Park operation system and method

Abstract

The invention provides a park operation system and method, and relates to the technical field of park operation, and the system comprises a control instruction output module which is used for collecting a user instruction, inputting the user instruction into a language model, and converting the user instruction into a control instruction; and the data display module is connected with the control instruction output module and is used for linking the physical entities in the park and the virtual entities in the generated virtual scene according to the control instruction and outputting display data for display. Compared with the prior art, the method has the beneficial effect that logic linkage between the physical entity and the virtual entity can be realized. By acquiring the user instruction, the physical entity and the virtual entity can be linked according to the demand of the user, so that the interaction between the user and the operation system is more friendly and convenient. Furthermore, the linkage result can be sent to be displayed, and the corresponding page can be automatically skipped to and the corresponding display operation can be executed according to the logic linkage between the physical entity and the virtual entity, so that the linkage result can be displayed to the user more visually.

Description

Technical Field

[0001] The present invention relates to the field of park operation technology, and in particular to a park operation system and method. Background Art

[0002] Digital twins use digital technology to make full use of physical models, sensor updates, operation history and other data, integrate multi-disciplinary, multi-physical quantity, multi-scale, and multi-probability simulation processes, and map real-world scenes to virtual scenes through three-dimensional digital modeling technology. At the same time, operations in virtual reality scenes also synchronously affect the operation of equipment in the real world.

[0003] The construction of smart parks is an important part of smart cities and is also an advanced stage of smart city development. Smart parks use the current advanced cloud computing, big data decision analysis, Internet of Things and other cutting-edge technologies to unify and integrate the current Internet architecture, equipment sensing technology, batch information processing and other information means, and apply perception, interconnection and intelligence to all levels of the park through intelligent response methods such as monitoring, analysis and integration, and connect the original scattered and independent basic physical facilities, basic information equipment, basic social equipment and basic commercial facilities in the park to jointly build a more advanced collection of intelligent information facilities, thereby upgrading the park to an intelligent whole with information interconnection, high control ability and high disaster resistance.

[0004] However, the current mainstream digital twin applications in industrial parks are mainly the interaction between two-dimensional planes and data panels, and are basically used to display local data and statistics. They lack interaction with users or the interaction methods with users are complex. Summary of the invention

[0005] In view of the problems existing in the prior art, the present invention provides a park operation system, comprising:

[0006] A control instruction output module is used to collect user instruction input language models and convert them into control instructions;

[0007] The data display module is connected to the control instruction output module and is used to link the physical entities in the park and the virtual entities in the generated virtual scene according to the control instruction and output display data for display.

[0008] Preferably, the control instruction output module includes:

[0009] A keyword extraction unit, used to extract keywords from the user instruction;

[0010] The instruction generation unit is connected to the keyword extraction unit and is used to input the keyword into the language model to convert it into the control instruction, and the control instruction includes a location instruction and a logic event instruction.

[0011] Preferably, the control instruction includes a voice instruction and a text instruction, and the data display module includes:

[0012] A voice broadcasting unit, used for broadcasting the voice form instruction to the user;

[0013] A text sending unit, used for sending the text form instruction to the data display module;

[0014] The data display module links the physical entities in the park and the virtual entities in the generated virtual scene according to the text instructions and outputs display data for display.

[0015] Preferably, the user instruction includes a voice instruction, a gesture instruction and a text instruction, and the keyword extraction unit includes:

[0016] A text extraction subunit, used for extracting keywords from the text instruction;

[0017] A speech recognition subunit, used for inputting the speech instruction into a speech recognition model to extract keywords in the speech instruction;

[0018] The image recognition subunit is used to input the gesture instruction into an image recognition model to extract keywords in the gesture instruction.

[0019] Preferably, the speech recognition model in the speech recognition subunit is an RNN-T prediction network model, the RNN-T prediction network model includes a prediction network, a decoding network, a joint model and a Softmax layer, and the decoding network includes 8 prediction networks;

[0020] The voice instruction serves as the input of the prediction network, the input of the joint model is connected to the output of the decoding network and the output of the prediction network, the output of the joint model is connected to the input of the Softmax layer, and the output of the Softmax layer serves as the output of the speech recognition model.

[0021] Preferably, the gesture instruction includes an upward / downward / leftward / rightward sliding action, and the image recognition model extracts the sliding action in the gesture instruction as the keyword by identifying the gesture instruction;

[0022] The instruction generation unit outputs a corresponding control instruction according to the keyword, and the data display module maps the physical entities in the park into virtual entities according to the control instruction, and performs sliding in a corresponding proportional direction.

[0023] Preferably, the data display module includes:

[0024] An acquisition unit, configured to acquire map data based on the location instruction and to acquire business data based on the logic event instruction;

[0025] A twin unit is connected to the acquisition unit and is used to link the physical entity with the virtual entity based on the map data and the business data and generate the display data for display.

[0026] Preferably, the twin unit comprises:

[0027] The twin layer is used to link physical entities with virtual entities based on map data and business data;

[0028] The panel layer is connected to the twin layer and is used to generate the display data for display based on the result of the linkage.

[0029] The present invention also provides a park operation method, which is applied to the above-mentioned park operation system, comprising:

[0030] Step S1, the park operation system collects user instructions and inputs the language model into control instructions;

[0031] Step S2: the park operation system links the physical entities in the park with the virtual entities in the generated virtual scene according to the control instruction and outputs display data for display.

[0032] Preferably, the step S1 comprises:

[0033] Step S11, the park operation system extracts keywords from the user instruction;

[0034] Step S12: the park operation system inputs the keyword into the language model to convert it into the control instruction, which includes a location instruction and a logic event instruction.

[0035] The above technical solution has the following advantages or beneficial effects:

[0036] Compared with the existing technology in which the park operation system can only display the park, the present invention can realize the logical linkage between the physical entity and the virtual entity. By obtaining the user's instructions, the physical entity and the virtual entity can be linked according to the user's needs, making the interaction between the user and the operation system more friendly and convenient. Furthermore, the linkage result can be sent for display, and according to the logical linkage between the physical entity and the virtual entity, it can automatically jump to the corresponding page and perform the corresponding display operation, so as to be displayed to the user more intuitively. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 This is a structural schematic diagram of a park operation system in a preferred embodiment of the present invention;

[0038] Figure 2 A schematic diagram of the training of a speech recognition model and the reasoning process of the model in a preferred embodiment of the present invention;

[0039] Figure 3 A schematic diagram of the structure of a speech recognition model in a preferred embodiment of the present invention;

[0040] Figure 4 A schematic diagram of a process flow of a park operation method in a preferred embodiment of the present invention;

[0041] Figure 5 This is a schematic diagram of a sub-process of step S1 in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0042] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. The present invention is not limited to this embodiment, and other embodiments may also fall within the scope of the present invention as long as they conform to the gist of the present invention.

[0043] In a preferred embodiment of the present invention, based on the above problems existing in the prior art, a park operation system is provided. Figure 1 As shown, including:

[0044] Control instruction output module 1, used to collect user instruction input language model and convert it into control instructions;

[0045] The data display module 2 is connected to the control instruction output module 1 and is used to link the physical entities in the park and the virtual entities in the generated virtual scene according to the control instructions and output display data for display.

[0046] Specifically, in this embodiment, the control instruction refers to the answer text corresponding to the user instruction. The data display module 2 maps the physical entity to the virtual entity for linkage according to the answer text. The answer text not only includes the coordinate data of the physical entity mapped to the virtual entity for linkage, but also includes the logical relationship when the physical entity and the virtual entity are linked according to the linked coordinate data, so as to realize the logical linkage between the physical entity and the virtual entity according to the user instruction. Among them, the location instruction is used to indicate the coordinate data of the physical entity mapped to the virtual entity for linkage, and the logical event instruction is used to indicate the logical relationship when the physical entity is mapped to the virtual entity for linkage. The logical relationship during linkage includes but is not limited to route planning, model scene switching, transition, lens zooming in, etc. For example, if the user instruction is "go to the parking lot", the control instruction output by the control instruction output module 1 is "route planning to the parking lot", and the output control instruction includes not only the location coordinates of the parking lot, but also the specific route from the current location to the parking lot required by the user. For another example, if the user instruction is "what is the name of the specified building", the control instruction output by the control instruction output module 1 is "specific building name", and the output control instruction includes not only the location coordinates of the specified building, but also the name of the building required by the user, etc.

[0047] It should be noted that the control instruction output by the control instruction output module 1 is not limited to one control instruction, but can also be a series of instructions. For example, if the user instruction is "go to the parking lot", the corresponding automatically output control instruction can be a path planning from the current location to the docking parking spot, the camera switches to the parking lot, and can also include a series of instructions such as switching the panel tab of the display module to the transportation service.

[0048] In a preferred embodiment of the present invention, the control instruction output module 1 includes:

[0049] A keyword extraction unit 11, used to extract keywords from user instructions;

[0050] The instruction generating unit 12 is connected to the keyword extracting unit 11 and is used to convert the keywords into the language model to obtain control instructions. The control instructions include location instructions and logic event instructions.

[0051] Further, in this embodiment, the control instruction includes a voice form instruction and a text form instruction, and the data display module 2 includes:

[0052] A voice broadcasting unit 21, used to broadcast voice instructions to the user;

[0053] The text sending unit 22 is used to link the physical entities in the park and the virtual entities in the generated virtual scene according to the text form instructions and output display data for display.

[0054] Furthermore, in this embodiment, the data display module 2 includes:

[0055] An acquisition unit 23, connected to the text sending unit 22, is used to acquire map data based on location instructions and to acquire business data based on logic event instructions;

[0056] The twin unit 24 is connected to the acquisition unit and is used to link the physical entity with the virtual entity based on the map data and the business data and generate display data for display.

[0057] Furthermore, in this embodiment, the twin unit 24 includes:

[0058] The twin layer 241 is used to link physical entities with virtual entities based on map data and business data;

[0059] The panel layer 242 is connected to the twin layer 241 and is used to generate display data for display based on the results of the linkage.

[0060] Specifically, in this embodiment, the logic event instructions include information query, path planning, highlight area, camera switching, floor splitting and aggregation. In some embodiments, the information query includes querying the name of the landmark building in the park, the name of the functional area in the park, the heat map of the designated location in the park, and the status of the parking lot.

[0061] Specifically, the data display module 2 obtains corresponding data according to the location instructions and logic event instructions output by the control instruction output module, maps the corresponding position coordinates in the physical entity to the coordinates in the virtual entity, and links the data in the physical entity corresponding to the logic event with the data in the virtual entity, so as to link the real physical entity in the virtual three-dimensional scene. At the same time, the display data is generated so that the display module switches to the corresponding display page according to the display data and displays it vividly and intuitively to the user.

[0062] This embodiment accurately identifies the control instruction corresponding to the user instruction output through the language model in the control instruction output module 1. The control instruction is the answer text corresponding to the user instruction, which not only includes the location coordinates for mapping the physical entity to the virtual entity for linkage, but also includes the logical relationship of linkage when the physical entity and the virtual entity are linked according to the location coordinates, so that the data display module can realize the logical linkage between the physical entity in the real world and the virtual entity in the digital twin system according to the output control instruction. Compared with the park operation system in the prior art that can only display the park, this embodiment can realize the logical linkage between the physical entity and the virtual entity. Further, the control instruction output module in this embodiment can link the physical entity with the virtual entity according to the needs of the user by obtaining the user instruction, so that the interaction between the user and the operation system is more friendly and convenient. Further, the linkage result can also be sent for display, and according to the logical linkage between the physical entity and the virtual entity, it can automatically jump to the corresponding page and perform the corresponding display operation so as to be more intuitively displayed to the user.

[0063] In a preferred embodiment of the present invention, the user instruction includes a voice instruction, a gesture instruction and a text instruction, and the keyword extraction unit 11 includes:

[0064] The text extraction subunit 111 is used to extract keywords from the text instruction;

[0065] The speech recognition subunit 112 is used to input the speech instruction into the speech recognition model to extract the keywords in the speech instruction;

[0066] The image recognition subunit 113 is used to input the gesture instruction into the image recognition model to extract keywords in the gesture instruction.

[0067] Specifically, in this embodiment, the user can send the user command to the control command output module through a third-party device. For example, the user can collect the user's voice command through a microphone and send it to the control command output module through transmission, or the user can collect the user's gesture command through a camera and send it to the control command output module through transmission.

[0068] In a preferred embodiment of the present invention, the speech recognition model in the speech recognition subunit is an RNN-T prediction network model, the RNN-T prediction network model includes a prediction network, a decoding network, a joint model and a Softmax layer, and the decoding network includes 8 prediction networks;

[0069] The voice command is used as the input of the prediction network. The input of the joint model is connected to the output of the decoding network and the output of the prediction network. The output of the joint model is connected to the input of the Softmax layer. The output of the Softmax layer is used as the output of the speech recognition model.

[0070] Specifically, in this embodiment, a schematic diagram of a speech recognition principle provided by the embodiment of the present disclosure is as follows: Figure 2 As shown in the figure, speech recognition includes two parts: model training and model reasoning. Specifically, on the one hand, after the speech input is preprocessed, feature vectors are extracted, and the extracted feature vectors are trained to obtain the model layer, and the speech input and feature vectors are stored in the template library for template matching during reasoning. On the other hand, during reasoning, the speech input is preprocessed to extract feature vectors, and the extracted feature vectors are reasoned with the trained model. The feature vector sequence of the input speech is compared with each template in the template library in turn for similarity, and the one with the highest similarity is output as the recognition result.

[0071] In some embodiments, the speech recognition model can be an RNN-T (Listen, Attend, and, Spell) model. The RNN-T prediction network model is a language model prediction network branch introduced for the conditional independence hypothesis of the CTC network (i.e., the current output is conditionally independent of the previous output). RNN-T combines the prediction network branch with the feedforward network through a joint feedforward neural network. When predicting the final output, it can simultaneously use acoustic and language feature information to more accurately extract keywords in voice commands.

[0072] This embodiment uses the RNN-T prediction network model to extract campus keywords, which can continuously process input voice commands and perform symbolic output. The feedback path of the RNN-T recognizer predicts the next symbol by inputting the symbols predicted by the RNN-T prediction network model back to itself, making the prediction result more accurate.

[0073] Figure 3 A schematic diagram of an RNN-T prediction network model provided in this embodiment. Figure 3 As shown in the figure, the RNN-T prediction network includes a prediction network Pred.Network, a decoding network Encoder, a joint model JointNetwork and a Softmax layer, wherein the prediction network Pred.Network consists of 2 layers with 2048 units and 1 projection layer with 640 dimensions, and the decoding network Encoder consists of 8 prediction network Pred.Network layers, i.e. 16 layers with 2048 units and 8 projection layers with 640 dimensions. x represents the input voice command data; y represents the predicted symbol. The predicted symbol (i.e. the output of the Softmax layer) y. is input back to the model through the RNN-T prediction network to ensure that the prediction takes into account both the current voice command and the past output. The prediction network Pred.Network and the decoding network Encoder are both LSTM-RNN, and the joint model JointNetwork is a feedforward network.

[0074] In a preferred embodiment of the present invention, the gesture instruction includes an upward / downward / leftward / rightward sliding action, and the image recognition model extracts the sliding action in the gesture instruction as a keyword by recognizing the gesture instruction;

[0075] The instruction generation unit 12 outputs corresponding control instructions according to the keywords, and the data display module 2 maps the physical entities in the park into virtual entities according to the control instructions, and performs sliding in the corresponding proportional direction.

[0076] Specifically, in this embodiment, the image recognition model can be a visual perception large model (Recognize Anything Model, RAM), or other image recognition models. Gesture instructions include sliding up / down / left / right. The image recognition model extracts keywords in the gesture instructions by identifying the gesture instructions. The instruction generation unit outputs corresponding control instructions according to the keywords. The driving data display module maps the physical entity to a virtual entity according to the control instruction, and slides in the corresponding proportional direction. Gesture instructions also include operations such as rotation.

[0077] Furthermore, in some embodiments, both the speech recognition model and the image recognition model can be generated based on neural network training.

[0078] In this embodiment, the keyword extraction unit extracts keywords from the voice command or gesture command input by the user, and the command generation unit generates corresponding control commands according to the keywords of the extracted user commands to drive the data display module to perform logical linkage and generate display data for the display module to display. This allows the user of the terminal to interact with the park system through voice input or gesture interaction, making the interaction between the user and the park operation system more convenient and simple to operate.

[0079] In some embodiments, the data display module 2 includes an acquisition unit 23 and a twin unit 24, wherein the acquisition unit 23 is used to acquire map data based on location instructions and to acquire business data based on logical event instructions; the twin unit 24 is used to link physical entities with virtual entities based on map data and business data and generate display data for display.

[0080] Specifically, after the data display module 2 receives the location instruction and the logical event instruction, the acquisition unit 23 first acquires the map data required for linkage according to the location instruction. The map data can be the real map data in the physical entity, or the map data of the virtual entity mapped according to the physical entity data. Then, the business data required for linkage is acquired according to the logical event instruction. The business data corresponds to the logical event instruction. The business data can be pre-stored data or real-time park data, etc. It can be understood that the business data can also be other data that links the physical entity with the virtual entity, and the present disclosure does not limit this. Finally, the twin unit 24 maps the corresponding position coordinates in the physical entity to the coordinates in the virtual entity according to the map data, and performs linkage according to the business data, and generates corresponding display data to be displayed to the user in the display module.

[0081] For example, when the logic event instruction is to query the status of a parking lot, the corresponding business data is the real-time data of the parking lot in the park; when the logic event instruction is to query the heat map of a specified location in the park, the corresponding business data is the real-time flow of people at the specified location in the park. When the logic event instruction is to query the name of a landmark building in the park or the name of a functional area in the park, the corresponding business data can be pre-stored building name data or pre-stored functional area name data. When the logic event instruction is path planning, the corresponding business data can be a pre-stored path planning route map, etc.

[0082] In some embodiments, the park operation system includes not only a control instruction output module 1 and a data display module 2, but also a storage module 3, and the storage module 3 is used to store the correspondence between the logic event instructions and the business data; the acquisition unit 23 is specifically used to acquire map data based on the location instructions and acquire business data from the storage module 3 based on the logic event instructions. It can be understood that the business data can also be interacted through the cloud, that is, the acquisition unit can also directly acquire the corresponding business data from the cloud according to the logic event instructions. By setting up a storage module, local deployment of business data can be achieved, speeding up data transmission, and business data can also be directly acquired through cloud interaction to reduce local storage. The embodiment of the present disclosure can select the interaction method according to the specific situation, and the interaction method is flexible.

[0083] To be more specific, a unified API standard specification is formulated, and the traditional three-dimensional GIS engine is used for development according to the formulated API standard specification. The twin layer 241 maps and converts the GIS coordinates of the physical entity with the spatial coordinates of the virtual entity. The panel layer 242 generates corresponding display data based on the results of the linkage. For example, when the user instruction is "route from A to B", the twin layer maps the planned route in the physical entity to the coordinates of the virtual entity, and the panel layer 242 is used to generate display data corresponding to the planned route to display the corresponding planned route. For another example, when the user instruction is the current flow of people in place A, the twin layer 241 maps the coordinates of A in the physical entity to the coordinates of the virtual entity, and at the same time calls the current real-time flow of people data in place A. The twin layer 241 generates heat map data based on the real-time flow of people data to display the current flow of people in place A.

[0084] In some embodiments, the display data includes chart data or list data, and the panel layer 242 is specifically used to generate chart data based on the results of the linkage; or the panel layer 242 is specifically used to generate list data based on the results of the linkage.

[0085] Specifically, the display layer 242 can generate display data that is convenient for users to view based on the specific display content. For example, when the user instruction is to query the attractions in the park, the twin layer 241 maps the coordinates of the attractions in the physical entity to the coordinates of the virtual entity, and the panel layer 242 generates list data based on the coordinates of each attraction in the virtual entity, so as to display the attractions in the park in the form of a list. For another example, when the user instruction is the current flow of people in place A, the twin layer 241 maps the coordinates of A in the physical entity to the coordinates of the virtual entity, and calls the current real-time flow of people data in place A. The twin layer 241 generates heat map data based on the real-time flow of people data to display the current flow of people in place A. When the user instruction is to query the trend chart of the flow of people in place A in the next 2 hours, the twin layer 241 maps the coordinates of A in the physical entity to the coordinates of the virtual entity, and calls the predicted data of the real flow of people data in place A in the next 2 hours. The twin layer 241 generates corresponding curve data or data charts based on the predicted data of the flow of people data in the next 2 hours, so as to display the trend of the flow of people in place A in the next 2 hours. Of course, the display data in the embodiment of the present disclosure can also be in other forms of expression. For example, when the user instruction is a route from A to B, the twin layer maps the planned route in the physical entity to the coordinates of the virtual entity, and the panel layer 242 is used to generate handwriting data corresponding to the planned route in order to display the corresponding planned route.

[0086] The present invention also provides a park operation method, which is applied to the above-mentioned park operation system, such as Figure 4 As shown, including:

[0087] Step S1, the park operation system collects user instructions and inputs the language model into control instructions;

[0088] Step S2: the park operation system links the physical entities in the park with the virtual entities in the generated virtual scene according to the control instructions and outputs display data for display.

[0089] In a preferred embodiment of the present invention, Figure 5 As shown, step S1 includes:

[0090] Step S11, the park operation system extracts keywords from the user instruction;

[0091] In step S12, the park operation system converts the keywords into a language model to obtain control instructions, which include location instructions and logic event instructions.

[0092] The above are only preferred embodiments of the present invention, and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should be aware that all solutions obtained by equivalent substitutions and obvious changes made using the contents of this specification and illustrations should be included in the protection scope of the present invention.

Claims

1. A park operation system, characterized in that: include: A control instruction output module is used to collect user instruction input language models and convert them into control instructions; The data display module is connected to the control instruction output module and is used to link the physical entities in the park and the virtual entities in the generated virtual scene according to the control instruction and output display data for display.

2. The park operation system according to claim 1, characterized in that: The control instruction output module comprises: A keyword extraction unit, used to extract keywords from the user instruction; The instruction generation unit is connected to the keyword extraction unit and is used to input the keyword into the language model to convert it into the control instruction, and the control instruction includes a location instruction and a logic event instruction.

3. The park operation system according to claim 1, characterized in that: The control instruction includes a voice instruction and a text instruction, and the data display module includes: A voice broadcasting unit, used for broadcasting the voice form instruction to the user; The text sending unit is used to link the physical entities in the park and the virtual entities in the generated virtual scene according to the text form instructions and output display data for display.

4. The park operation system according to claim 2, characterized in that: The user instruction includes a voice instruction, a gesture instruction and a text instruction, and the keyword extraction unit includes: A text extraction subunit, used for extracting keywords from the text instruction; A speech recognition subunit, used for inputting the speech instruction into a speech recognition model to extract keywords in the speech instruction; The image recognition subunit is used to input the gesture instruction into an image recognition model to extract keywords in the gesture instruction.

5. The park operation system according to claim 4, characterized in that: The speech recognition model in the speech recognition subunit is an RNN-T prediction network model, the RNN-T prediction network model includes a prediction network, a decoding network, a joint model and a Softmax layer, and the decoding network includes 8 prediction networks; The voice instruction serves as the input of the prediction network, the input of the joint model is connected to the output of the decoding network and the output of the prediction network, the output of the joint model is connected to the input of the Softmax layer, and the output of the Softmax layer serves as the output of the speech recognition model.

6. The park operation system according to claim 4, characterized in that: The gesture instruction includes an upward / downward / leftward / rightward sliding action, and the image recognition model extracts the sliding action in the gesture instruction as the keyword by recognizing the gesture instruction; The instruction generation unit outputs a corresponding control instruction according to the keyword, and the data display module maps the physical entity in the park into a virtual entity according to the control instruction, and performs sliding in a corresponding proportional direction.

7. The park operation system according to claim 2, characterized in that: The data display module includes: An acquisition unit, configured to acquire map data based on the location instruction and to acquire business data based on the logic event instruction; A twin unit is connected to the acquisition unit and is used to link the physical entity with the virtual entity based on the map data and the business data and generate the display data for display.

8. The park operation system according to claim 7, characterized in that: The twin unit comprises: The twin layer is used to link physical entities with virtual entities based on map data and business data; The panel layer is connected to the twin layer and is used to generate the display data for display based on the result of the linkage.

9. A park operation method, characterized in that: The park operation system as claimed in any one of claims 1 to 8 comprises: Step S1, the park operation system collects user instructions and inputs the language model into control instructions; Step S2: the park operation system links the physical entities in the park with the virtual entities in the generated virtual scene according to the control instruction and outputs display data for display.

10. The park operation method according to claim 9, characterized in that: The step S1 comprises: Step S11, the park operation system extracts keywords from the user instruction; Step S12: the park operation system inputs the keyword into the language model to convert it into the control instruction, which includes a location instruction and a logic event instruction.

Images