusion of simulated annealing and tabu search is adopted, using the difference between the adjusted path and the original path, and the change in the objective function value of the adjusted path as evaluation indicators. First, an initial adjustment scheme is randomly generated, and the objective function value of the current scheme is calculated. Then, neighborhood schemes are generated according to certain rules, and the objective function value of the neighborhood schemes is calculated. ,like If so, then accept the neighborhood scheme as the new current scheme. Then, based on probability Accept the proposal, in which For temperature parameters, Represents the natural constant, which increases with the number of iterations. Attenuation, The initial temperature. The cooling coefficient, To determine the number of iterations, a tabu search strategy is introduced, and a tabu table is set up to record recently visited solutions to avoid the algorithm getting stuck in local optima. By combining simulated annealing and tabu search, a better local adjustment solution can be quickly found during the path adjustment process after feedback from multiple parties, thereby improving the efficiency and quality of path optimization.

[0030] Furthermore, the process of analyzing and learning data using machine learning algorithms includes: employing a path optimization algorithm based on deep reinforcement learning to construct an agent-environment interaction model, where the agent is the path optimization algorithm, the environment is the metaverse space and the real-world logistics scenario, and the state of the agent... The actions of the intelligent agent are composed of a multi-dimensional vector consisting of order information, vehicle location, traffic conditions, and historical route data. The reward function is used to adjust the cargo delivery route planning scheme. The design comprehensively considers factors such as the reduction in delivery time, the reduction in cost, and the improvement in customer satisfaction. The calculation formula is as follows:

[0031]

[0032] in , , The weighting coefficients were determined through regression analysis of historical delivery data. To reduce delivery time, To reduce costs and increase volume, To improve customer satisfaction, the agent selects actions based on the current state, interacts with the environment to obtain rewards and new states, and learns through continuous trial and error, using a deep neural network to fit a state-action value function. We employ experience replay and target network techniques to improve the stability and convergence speed of the algorithm, thereby enabling continuous optimization of the path planning algorithm.

[0033] Furthermore, the process of creating virtual roles for logistics dispatchers, delivery personnel, and customers, and providing corresponding terminal devices to access the metaverse space, includes:

[0034] By adopting low-latency access technology based on the convergence of edge computing and 5G, edge computing nodes are deployed in locations such as logistics parks and delivery stations. Terminal devices preprocess and analyze some data at the edge computing nodes to reduce the amount of data transmission. Taking advantage of the high bandwidth and low latency characteristics of the 5G network, the processed data is quickly transmitted to the server.

[0035] The terminal device receives metaverse space data returned by the server. To ensure the reliability of data transmission, a multi-path transmission protocol is adopted to divide the data into multiple data packets and transmit them through different 5G network channels. The data is then reassembled at the receiving end. Interactive data with high real-time requirements is prioritized for transmission through low-latency 5G network slices to ensure a smooth interactive experience for users in the metaverse space and achieve efficient immersive collaborative path planning.

[0036] Furthermore, when displaying the expected route of the delivery vehicle in a virtual scene, a visualization enhancement technology based on the fusion of augmented reality and virtual reality is employed, including:

[0037] For dispatchers and customers, in the virtual scene displayed by VR devices, the ray tracing algorithm simulates the real lighting effect, making the delivery vehicles and road elements present a realistic three-dimensional visual effect. At the same time, the particle system is used to simulate the dynamic effects of exhaust gas and dust during vehicle operation, enhancing the realism of the scene.

[0038] For AR glasses used by delivery drivers, virtual delivery route guidance information is overlaid on the real scene. Spatial registration technology is used to ensure accurate matching between virtual routes and real roads. When delivery vehicles encounter special situations, the affected road sections are highlighted with bright and flashing effects in VR devices and AR glasses, and the adjusted route information is displayed simultaneously to help all parties understand the delivery route situation more intuitively and accurately, and to assist in decision-making and execution.

[0039] Compared with existing technologies, this intelligent goods delivery route optimization method has the following advantages:

[0040] I. This invention constructs a metaverse space for logistics and distribution, enabling real-time synchronization between the real world and the virtual space. This includes the synchronous updating of order information, vehicle location, and traffic data. This helps logistics dispatchers plan more accurate delivery routes in the virtual environment. Delivery personnel can also view and provide feedback on delivery route plans through virtual characters before executing tasks, thereby achieving immersive route planning through multi-party collaboration. This method can significantly improve the efficiency and accuracy of goods delivery and reduce delivery delays or errors caused by information asymmetry or poor communication.

[0041] Second, the system allows customers to view the delivery progress and estimated arrival time of their goods in real time within the metaverse space. They also have the right to negotiate with logistics dispatchers and delivery personnel through virtual roles and make special requests. The system adopts a demand response mechanism based on preference mining. Based on customer operation behavior data and historical order data, it constructs a customer preference model, thereby prioritizing the generation of route adjustment plans that meet customer preferences. This not only enhances customers' sense of participation and control over the delivery process, but also significantly improves customer satisfaction and recognition of the delivery plan.

[0042] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0044] Figure 1 A flowchart illustrating the process of optimizing intelligent goods delivery routes. Detailed Implementation

[0045] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0046] Example 1

[0047] First, technicians used drones equipped with lidar to acquire point cloud data. ,in Indicating the first point in the point cloud data The three-dimensional coordinates of each point were used to perform a comprehensive scan of the logistics warehouse in Pudong New Area and the roads within a 10-kilometer radius. The LiDAR rapidly rotated to acquire massive amounts of three-dimensional point cloud data, while high-definition cameras simultaneously captured scene image data. This adds visual detail to the virtual space. By using a point cloud registration algorithm, point cloud data acquired from different flight perspectives are precisely aligned to eliminate data deviations. Combined with a multi-view stereo matching algorithm, point clouds and images are fused to construct a high-precision 3D model, allowing details such as warehouse shelf layout and road curves to be reproduced in the virtual space.

[0048] Next, digital twin technology is introduced to establish a real-time mapping between real-world logistics scenarios and the metaverse space. Warehouse goods entry and exit information, such as a mobile phone moving from shelf A to sorting area B, will be synchronized to the virtual space in real time through a data interface. Real-time traffic flow data on roads, such as congestion on a certain road section, will immediately display a red warning at the corresponding location on the virtual road. To ensure data security and immutability, blockchain technology is used to encrypt and store the synchronized data. Each data block contains the hash value and timestamp of the preceding and following data, forming a reliable data chain. Finally, a virtual warehouse, roads, and delivery vehicles that are completely consistent with reality are built in the metaverse, and all logistics data is updated in real time.

[0049] The platform created virtual characters for 5 logistics dispatchers, 80 deliverymen, and 300 customers in the logistics dispatch center. The virtual images of the logistics dispatchers are dressed in blue work clothes, while the deliverymen are dressed in orange work clothes. Customer characters can customize their avatars to make it easier to distinguish their identities in the metaverse.

[0050] Logistics dispatchers are equipped with high-performance VR headsets and controllers. The headsets' high-definition displays can present detailed virtual scenes, and the controllers can perform operations such as dragging and clicking. Delivery personnel wear AR glasses with camera and voice functions, and virtual information can be overlaid on the lenses of the glasses. They are also equipped with smart terminals to receive messages. Customers access the system through a mobile app with a simple and intuitive interface to check delivery progress. When users put on the device or open the app, the virtual character "steps into" the metaverse space. Logistics dispatchers can overlook the entire situation in the virtual dispatch room, the virtual image of the delivery personnel appears at the entrance of the virtual warehouse, and customers can see their delivery address on the virtual map, achieving immersive access for multiple roles.

[0051] In the virtual dispatch room of the metaverse, the logistics dispatcher faces a giant holographic screen. Based on the order information aggregated by the system (including delivery address, cargo weight, delivery time requirements, etc.), they invoke the initial route planning algorithm. The formula is: [Delivery time...] Delivery costs Carbon emissions As the optimization objective, construct the objective function. ,in , , Using weighted coefficients, preliminary delivery routes are automatically generated on the virtual map. The expected driving trajectory of each delivery vehicle is displayed as a line of a different color, and key locations such as intersections and community gates are marked.

[0052] Before setting off, delivery drivers enter the metaverse through AR glasses, where the real-world scene is overlaid with the virtual route. One delivery driver responsible for the Lujiazui area discovered that a section of the route planned by the system would be closed to traffic the next day due to an event. He immediately reported to the logistics dispatcher via voice function: "Century Avenue will be closed tomorrow. We suggest you take Dongfang Road instead." The logistics dispatcher checked the real-time traffic announcement for that section in the virtual space, confirmed the situation, and adjusted the route with the delivery driver via voice communication.

[0053] Meanwhile, a customer in a certain community saw on their mobile app that their package was scheduled to arrive at 1 p.m., but they didn't get home until 2 p.m. They then sent a message to the logistics dispatcher through their virtual avatar: "Could the delivery time be changed to after 3 p.m.?" Upon receiving the request, the dispatcher, considering the delivery status of other orders in the area, replanned the route in the metaverse, shifting the customer's package delivery order backwards and showing the customer the adjusted estimated arrival time. After the customer confirmed the change, the negotiation was completed. Throughout the process, the logistics dispatcher adjusted route nodes in the metaverse in real time based on feedback from both the delivery person and the customer, and the virtual vehicle's trajectory was dynamically updated accordingly.

[0054] During the operation of the metaverse, various types of data are continuously collected: the content of every voice communication between logistics dispatchers and deliverymen, deliverymen's suggestions for route modifications, vehicle speeds and stopping times in virtual scenarios, and sudden traffic control information in the real world. This data is transmitted to the backend server and analyzed using machine learning algorithms. The formula is: Construct an agent-environment interaction model, where the agent is the path optimization algorithm, the environment is the metaverse space and the real logistics scenario, and the state of the agent... The actions of the intelligent agent are composed of a multi-dimensional vector consisting of order information, vehicle location, traffic conditions, and historical route data. The reward function is used to adjust the path planning scheme. The design comprehensively considers factors such as the reduction in delivery time, the reduction in cost, and the improvement in customer satisfaction. The calculation formula is as follows:

[0055]

[0056] in , , These are the weighting coefficients. To reduce delivery time, To reduce costs and increase volume, To improve customer satisfaction, the agent selects actions based on the current state, interacts with the environment to obtain rewards and new states, and learns through continuous trial and error, using a deep neural network to fit a state-action value function. We employ empirical replay and target network techniques to improve the stability and convergence speed of the algorithm.

[0057] Example 2

[0058] The technical team used a combination of 3D reconstruction and digital twin technology to digitally model the cold chain warehouse and surrounding roads in Chaoyang District. Drones equipped with LiDAR hovered over the warehouse, scanning the cold storage zones, shelf arrangements, and the road network within a 50-kilometer radius, generating dense point cloud data. This data was accurate down to the location of the cold storage doors and the slope of the roads. High-definition cameras simultaneously captured images of the warehouse interior and road scenes, acquiring rich texture information. Point cloud registration algorithms were used to integrate point cloud data from different angles, and then multi-view stereo matching algorithms were used to combine the point cloud data with the images to construct a 3D model that included details such as cold storage temperature zones and real-time road conditions.

[0059] After introducing digital twin technology, warehouse goods entry and exit information (such as a batch of strawberries being moved from cold storage area A to the loading port), real-time traffic flow on roads, and temperature monitoring data of refrigerated trucks (such as whether the temperature inside the truck compartment is stable at 2℃) are synchronized to the metaverse space at short intervals through data interfaces. Each data block is encrypted to ensure the authenticity and integrity of key data such as temperature data and order information. Ultimately, a dynamic cold chain logistics scenario is presented in the metaverse. When the refrigerated truck is driving on the virtual road, the virtual thermometer inside the truck compartment will display the temperature changes in real time, completely synchronized with the real vehicle.

[0060] The platform creates virtual roles for cold chain logistics dispatchers, refrigerated truck drivers, and supermarket receiving clerks. The virtual image of the logistics dispatcher wears a work badge, the driver wears a cold-weather work uniform, and the receiving clerk wears a supermarket uniform.

[0061] Logistics dispatchers use high-end VR devices to access the metaverse. The immersive experience of the devices makes them feel as if they are in a virtual dispatch center. The screens around them display the status of each refrigerated truck in real time. Refrigerated truck drivers are equipped with AR glasses and on-board terminals that integrate temperature monitoring functions. The AR glasses can display virtual routes and temperature alarms in the truck compartment during driving. Supermarket receiving clerks access the system through computers and can view the real-time location and estimated arrival time of delivery vehicles in the virtual interface. Once all users are connected, logistics dispatchers can "inspect" the stock status of each cold storage in the virtual space. The virtual image of the driver appears next to the driver's seat of the refrigerated truck, while the receiving clerks wait at the receiving point of the virtual supermarket, forming a virtual collaborative network for the entire process.

[0062] In the virtual dispatch center of the metaverse, the logistics dispatcher initiates an initial route planning algorithm based on order information from 50 supermarkets (including product type, quantity, temperature requirements, and arrival time). The algorithm fully considers the special characteristics of cold chain distribution, optimizing not only delivery time and cost but also incorporating temperature maintenance into the objective. It plans the route for each refrigerated truck on a virtual map. For example, vehicles transporting meat will prioritize routes with smooth road conditions and short travel times to reduce temperature fluctuations; vehicles transporting vegetables will avoid congested sections with frequent starts and stops to reduce energy consumption.

[0063] Before setting off, a driver responsible for transporting meat checked the route plan through AR glasses and found a tunnel in the system's planned route. Concerned that the unstable signal in the tunnel might affect temperature monitoring, he suggested to the logistics dispatcher through the text input function of the vehicle terminal: "There may be a delay in temperature data in the tunnel. I suggest taking a detour via surface roads." The logistics dispatcher checked the historical signal records of the tunnel in the metaverse and confirmed the existence of a signal blind spot. After that, he discussed with the driver to adjust the route and choose a slightly longer road with a stable signal.

[0064] A supermarket receiving clerk accessed the virtual map and saw that the vehicle delivering the supermarket's fruit was scheduled to arrive at 8:50 AM. However, the supermarket's unloading platform was only available after 9:00 AM. The clerk immediately sent a message to the logistics dispatcher via their virtual avatar: "Please adjust the arrival time to after 9:10 AM to avoid vehicle waiting." Upon receiving the request, the logistics dispatcher re-planned the vehicle's delivery sequence in the virtual map. The adjusted route required the vehicle to first deliver to another nearby supermarket before heading to the new supermarket, with an estimated arrival time of 9:15 AM. The dispatcher showed the receiving clerk the adjusted route and time, explaining that the adjustment would not affect the fruit's freshness. After the receiving clerk confirmed, the coordination was complete. Throughout the process, the logistics dispatcher updated the delivery route in real time by dragging vehicle nodes on the virtual map, and the virtual refrigerated truck's travel time and temperature change curve were displayed synchronously to ensure the adjusted plan met cold chain requirements.

[0065] During the operation of the metaverse, the system continuously collects multi-dimensional data: interaction records between logistics dispatchers and drivers, the impact of refrigerated trucks' driving speed on different road sections on the temperature of the truck compartment, and route changes caused by sudden road construction in reality. This data is transmitted to the backend and analyzed by machine learning algorithms.

[0066] The algorithm learns from historical data, such as the impact of road temperatures at different times of the year on the energy consumption of refrigerated trucks, and the temperature sensitivity of different goods, to continuously optimize route planning strategies. For example, after a period of learning, the algorithm discovers that the surface temperature of a certain section of elevated road in the summer afternoon rises due to sun exposure, which increases the energy consumption of refrigerated trucks. Therefore, when planning routes, it automatically reduces the selection of this section and prioritizes ground roads with shade. Through continuous data-driven optimization, the temperature anomaly rate of cold chain delivery has been reduced by 10%, and vehicle energy consumption has been effectively controlled, achieving a balance between preservation and cost.

[0067] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for optimizing intelligent cargo delivery routes, characterized in that, The specific steps of this method are as follows: S1. Metaverse Space Construction: Construct a logistics and distribution metaverse space. In this space, virtual warehouses, roads, and delivery vehicle elements are built according to actual logistics scenarios and synchronized with real-world logistics data in real time, including order information, vehicle location, and traffic conditions. The construction of the logistics and distribution metaverse space adopts the fusion technology of 3D reconstruction and digital twin. S2. User Role Creation and Access: Virtual roles are created for logistics dispatchers, delivery personnel, and customers, and corresponding terminal devices are provided to access the metaverse space. Users can move and interact freely in the metaverse space in the form of virtual roles through the terminal devices. S3, Immersive Collaborative Route Planning: In the metaverse space, logistics dispatchers formulate preliminary cargo delivery route planning schemes based on order information and route planning algorithms, and display the expected driving routes of delivery vehicles in a virtual scene; Before carrying out a delivery task, the delivery person enters the metaverse space and views the cargo delivery route plan formulated by the logistics dispatcher through a virtual avatar. Simultaneously, based on their own experience, they can raise questions or suggestions about the route in the virtual scenario and engage in real-time voice, text, or gesture interaction with the logistics dispatcher. The real-time voice, text, or gesture interaction between the delivery person and the logistics dispatcher employs an interaction optimization mechanism based on semantic understanding and sentiment analysis, specifically: Natural language processing (NLP) technology is used to segment, tag, and syntactically analyze the voice or text information input by delivery personnel, converting the text into word vector representations. ,in For the first Vector representation of each word , This represents the total number of word vectors. A sentiment analysis model is constructed using an improved recurrent neural network structure, incorporating an attention mechanism to weight word vectors and calculate sentiment scores. The calculation formula is: ,in Attention weights are calculated by determining the similarity between word vectors and the global semantic vector. For word vectors Feature extraction function, For activation functions; Logistics dispatchers judge the deliveryman's feedback attitude based on the emotion score, and make targeted adjustments to the goods delivery route planning scheme based on the semantic understanding results. At the same time, a visual emotion feedback icon is set in the interactive interface to display the deliveryman's emotional state in real time. Customers can also view the delivery progress and estimated arrival time of goods in the metaverse space. If there are special needs, they can negotiate with logistics dispatchers and deliverymen through virtual characters. The logistics dispatcher will adjust the goods delivery route planning scheme in real time in the metaverse space based on feedback from multiple parties. S4. Path Optimization Algorithm and Data-Driven Approach: During the operation of the metaverse space, continuously collect interaction data from all parties, operational data in virtual scenarios, and real-time logistics data from the real world. Utilize machine learning algorithms to analyze and learn from the data, continuously optimize the path planning algorithm, and make the cargo delivery path planning scheme fit actual needs.

2. The intelligent goods delivery route optimization method according to claim 1, characterized in that, The construction of the logistics and distribution metaverse space employs a fusion technology of 3D reconstruction and digital twins, specifically including: Using drones equipped with LiDAR to perform omnidirectional scanning of actual logistics scenarios and acquire point cloud data. ,in Indicating the first point in the point cloud data The three-dimensional coordinates of the points , This represents the total number of point cloud data. Use a high-definition camera to capture scene image data , For the first The scene data is aligned with point cloud data from different perspectives using a point cloud registration algorithm. Based on a multi-view stereo matching algorithm, a high-precision 3D model is constructed by combining point cloud data and image data. Digital twin technology is introduced to establish a real-time mapping relationship between the real logistics scene and the virtual metaverse space. For warehouse goods entry and exit information and real-time road traffic flow data, data is transmitted at time intervals via a data interface. Synchronous updates are performed, and blockchain technology is used to encrypt, store, and verify the synchronized data. Each data block contains the hash value of the previous data block, the hash value of the current data, and timestamp information, thus constructing a blockchain data chain. ,in Represents the first in the blockchain One data block, , This represents the total number of data blocks.

3. The intelligent goods delivery route optimization method according to claim 1, characterized in that, The path planning algorithm employs a multi-objective path planning algorithm based on improved particle swarm optimization, which considers delivery time... Delivery costs Carbon emissions As the optimization objective, construct the objective function. ,in , , These are the weighting coefficients, and The particle's position represents the delivery path, and its velocity represents the direction and step size of path adjustment, introducing dynamic inertia weights. Its calculation formula is ,in and These are the maximum and minimum values ​​of the inertia weight, respectively. This represents the current iteration number. To maximize the number of iterations, an elite learning strategy is employed, selecting the best particle from each iteration as the elite particle to guide other particles in learning its superior features. Accelerate algorithm convergence.

4. The intelligent goods delivery route optimization method according to claim 1, characterized in that, When the customer negotiates with the logistics dispatcher and delivery person in the metaverse space, it includes: A demand response mechanism based on preference mining is adopted, which constructs a customer preference model by using customer operational behavior data in the metaverse space and historical order data; Association rule mining algorithms are used to uncover the relationships between customer needs and delivery elements, and to calculate customer preferences for different delivery options. The calculation formula is: in, For the first The weights of the related factors For the first The response values ​​of each related factor; Based on customer preferences, logistics dispatchers prioritize generating delivery route adjustment plans that align with those preferences and then present the advantages and expected results of these plans to customers within the metaverse space.

5. The intelligent goods delivery route optimization method according to claim 1, characterized in that, The process of adjusting the cargo delivery route planning scheme in real time in the metaverse space includes: A local optimization algorithm based on a fusion of simulated annealing and tabu search is adopted. The difference between the adjusted path and the original path, and the change in the objective function value of the adjusted path are used as evaluation indicators. An initial adjustment scheme is randomly generated, and the objective function value of the current scheme is calculated. It generates neighborhood schemes and calculates the objective function value of the neighborhood schemes. ,like If so, then accept the neighborhood scheme as the new current scheme. Then, based on probability Accept the proposal, in which, For temperature parameters, Represents the natural constant, which increases with the number of iterations. Attenuation, The initial temperature. The cooling coefficient, The iteration count is used; at the same time, a tabu search strategy is introduced, and a tabu table is set to record recently visited solutions to avoid the algorithm getting trapped in local optima.

6. The intelligent goods delivery route optimization method according to claim 1, characterized in that, The process of using machine learning algorithms to analyze and learn data includes: A path optimization algorithm based on deep reinforcement learning is used to construct an agent-environment interaction model. The agent is the path optimization algorithm, the environment is the metaverse space and the real logistics scenario, and the state of the agent is... The actions of the intelligent agent are composed of a multi-dimensional vector consisting of order information, vehicle location, traffic conditions, and historical route data. The reward function is used to adjust the cargo delivery route planning scheme. The design comprehensively considers factors such as the reduction in delivery time, the reduction in cost, and the improvement in customer satisfaction. The calculation formula is as follows: in , , These are the weighting coefficients. To reduce delivery time, To reduce costs and increase volume, To improve customer satisfaction, the agent selects actions based on the current state, interacts with the environment to obtain rewards and new states, and learns through continuous trial and error, using a deep neural network to fit a state-action value function. We employ empirical replay and target network techniques to improve the stability and convergence speed of the algorithm.

7. The intelligent goods delivery route optimization method according to claim 1, characterized in that, When creating virtual roles for logistics dispatchers, delivery personnel, and customers, and providing corresponding terminal devices to access the metaverse space, the following is included: By adopting low-latency access technology based on the convergence of edge computing and 5G, edge computing nodes are deployed in logistics parks and delivery stations. Terminal devices preprocess and analyze some data at the edge computing nodes and transmit the processed data to the server. The terminal device receives metaverse space data returned by the server and uses a multipath transmission protocol to divide the data into multiple data packets, which are then transmitted through different 5G network channels. The data is then reassembled at the receiving end, and interactive data with high real-time requirements is preferentially transmitted through low-latency 5G network slices.

8. The intelligent goods delivery route optimization method according to claim 1, characterized in that, When displaying the expected route of the delivery vehicle in a virtual scene, a visualization enhancement technology based on the fusion of augmented reality and virtual reality is used, including: For logistics dispatchers and customers, the virtual scene displayed by the VR device uses ray tracing algorithm to simulate real lighting effects, making delivery vehicles and road elements present a realistic three-dimensional visual effect. At the same time, the particle system is used to simulate the dynamic effects of exhaust fumes and dust during vehicle operation, enhancing the realism of the scene. For AR glasses used by delivery drivers, virtual delivery route guidance information is overlaid on the real scene. Spatial registration technology is used to ensure accurate matching between virtual routes and real roads. When delivery vehicles encounter special situations, the affected road sections are highlighted with bright and flashing effects in VR devices and AR glasses, and the adjusted route information is displayed simultaneously.

Images