Path planning method and system in digital twin environment and electronic equipment

Abstract

The invention provides a path planning method and system in a digital twin environment and electronic equipment, and belongs to the field of path planning. The method comprises the steps that a user instruction is acquired, and an execution task in the instruction is extracted; the method comprises the following steps: abstracting key geographic position information and a passage path in a real environment into nodes and edges in a graph structure based on a digital twinborn sum graph structure construction method, and generating a sparse directed graph; and in combination with the execution task and the sparse directed graph, performing global path exploration by using a Go-Explore algorithm to generate a plurality of candidate paths, and performing local optimization on the plurality of candidate paths by using an ERL algorithm to obtain a target path. Through combination of digital twin environment graph structure modeling and algorithm learning, efficient cooperation of global path exploration and local optimization is realized, the path planning efficiency, task execution flexibility and user interaction experience of a rover in a digital twin environment are remarkably improved, and a comprehensive and accurate solution is provided for navigation and decision making of the rover.

Description

Technical Field

[0001] The present invention belongs to the technical field of path planning, and specifically relates to a path planning method, system and electronic equipment in a digital twin environment. Background Art

[0002] With the rapid development of artificial intelligence and digital twin technology, the demand for the application of intelligent agents in complex environments is increasing. Especially in industrial scenarios such as parks and buildings, the efficient navigation and decision-making capabilities of intelligent agents have become the key to improving operational efficiency and enhancing user experience. However, traditional navigation and decision-making technologies are difficult to effectively deal with uncertainties in complex environments, limiting the application potential of intelligent agents. Therefore, there is an urgent need for a technical solution that can provide efficient, flexible navigation and intelligent decision-making in complex environments.

[0003] In the current application of digital twin technology, although the system can provide a virtualized environment presentation through high-precision modeling, its functions are often limited to path planning in a single scenario, and it is difficult to achieve multi-addressing under different user problems and tasks. At the same time, due to the lack of effective autonomous control capabilities, traditional digital twin systems usually require frequent user intervention, which reduces operational efficiency. In terms of humanized interaction design, existing digital twin systems generally lack convenient and intuitive operation methods, and it is difficult to provide a natural and friendly interactive experience. Most systems rely on traditional button or text input and fail to integrate multimodal interaction methods, such as voice commands, touch gestures and other more intuitive control methods. This single interaction mode increases the complexity of user operations. Especially for non-professional users, the cumbersome control process and command input reduce the ease of use and user experience of the system. Summary of the invention

[0004] The purpose of the embodiments of the present invention is to provide a path planning method, system and electronic device in a digital twin environment, which is used to completely or at least partially solve the technical problems existing in the above-mentioned prior art, such as the difficulty in realizing multi-dimensional addressing under different user problems and tasks and the low navigation efficiency in a digital twin environment.

[0005] In a first aspect, an embodiment of the present application provides a path planning method in a digital twin environment, including: Obtaining user instructions and extracting execution tasks contained in the instructions; Based on digital twin technology and graph structure construction methods, the key geographic location information and travel paths in the real environment are abstracted into nodes and edges between nodes in the graph structure to generate a sparse directed graph. In combination with the execution task and the sparse directed graph, the Go-Explore algorithm is used to perform global path exploration to generate multiple candidate paths, and the ERL algorithm is used to locally optimize the multiple candidate paths to obtain the target path, wherein the ERL algorithm includes the NSGA-II algorithm and the PPO algorithm.

[0006] Optionally, obtaining a user instruction and extracting an execution task contained in the instruction includes: If the user command is a voice command, the ffmpeg tool is used to unify the file encoding format into mp4 format, and the mp4 file is sent to OpenAI's Whisper model for voice-to-text conversion to obtain the converted text command; The converted text instructions are transferred to the LLM large model so that the LLM large model can perform semantic interpretation on the text instructions and extract the execution tasks contained in the text instructions in combination with the preset task prompt words.

[0007] Optionally, in combination with the execution task and the sparse directed graph, a Go-Explore algorithm is used to perform global path exploration to generate multiple candidate paths, including: Initialize the digital twin path planning space state, and archive and explore the path planning space state; Based on the set of explored paths, the cumulative reward of each path is calculated and multiple paths are output as candidate paths.

[0008] Optionally, the state archiving mechanism function and cumulative reward function are as follows: ; ; In the formula, S represents the path planning space state, s represents an exploration state in the path planning process, Represents cumulative rewards, Indicates the threshold for archiving. represents the exploration state at step t, represents the action taken at step t, Represents taking action at step t The instant reward is , and T represents the total number of steps in the path.

[0009] Optionally, the ERL algorithm is used to locally optimize multiple candidate paths to obtain a target path, including: Initialize the population based on the NSGA-II algorithm, define the objective optimization function to perform non-dominated sorting on the individuals in the population, and identify the optimal solution set under multi-objective conditions, where each individual in the population comes from the candidate path generated by Go-Explore; The PPO algorithm is used to limit the strategy update range through the strategy clipping mechanism, and the execution strategy of the path is gradually optimized in combination with real-time environmental feedback. The path strategy is partially adjusted to obtain the target path.

[0010] Optionally, the objective optimization function is as follows: ; In the formula, Represents the weight of the k-th target.

[0011] Optionally, the objective function non-dominated sorting formula is as follows: ; ; In the formula, A set of target optimization functions representing the path, represents the i-th path individual in the path population, represents the performance value of the kth objective, Representative path and non-dominated paths For all targets k, the path Performance values ​​are better than or equal to , and there is at least one target l, path The performance value is strictly better than .

[0012] Optionally, the policy clipping mechanism function and policy ratio function are as follows: ; ; In the formula, represents the policy update objective function, represents the strategy ratio, represents the new strategy probability, represents the old strategy probability, represents the advantage function, Represents the clipping range.

[0013] In a second aspect, the embodiment of the present application further provides a path planning system in a digital twin environment, including: The task parsing and instruction generation module is used to obtain user instructions and extract the execution tasks contained in the instructions; A graph construction module is used to abstract key geographic location information and travel paths in the real environment into nodes and edges between nodes in a graph structure based on digital twin technology and graph structure construction methods to generate a sparse directed graph; A path optimization and strategy adjustment module is used to combine the execution task and the sparse directed graph, use the Go-Explore algorithm to perform global path exploration to generate multiple candidate paths, and use the ERL algorithm to locally optimize the multiple candidate paths to obtain the target path, wherein the ERL algorithm includes the NSGA-II algorithm and the PPO algorithm; The human-computer interaction and feedback module is used to visualize the target path, view the navigation path, current position and target position through a 3D interface, and obtain real-time operation guidance through real-time voice.

[0014] In a third aspect, an embodiment of the present application also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the path planning method in the digital twin environment described above are implemented.

[0015] It can be seen from the above technical solutions that the present invention has the following advantages: The present application provides a path planning method, system and electronic device in a digital twin environment. By combining digital twin environment graph structure modeling and evolutionary NSGA-II + PPO reinforcement learning, efficient coordination of global path exploration and local optimization is achieved, which significantly improves the path planning efficiency, task execution flexibility and user interaction experience of the rover in the digital twin environment, and provides a comprehensive and accurate solution for rover navigation and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the description will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0017] Figure 1 A flow chart of a path planning method in a digital twin environment provided by an embodiment of the present invention; Figure 2 An ERL architecture diagram provided for an embodiment of the present invention; Figure 3 A detailed flow chart of a path planning method in a digital twin environment provided by an embodiment of the present invention; Figure 4 A schematic diagram of the structure of a path planning system in a digital twin environment provided by an embodiment of the present invention; Figure 5 A schematic diagram of the hardware structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In the detailed description below, various embodiments of the present disclosure will be described more fully. The present disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but rather the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0019] Hereinafter, the terms "include" or "may include" that may be used in various embodiments of the present disclosure indicate the presence of the disclosed functions or operations, and do not limit the addition of one or more functions or operations. In addition, as used in various embodiments of the present disclosure, the terms "include", "have" and their cognates are intended only to indicate specific features, numbers, steps, operations, or combinations of the foregoing items, and should not be understood as first excluding the presence of one or more other features, numbers, steps, operations, or combinations of the foregoing items or the possibility of adding one or more features, numbers, steps, operations, or combinations of the foregoing items.

[0020] In various embodiments of the present disclosure, the expression "or" or "at least one of A or / and B" includes any combination or all combinations of the words listed at the same time. For example, the expression "A or B" or "at least one of A or / and B" may include A, may include B, or may include both A and B.

[0021] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0022] See also Figure 1 FIG. 1 is a flow chart of a path planning method in a digital twin environment in a specific embodiment, including the following execution steps: Step 100: Obtain a user instruction and extract the execution task contained in the instruction.

[0023] Specifically, if the user instruction is a voice instruction, the ffmpeg tool is used to unify the file encoding format into mp4 format, and the mp4 file is sent to OpenAI's Whisper model for voice-to-text conversion to obtain the converted text instruction; the converted text instruction is handed over to the LLM large model so that the LLM large model can perform semantic interpretation on the text instruction, and extract the execution task contained in the text instruction in combination with the preset task prompt words.

[0024] Furthermore, the large model uniformly returns a fixed JSON format, in which an id is defined to distinguish different tasks; a type field is defined to distinguish different types of tasks; an args field is defined as an input parameter; and returns is defined as the return content. The above are key parameters.

[0025] Exemplarily, user task types include five categories: tasks with a fixed starting point and a fixed end point, tasks with a fixed starting point and a fixed end point and including stop points, tasks with a fixed starting point and an end point in a range area, tasks with a fixed starting point and an end point of a specified type of location (multiple locations distributed in different areas), and tasks with a fixed starting point and no end point and containing instruction actions.

[0026] Step 101: Based on the digital twin technology and the graph structure construction method, the key geographic location information and the travel path in the real environment are abstracted into nodes and edges between nodes in the graph structure to generate a sparse directed graph.

[0027] Specifically, the system abstracts key locations (including major nodes, intersections, target points, etc.) as nodes in the graph structure, and abstracts traversable paths as edges between nodes. On this basis, each edge is given an initial weight according to the different characteristics of the path (such as length, width, slope, etc.) to accurately reflect its difficulty or priority. Finally, the system constructs a sparse directed graph through the reasonable abstraction of key locations and traversable paths, ensuring the efficient navigation and task execution of the rover in a complex environment. Through this sparse directed graph, the rover can achieve flexible and efficient path selection and optimization.

[0028] Step 102: In combination with the execution task and the sparse directed graph, a Go-Explore algorithm is used to perform global path exploration to generate multiple candidate paths, and an ERL algorithm is used to locally optimize the multiple candidate paths to obtain a target path, wherein the ERL algorithm includes an NSGA-II algorithm and a PPO algorithm.

[0029] In this embodiment, by combining digital twin environment graph structure modeling and evolutionary NSGA-II+PPO reinforcement learning, efficient coordination of global path exploration and local optimization is achieved, which significantly improves the path planning efficiency of the rover in the digital twin environment.

[0030] In a specific implementation, when executing step 102, the following steps can be specifically performed: initializing the digital twin path planning space state, and archiving and exploring the path planning space state; based on the explored path set, calculating the cumulative reward of each path and outputting multiple paths as candidate paths.

[0031] It should be noted that initializing the digital twin state space includes completing the initialization of the digital twin environment, establishing an empty state archive, and defining a reward function to evaluate the pros and cons of the state during the exploration process.

[0032] In some implementations, the state archiving mechanism function and the cumulative reward function are as follows: ; ; In the formula, S represents the path planning space state (including the current position and related environmental information), s represents an exploration state in the path planning process, Represents cumulative rewards, Indicates the threshold for archiving (used to filter which states can be archived). represents the exploration state at step t, represents the action taken at step t, Represents taking action at step t The instant reward is , and T represents the total number of steps in the path.

[0033] Archive and explore the state. During the exploration process, the system tries multiple possible actions a∈A() from the current state to generate a new state. Evaluate the new state according to the value of the reward function, and store the state that meets the reward condition in the state archive S.

[0034] In one embodiment of the present invention, the ERL (NSGA-II+PPO) architecture diagram is shown in Figure 2 As shown, based on the use of the ERL algorithm in step 102 to locally optimize multiple candidate paths to obtain a target path, a possible embodiment will be given below to non-limitingly illustrate its specific implementation scheme.

[0035] S1: Initialize the population based on the NSGA-II algorithm, define the objective optimization function to perform non-dominated sorting on the individuals in the population, and identify the optimal solution set under multi-objective conditions, where each individual in the population comes from the candidate path generated by Go-Explore.

[0036] Specifically, multi-objective optimization is performed on the population. First, according to the defined objective optimization function (path length, width, slope), the individuals in the population are non-dominated and sorted to identify the optimal solution set under multi-objective conditions. Then, the next generation of population is generated through selection, crossover and mutation operations to improve the adaptability and diversity of the path. Multi-objective optimization ensures that the generated path is balanced in multiple performance indicators and provides a reliable basis for path optimization in the digital twin environment.

[0037] Among them, the objective optimization function is as follows: ; In the formula, Represents the weight of the k-th target (including path length, width, and slope).

[0038] The objective function non-dominated sorting formula is as follows: ; ; In the formula, A set of target optimization functions representing the path, represents the i-th path individual in the path population, represents the performance value of the kth objective, Representative path and non-dominated paths For all targets k, the path Performance values ​​are better than or equal to , and there is at least one target l, path The performance value is strictly better than .

[0039] S2: Use the PPO algorithm to limit the strategy update range through the strategy clipping mechanism, and gradually optimize the path execution strategy in combination with real-time environmental feedback, make local adjustments to the path strategy, and obtain the target path.

[0040] Among them, the policy clipping mechanism function and policy ratio function are as follows: ; ; In the formula, represents the policy update objective function, represents the strategy ratio, represents the new strategy probability, represents the old strategy probability, represents the advantage function, Represents the clipping range.

[0041] In one embodiment, Figure 3 This is a detailed flow chart of a path planning method in a digital twin environment provided according to an embodiment of the present invention. This embodiment is further optimized and expanded on the basis of the above embodiments.

[0042] First, the natural language instructions input by the user are received and converted into tasks that the system can execute. Users can express their needs through voice or text, such as "plan the shortest path from A to B" or "find nearby restaurants". The Whisper model is used to convert voice instructions into text, and then the text content is semantically parsed through the large language model (LLM) to extract the user's target needs and constraints.

[0043] After parsing is completed, combined with predefined task prompts, tasks are classified into fixed path planning, multi-target path search, or range area tasks. Finally, the parsing results are output in the form of standardized JSON instructions, including task ID, task type, input parameters, and expected return values. These instructions are then passed on for further processing and execution.

[0044] At the same time, key locations (such as nodes, intersections, and destinations) are abstracted as nodes in the graph, and traversable paths are abstracted as edges in the graph. Based on the characteristics of the path (length, width, slope), initial weights are assigned to the edges to complete the construction of a sparse directed graph. This graph not only fully represents the topological structure of the environment, but also provides basic support for efficient path planning.

[0045] Next, combined with the output task instructions and the generated sparse directed graph, the Go-Explore and ERL (NSGA-II + PPO) algorithms are used (if it is the first task execution, the Go-Explore algorithm is used, otherwise the ERL (NSGA-II + PPO) algorithm is used) to achieve global exploration and local optimization of path planning. First, the Go-Explore algorithm is responsible for global path exploration, generates a set of candidate paths through breadth-first search, and scores each path based on the reward function (path length, width, slope). High-quality paths are stored in the state archive to provide a reference for subsequent optimization.

[0046] On this basis, the NSGA-II algorithm performs multi-objective optimization on candidate paths, including path length and traffic efficiency, to generate a Pareto optimal solution set. Subsequently, the PPO algorithm dynamically adjusts the optimization results in combination with path weight feedback, for example, replanning the optimal solution when the path changes due to obstacles or congestion. Through the collaborative work of global exploration and local optimization, the system can achieve optimal path planning in a digital twin environment.

[0047] Finally, multimodal interaction technology is used to provide users with real-time feedback and task progress display to show the current location and target location. In addition, voice feedback is generated through WaveNet technology to provide users with clear guidance. For example, when the rover reaches the destination, it will prompt: "You have arrived at the destination!" This multimodal feedback method not only significantly improves the user experience, but also enhances the ease of use and interaction efficiency of the system.

[0048] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0049] like Figure 4 As shown, the following is an embodiment of the path planning system in a digital twin environment provided by an embodiment of the present disclosure, which belongs to the same inventive concept as the path planning method in a digital twin environment in the above-mentioned embodiments. For details not fully described in the embodiment of the path planning system in a digital twin environment, please refer to the embodiment of the path planning method in the above-mentioned digital twin environment.

[0050] The task parsing and instruction generating module 40 is used to obtain the user instruction and extract the execution task contained in the instruction; A graph construction module 41 is used to abstract key geographic location information and travel paths in a real environment into nodes and edges between nodes in a graph structure based on digital twin technology and a graph structure construction method, and generate a sparse directed graph; A path optimization and strategy adjustment module 42 is used to combine the execution task and the sparse directed graph, use the Go-Explore algorithm to perform global path exploration to generate multiple candidate paths, and use the ERL algorithm to locally optimize the multiple candidate paths to obtain a target path, wherein the ERL algorithm includes the NSGA-II algorithm and the PPO algorithm; The human-computer interaction and feedback module 43 is used to visualize the target path, view the navigation path, current position and target position through a 3D interface, and obtain real-time operation guidance through real-time voice.

[0051] The system is particularly suitable for path planning needs in digital twin environments, such as campus management, smart transportation, and digital twin systems. By combining advanced algorithms and interactive design, the system can help users efficiently complete multi-objective task planning, improve resource utilization efficiency, and ensure the stability and continuity of business processes.

[0052] Figure 5 It is a schematic diagram of the hardware structure of an electronic device for implementing various embodiments of the present invention.

[0053] The path planning method in the digital twin environment provided in the embodiment of the present application can be applied to electronic devices. It can be understood by those skilled in the art that the electronic device structure involved in the embodiment of the present invention does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown, or combine certain components, or arrange different components. In an embodiment of the present invention, the electronic device includes but is not limited to a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described and / or required herein.

[0054] The electronic device may include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charging management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, buttons, a camera, a display, and a SIM card interface, etc.

[0055] It is to be understood that the structure illustrated in the embodiments of the present application does not constitute a specific limitation on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0056] The processor may include one or more processing units, for example, the processor may include a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units may be independent devices or integrated into one or more processors.

[0057] The processor can be the nerve center and command center of the electronic device. The controller can generate an operation control signal according to the instruction operation code and timing signal to complete the control of fetching and executing instructions.

[0058] A memory may also be provided in the processor for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. The memory may store instructions or data that the processor has just used or is cyclically used. If the processor needs to use the instruction or data again, it may be directly called from the memory. This avoids repeated access, reduces the waiting time of the processor, and thus improves system efficiency.

[0059] The external memory interface can be used to connect an external memory card, such as a MicroSD card, to expand the storage capacity of the electronic device. The external memory card communicates with the processor through the external memory interface to implement data storage functions. For example, files such as music and videos can be saved in the external memory card.

[0060] The internal memory can be used to store computer executable program codes, which include instructions. The processor executes various functional applications and data processing of the electronic device by running the instructions stored in the internal memory. The internal memory may include a program storage area and a data storage area. The internal memory may include a high-speed random access memory and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, a universal flash storage (UFS), etc.

[0061] The wireless communication function of an electronic device can be realized through an antenna, a wireless communication module, a modem processor, and a baseband processor.

[0062] The wireless communication module can provide wireless communication solutions for electronic devices, including wireless local area networks (WLAN) (such as wireless fidelity (Wi-Fi) networks), Bluetooth (BT), global navigation satellite system (GNSS), frequency modulation (FM), near field communication (NFC), infrared technology (IR), etc.

[0063] Electronic devices can implement audio functions, etc. through audio modules, speakers, receivers, microphones, headphone jacks, and application processors.

[0064] Electronic devices can achieve shooting functions through ISP, camera, video codec, GPU, display and application processor.

[0065] Electronic devices can achieve display functions through GPU, display screen and application processor.

[0066] The GPU is a microprocessor for image processing that connects the display screen and the application processor. The GPU is used to perform mathematical and geometric calculations for graphics rendering. The processor may include one or more GPUs that execute program instructions to generate or change display information.

[0067] The display screen is used to display images, videos, etc. The display screen includes a display panel.

[0068] The electronic device implements the acquisition of user instructions of the subject name of the path planning method in the digital twin environment of the present application, and extracts the execution tasks contained in the instructions; Based on digital twin technology and graph structure construction methods, the key geographic location information and travel paths in the real environment are abstracted into nodes and edges between nodes in the graph structure to generate a sparse directed graph. In combination with the execution task and the sparse directed graph, the Go-Explore algorithm is used to perform global path exploration to generate multiple candidate paths, and the ERL algorithm is used to locally optimize the multiple candidate paths to obtain the target path. The ERL algorithm includes the NSGA-II algorithm and the PPO algorithm technical solutions, which achieves the beneficial effect of improving the path planning efficiency of the rover in the digital twin environment.

[0069] The storage medium provided in the present application stores a program product that can implement a path planning method in a digital twin environment.

[0070] The path planning method in a digital twin environment includes: obtaining user instructions and extracting execution tasks contained in the instructions; based on digital twin technology and graph structure construction methods, key geographical location information and travel paths in the real environment are abstracted into nodes and edges between nodes in the graph structure to generate a sparse directed graph; combining the execution tasks and the sparse directed graph, using the Go-Explore algorithm to perform global path exploration to generate multiple candidate paths, and using the ERL algorithm to locally optimize the multiple candidate paths to obtain the target path, wherein the ERL algorithm includes the NSGA-II algorithm and the PPO algorithm.

[0071] In some possible implementations, the subject matter of the present disclosure, namely, the path planning method, system, and electronic device in a digital twin environment, can be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps of various exemplary implementations of the present disclosure described in the above “Exemplary Method” section of this specification.

[0072] The storage medium of the present disclosure can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0073] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A path planning method in a digital twin environment, characterized in that: include: Obtaining user instructions and extracting execution tasks contained in the instructions; Based on digital twin technology and graph structure construction methods, the key geographic location information and travel paths in the real environment are abstracted into nodes and edges between nodes in the graph structure to generate a sparse directed graph. In combination with the execution task and the sparse directed graph, the Go-Explore algorithm is used to perform global path exploration to generate multiple candidate paths, and the ERL algorithm is used to locally optimize the multiple candidate paths to obtain the target path, wherein the ERL algorithm includes the NSGA-II algorithm and the PPO algorithm.

2. The path planning method in a digital twin environment according to claim 1, characterized in that: Obtain user instructions and extract execution tasks contained in the instructions, including: If the user command is a voice command, the ffmpeg tool is used to unify the file encoding format into mp4 format, and the mp4 file is sent to OpenAI's Whisper model for voice-to-text conversion to obtain the converted text command; The converted text instructions are transferred to the LLM large model so that the LLM large model can perform semantic interpretation on the text instructions and extract the execution tasks contained in the text instructions in combination with the preset task prompt words.

3. The path planning method in a digital twin environment according to claim 1, characterized in that: Combining the execution task and the sparse directed graph, the Go-Explore algorithm is used to perform global path exploration to generate multiple candidate paths, including: Initialize the digital twin path planning space state, and archive and explore the path planning space state; Based on the set of explored paths, the cumulative reward of each path is calculated and multiple paths are output as candidate paths.

4. The path planning method in a digital twin environment according to claim 3 is characterized in that: The state archiving mechanism function and cumulative reward function are as follows: ； ； In the formula, S represents the path planning space state, s represents an exploration state in the path planning process, Represents cumulative rewards, Indicates the threshold for archiving. represents the exploration state at step t, represents the action taken at step t, Represents taking action at step t The instant reward is , and T represents the total number of steps in the path.

5. The path planning method in a digital twin environment according to claim 1, characterized in that: The ERL algorithm is used to locally optimize multiple candidate paths to obtain the target path, including: Initialize the population based on the NSGA-II algorithm, define the objective optimization function to perform non-dominated sorting on the individuals in the population, and identify the optimal solution set under multi-objective conditions, where each individual in the population comes from the candidate path generated by Go-Explore; The PPO algorithm is used to limit the strategy update range through the strategy clipping mechanism, and the execution strategy of the path is gradually optimized in combination with real-time environmental feedback. The path strategy is partially adjusted to obtain the target path.

6. The path planning method in a digital twin environment according to claim 5, characterized in that: The objective optimization function is as follows: ； In the formula, Represents the weight of the k-th target.

7. The path planning method in a digital twin environment according to claim 6, characterized in that: The objective function non-dominated sorting formula is as follows: ； ； In the formula, A set of target optimization functions representing the path, represents the i-th path individual in the path population, represents the performance value of the kth objective, Representative path and non-dominated paths For all targets k, the path Performance value is better than or equal to , and there is at least one target l, path The performance value is strictly better than .

8. The path planning method in a digital twin environment according to claim 5, characterized in that: The policy clipping mechanism function and policy ratio function are as follows: ； ； In the formula, represents the policy update objective function, represents the strategy ratio, represents the probability of the new strategy, represents the old strategy probability, represents the advantage function, Represents the clipping range.

9. A path planning system in a digital twin environment, characterized in that: include: The task parsing and instruction generation module is used to obtain user instructions and extract the execution tasks contained in the instructions; A graph construction module is used to abstract key geographic location information and travel paths in the real environment into nodes and edges between nodes in a graph structure based on digital twin technology and graph structure construction methods to generate a sparse directed graph; A path optimization and strategy adjustment module is used to combine the execution task and the sparse directed graph, use the Go-Explore algorithm to perform global path exploration to generate multiple candidate paths, and use the ERL algorithm to locally optimize the multiple candidate paths to obtain the target path, wherein the ERL algorithm includes the NSGA-II algorithm and the PPO algorithm; The human-computer interaction and feedback module is used to visualize the target path, view the navigation path, current position and target position through a 3D interface, and obtain real-time operation guidance through real-time voice.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the path planning method in the digital twin environment as described in any one of claims 1 to 8 are implemented.

Images