Elevator control method and system

Abstract

The invention discloses an elevator control method and system, and belongs to elevator system design. And voice interaction with a user and personalized elevator control are realized by utilizing an artificial intelligence technology. The system comprises a voice acquisition and processing module which converts a user instruction into a structured text through voice recognition and ensures personalized service in combination with voiceprint recognition; the language understanding module performs intention recognition and context analysis on the instruction to generate intention and entity information; the personalized control module dynamically adjusts elevator operation parameters according to historical behaviors of a user; the safety verification module ensures that elevator operation meets the safety standard; the feedback and optimization module continuously optimizes an elevator control strategy according to user feedback, and the response speed and accuracy are improved. According to the invention, through intelligent and personalized control and optimization mechanisms, the user experience and safety of the home elevator are improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of elevator system design, and in particular to an elevator control method and system. Background Art

[0002] With the rise in height and intelligence of modern buildings, elevator systems are becoming increasingly important in daily life. This is especially true in villas, high-end residences, and large shopping malls. Elevators not only provide vertical transportation for people but also ensure comfort, efficiency, and safety. Traditional elevator systems, which mostly rely on physical buttons, are relatively simple to operate and offer limited functionality. However, with technological advancements, user demands for elevators are becoming increasingly diverse, with increasing expectations for convenience, personalized service, and safety.

[0003] Existing elevator systems face multiple technical challenges. Traditional elevator systems are controlled solely through physical buttons or simple touchscreens, lacking intelligent user interaction. With the increasing popularity of smart homes, users desire more natural and convenient ways to interact with elevators, such as voice control. However, existing technologies lack the ability to communicate with users through natural language.

[0004] Existing elevator systems typically use fixed preset floors and operating modes, rather than dynamically adjusting to individual user needs. This lack of intelligent control parameter adjustments based on user behavior, habits, and actual needs results in a poor user experience.

[0005] Elevator control systems must ensure strict safety regulations when executing control commands, particularly regarding target floor selection, passenger safety, and equipment protection during elevator operation. However, existing elevator systems often lack real-time safety verification mechanisms when generating control commands, potentially leading to unexpected situations.

[0006] Traditional elevator systems lack a real-time mechanism for processing user feedback after executing control commands. User feedback (e.g., whether the operation was completed smoothly and whether the user was satisfied) is not effectively utilized to optimize the elevator system, resulting in the inability to continuously improve the elevator control system during use. Summary of the Invention

[0007] The present invention provides an elevator control method and system. The technical solution is as follows.

[0008] According to one aspect of the present application, there is provided an elevator control system, the system comprising:

[0009] The voice collection and processing module is used to receive user voice commands, perform voice recognition and convert them into structured text commands, and output text data containing timestamp, voiceprint ID and confidence level;

[0010] The language understanding module is used to identify intent, extract entities, and analyze context for structured text instructions, generate intent, entity, and context information, and fill in relevant dialogue slots;

[0011] Personalized control module, used to generate personalized control parameters based on user historical behavior, including target floor, operation mode and priority;

[0012] Safety verification module, used to verify the legality and safety of personalized control parameters and generate elevator control instructions;

[0013] The feedback and optimization module is used to receive elevator execution results and optimize the system model based on user feedback.

[0014] Optionally, the semantic input preprocessing and intent recognition module includes:

[0015] A natural language processing unit, which uses a deep learning model to convert natural language questions into structured semantic representations;

[0016] Intent recognition unit, used to identify user intent and label the intent through artificial intelligence classification models;

[0017] Entity extraction unit, used to extract and construct entity lists from questions using named entity recognition technology;

[0018] A confidence assessment unit is used to use an artificial intelligence model to evaluate the accuracy of the extracted intent labels and entity lists and generate a confidence value.

[0019] Optionally, the language understanding module performs intent recognition, entity extraction, and context filling based on a natural language processing model, outputs intent, entity, and dialogue status, and passes them as input to subsequent modules.

[0020] Optionally, the personalized control module generates personalized elevator operating parameters through a deep learning model based on user behavior data, and adjusts the elevator operating mode, priority and target floor.

[0021] Optionally, the safety verification module ensures that the elevator control instruction complies with safety regulations through compliance checks, and generates an elevator control instruction with a target floor, an action type, and a safety code.

[0022] Optionally, the feedback and optimization module optimizes the elevator control system through a reinforcement learning algorithm and automatically adjusts the model and parameters to improve the elevator response speed and accuracy.

[0023] In another aspect, an elevator control method is provided, the method comprising:

[0024] Receive user voice commands and convert them into structured text commands through voice recognition, and output timestamp, voiceprint ID and confidence level;

[0025] Perform intent recognition, entity extraction, and context analysis on structured text instructions to generate intent, entity, and context information and fill in dialogue slots.

[0026] Generate personalized control parameters based on user historical data, including target floor, operation mode and priority;

[0027] Verify the legality and safety of personalized control parameters and generate elevator control instructions;

[0028] Receive feedback on elevator execution results and optimize the system model, performing self-learning and adjustments based on user feedback.

[0029] Optionally, the method further includes:

[0030] Use natural language processing technology to identify intent and entities, use context analysis to generate intent and conversation status, and fill conversation slots;

[0031] The method models user behavior through a deep learning model, generates personalized elevator control parameters, and improves user interaction experience;

[0032] The method uses a reinforcement learning algorithm to dynamically optimize the operation strategy of the elevator control system based on user feedback, continuously improving the operating efficiency and fault diagnosis accuracy of the elevator.

[0033] Artificial intelligence technology enables voice interaction with users and personalized elevator control. The system includes a voice acquisition and processing module, which converts user commands into structured text through voice recognition and integrates voiceprint recognition to ensure personalized service. A language understanding module identifies the intent and analyzes the context of commands, generating intent and entity information. A personalized control module dynamically adjusts elevator operating parameters based on historical user behavior. A safety verification module ensures that elevator operations meet safety standards. A feedback and optimization module continuously optimizes elevator control strategies based on user feedback, improving response speed and accuracy. This invention enhances the user experience and safety of home elevators through intelligent and personalized control and optimization mechanisms. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a flowchart of an elevator control method provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0035] To make the objectives, technical solutions and advantages of the present invention more clear, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0036] In this document, "plurality" refers to two or more. "And / or" describes a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone. The character " / " generally indicates an "or" relationship between the associated objects.

[0037] Example 1

[0038] An elevator control system is provided, comprising the following contents.

[0039] The voice collection and processing module is used to receive user voice commands, perform voice recognition and convert them into structured text commands, and output text data containing timestamp, voiceprint ID and confidence.

[0040] The language understanding module is used to perform intent recognition, entity extraction, and context analysis on structured text instructions, generate intent, entity, and context information, and fill in relevant dialogue slots.

[0041] The personalized control module is used to generate personalized control parameters based on the user's historical behavior, including the target floor, operation mode and priority.

[0042] The safety verification module is used to verify the legality and safety of personalized control parameters and generate elevator control instructions.

[0043] The feedback and optimization module is used to receive elevator execution results and optimize the system model based on user feedback.

[0044] The elevator control system comprises multiple modules, the most critical of which is the voice acquisition and processing module. When a user issues a voice command, such as "Go to the study on the second floor," the system converts the command into structured text using a built-in speech recognition algorithm. Voiceprint recognition technology then verifies the user's identity, ensuring the accuracy of the voice command and the correctness of the personalized service. The text command received by the system is accompanied by a timestamp, voiceprint ID, and confidence level, ensuring the precision of subsequent processing.

[0045] Through speech recognition technology based on deep neural networks (DNN) and long short-term memory networks (LSTM), user speech is accurately converted into text; voiceprint recognition technology is used to identify the user's identity so that personalized control can be performed based on the user's historical behavior and preferences; timestamp alignment and confidence scoring enable voice input to be accurately located in the system, improving the system's response speed and accuracy.

[0046] This module realizes efficient voice command acquisition, conversion and processing, provides accurate and real-time data support for subsequent elevator control, and meets users' needs for efficiency and personalization of smart home systems.

[0047] Example 2

[0048] Optionally, the semantic input preprocessing and intent recognition module includes the following contents.

[0049] The natural language processing unit is used to convert natural language questions into structured semantic representations using a deep learning model; the intent recognition unit is used to identify user intent and label intent through an artificial intelligence classification model; the entity extraction unit is used to extract and construct an entity list from the question through named entity recognition technology; the confidence assessment unit is used to use an artificial intelligence model to evaluate the accuracy of the extracted intent labels and entity lists and generate a confidence value.

[0050] The system's semantic input preprocessing and intent recognition module uses natural language processing (NLP) and deep learning models (such as BERT and GPT) to perform intent recognition and entity extraction. After converting user speech into structured text, the module uses named entity recognition (NER) technology to extract entities related to elevator control (such as the target floor and operating mode). It then uses an artificial intelligence classification model to identify the user's operational intent (such as "going upstairs" or "going downstairs").

[0051] A Transformer-based deep learning model is used to perform text intent recognition and entity extraction; NER processing is performed on user input text to extract entity information such as "second-floor study"; and a technology based on a combination of convolutional neural networks and recurrent neural networks is used to conduct confidence assessment on the results of intent recognition and entity extraction to ensure the accuracy of system output.

[0052] Through intent recognition and entity extraction, the system can accurately understand the user's elevator control needs and generate structured semantic representations to ensure the efficient execution of elevator control instructions.

[0053] Example 3

[0054] Optionally, the language understanding module performs intent recognition, entity extraction, and context filling based on a natural language processing model, outputs intent, entity, and dialogue status, and passes them as input to subsequent modules.

[0055] Example 4

[0056] Optionally, the personalized control module generates personalized elevator operating parameters through a deep learning model based on user behavior data, and adjusts the elevator operating mode, priority and target floor.

[0057] The personalized control module automatically adjusts elevator control parameters based on user behavior data, historical preferences, and usage habits. For example, if a user frequently uses the elevator during peak hours, the system will automatically select the fastest operating mode or slow down the elevator in advance when the user reaches a specific floor to improve the user experience. A deep learning model analyzes the user's historical usage data to generate a user profile; based on this user profile, elevator operating parameters such as priority, target floor, and door opening and closing times are adjusted. The personalized control module provides personalized elevator service, which not only improves the user experience but also enables the elevator system to operate more efficiently in high-demand scenarios.

[0058] Example 5

[0059] Optionally, the safety verification module ensures that the elevator control instruction complies with safety regulations through compliance checks, and generates an elevator control instruction with a target floor, an action type, and a safety code.

[0060] The safety verification module performs compliance checks on generated elevator control commands to ensure they meet elevator safety standards. This module uses a predefined safety rule library to verify the target floor, action type, and safety code to prevent dangerous operations such as illegal target floors and rapid lifting and lowering.

[0061] Through the safety rule library, it is ensured that the generated control instructions comply with the hardware and software safety standards of the elevator; according to the type of control instruction, the corresponding safety code is generated to ensure the legality of the execution of the instruction.

[0062] This module adds a layer of security to the elevator control system, ensuring that each control instruction is verified for legitimacy and safety when executed, greatly improving the safety of elevator operation.

[0063] Example 6

[0064] Optionally, the feedback and optimization module optimizes the elevator control system through a reinforcement learning algorithm and automatically adjusts the model and parameters to improve the elevator response speed and accuracy.

[0065] The Feedback and Optimization module automatically learns and adjusts the elevator control model using a reinforcement learning algorithm. For example, after an elevator executes a control command within a certain time period, the system adjusts the elevator's response strategy based on user feedback (such as satisfaction ratings and fault reports) to optimize its operating efficiency. This module automatically adjusts the elevator control strategy using reinforcement learning algorithms (such as Deep Q-Networks (DQN)). Based on user feedback, the elevator's operating strategy and behavior patterns are dynamically adjusted. The Feedback and Optimization module enables the elevator control system to continuously optimize during use, adapting to changing user needs and equipment status, thereby improving overall efficiency and user satisfaction.

[0066] Example 7

[0067] In another aspect, an elevator control method is provided, the method comprising:

[0068] Step 201: Receive user voice commands and convert them into structured text commands through voice recognition, and output timestamp, voiceprint ID and confidence level.

[0069] Through efficient speech recognition technology, the user's voice commands are converted into structured text commands in real time. The system uses deep learning models (such as convolutional neural networks (CNN) and recurrent neural networks (RNN)) to analyze voice signals to obtain accurate text conversion results. In addition, the system also introduces voiceprint recognition technology to ensure the personalization and security of commands, and identifies the user's identity based on the user's unique voiceprint ID. To ensure that the converted commands have a time dimension, the system will attach a timestamp to each command to facilitate the subsequent processing of the sequential execution of the commands. Furthermore, through a confidence assessment mechanism (based on the fusion of deep neural networks and acoustic models), each voice command will be assigned a corresponding confidence value to indicate the accuracy of its recognition. This technical solution is innovative in elevator control. First, it realizes accurate voice-to-text conversion during user interaction, and enhances personalized services and security through voiceprint recognition.

[0070] By combining voiceprint recognition technology with voice command conversion, not only can the command content be identified, but the identity of the command source can also be ensured. This innovation enhances the personalized service of the elevator system. The introduction of timestamps and confidence outputs not only ensures the timing and execution accuracy of commands, but also provides reliable data support and feedback mechanisms for subsequent modules.

[0071] Step 202: perform intent recognition, entity extraction, and context analysis on the structured text instructions to generate intent, entity, and context information, and fill in the dialogue slots.

[0072] In this step, the system uses a natural language processing (NLP) model to conduct an in-depth analysis of structured text instructions. The system first uses a deep learning model (such as BERT or Transformer) to identify intent and convert the user's input into commands that the elevator system can understand. Next, named entity recognition (NER) technology is used to extract relevant entities (such as target floor, operating mode, priority, etc.) from the text. These entities are crucial for elevator control instructions. Subsequently, the system uses context analysis to ensure that the instructions can be accurately embedded in the conversation state and historical context, further filling in the relevant conversation slots. This slot will be flexibly updated according to the needs of different users and the current situation to provide effective information input for subsequent modules.

[0073] Combining contextual analysis and intent recognition, the elevator system not only identifies the user's current intent but also understands the underlying needs and context, ensuring that the system's responses are more aligned with the user's actual needs. By automatically filling conversation slots and dynamically updating conversation status, the system enables multiple rounds of interaction, maintaining a continuous dialogue with the user, and improving system flexibility and interactive experience.

[0074] Step 203: Generate personalized control parameters based on the user's historical data, including target floor, operation mode and priority.

[0075] The personalized control module relies on deep learning algorithms and user behavior data to generate personalized elevator operating parameters. By modeling users' historical behavior, the system can identify their usage patterns, preferences, and needs (such as frequently used floors and travel time periods). For example, the system may discover that a user typically chooses to reach the top floor in the morning and selects a preferred operating mode (such as energy-saving mode) during the evening rush hour. Based on these behavioral patterns, the system learns from the user's historical data through deep learning (such as LSTM networks) and dynamically adjusts the elevator's control parameters. Through these personalized adjustments, the system can automatically optimize the elevator's operating efficiency while providing users with a personalized elevator experience.

[0076] Personalized modeling with deep learning models: Utilizing deep learning models (such as LSTM) to automatically learn and generate personalized elevator control parameters, the system can dynamically adjust its operating mode based on user habits, enhancing system intelligence. The system can dynamically adjust control parameters based on user historical behavior in real time without manual intervention, improving elevator automation and convenience.

[0077] Step 204: verify the legality and safety of the personalized control parameters and generate elevator control instructions.

[0078] The safety verification module verifies the legality and safety of customized control parameters to ensure that elevator operation poses no safety risks. The system first verifies that control parameters such as the target floor, operating mode, and passenger priority meet safety standards and regulations for elevator operation through compliance check rules. For example, during peak hours, the system automatically verifies whether there is an overload risk and restricts the elevator's rapid ascent. Furthermore, the system generates safety codes related to elevator control to ensure that every control instruction is effectively executed and does not cause elevator equipment failure or accidents.

[0079] Through customized compliance checking algorithms, the system automatically detects and prevents potential safety hazards, ensuring that elevator operations comply with national safety standards. By generating a safety code for each control command, the elevator control system not only verifies command legitimacy but also enhances safety and traceability during the elevator control process.

[0080] Step 205: Receive elevator execution result feedback and optimize the system model, and perform self-learning and adjustment through user feedback.

[0081] The Feedback and Optimization module uses a reinforcement learning algorithm to enable the elevator system to receive timely user feedback after executing control commands and use this feedback to optimize the system model. For example, if a user fails to reach their destination floor on time, the system analyzes the cause (perhaps due to improper elevator scheduling or incorrect user priority settings) and automatically adjusts the elevator scheduling algorithm and control strategy. Furthermore, the system continuously learns and adjusts parameters to improve response speed and accuracy, further enhancing the user experience.

[0082] Through reinforcement learning, the system automatically adjusts elevator control strategies based on user feedback and continuously optimizes them during actual operation. This self-learning and adjustment mechanism enables the elevator system to not only resolve current problems but also predict and prevent future failures. By obtaining real-time user feedback and adjusting its strategies accordingly, the system continuously improves its fault diagnosis capabilities and response efficiency, significantly enhancing the user experience.

[0083] This method receives user voice commands and performs speech recognition, converting them into structured text. The system then uses natural language processing models to identify intent and analyze context, generating personalized control parameters and performing safety checks. Finally, it continuously optimizes the elevator control strategy based on user feedback.

[0084] The system uses automatic speech recognition (ASR) technology to convert speech into structured text; uses NLP and deep learning models to perform intent recognition and contextual reasoning on text; and dynamically adjusts control parameters based on user portraits and historical data to provide personalized services.

[0085] This method significantly improves the intelligence of the elevator system and user experience through intelligent voice interaction, personalized control and self-optimization mechanism.

[0086] Optionally, the method further includes:

[0087] Use natural language processing technology to identify intent and entities, use context analysis to generate intent and conversation status, and fill conversation slots;

[0088] The method models user behavior through a deep learning model, generates personalized elevator control parameters, and improves user interaction experience;

[0089] The method uses a reinforcement learning algorithm to dynamically optimize the operation strategy of the elevator control system based on user feedback, continuously improving the operating efficiency and fault diagnosis accuracy of the elevator.

[0090] This method uses natural language processing technology to identify intent and entities, combining contextual analysis to generate intent, entities, and conversation states. A deep learning model is used to model user behavior and generate personalized elevator control parameters, enhancing the user interaction experience.

[0091] A deep learning-based NLP model extracts intent and entities from text. Contextual analysis using LSTM or Transformer models ensures efficient interaction within the elevator control system. Furthermore, the method uses a reinforcement learning algorithm to dynamically optimize the elevator control system's operational strategy based on user feedback. By continuously learning from user behavior and feedback, the system achieves self-optimization of the elevator control strategy.

[0092] The above description is merely an optional embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. An elevator control system, characterized in that: The system comprises: The voice collection and processing module is used to receive user voice commands, perform voice recognition and convert them into structured text commands, and output text data containing timestamp, voiceprint ID and confidence level; The language understanding module is used to identify intent, extract entities, and analyze context for structured text instructions, generate intent, entity, and context information, and fill in relevant dialogue slots; Personalized control module, used to generate personalized control parameters based on user historical behavior, including target floor, operation mode and priority; Safety verification module, used to verify the legality and safety of personalized control parameters and generate elevator control instructions; The feedback and optimization module is used to receive elevator execution results and optimize the system model based on user feedback.

2. The elevator control system according to claim 1, characterized in that: The semantic input preprocessing and intention recognition module includes: A natural language processing unit, which uses a deep learning model to convert natural language questions into structured semantic representations; Intent recognition unit, used to identify user intent and label the intent through artificial intelligence classification models; Entity extraction unit, used to extract and construct entity lists from questions using named entity recognition technology; A confidence assessment unit is used to use an artificial intelligence model to evaluate the accuracy of the extracted intent labels and entity lists and generate a confidence value.

3. The elevator control system according to claim 1, characterized in that: The language understanding module performs intent recognition, entity extraction, and context filling based on the natural language processing model, outputs intent, entity, and dialogue status, and passes them as input to subsequent modules.

4. The elevator control system according to claim 1, characterized in that: The personalized control module generates personalized elevator operating parameters based on user behavior data through a deep learning model, and adjusts the elevator operation mode, priority and target floor.

5. The elevator control system according to claim 1, characterized in that: The safety verification module ensures that the elevator control instruction complies with safety regulations through compliance checks and generates an elevator control instruction with a target floor, action type and safety code.

6. The elevator control system according to claim 1, characterized in that: The feedback and optimization module optimizes the elevator control system through a reinforcement learning algorithm and automatically adjusts the model and parameters.

7. An elevator control method, characterized in that: The method comprises: Receive user voice commands and convert them into structured text commands through voice recognition, and output timestamp, voiceprint ID and confidence level; Perform intent recognition, entity extraction, and context analysis on structured text instructions to generate intent, entity, and context information and fill in dialogue slots. Generate personalized control parameters based on user historical data, including target floor, operation mode and priority; Verify the legality and safety of personalized control parameters and generate elevator control instructions; Receive feedback on elevator execution results and optimize the system model, performing self-learning and adjustments based on user feedback.

8. The elevator control method according to claim 7, characterized in that: The method further comprises: Use natural language processing technology to identify intent and entities, use context analysis to generate intent and conversation status, and fill conversation slots; The method models user behavior through a deep learning model and generates personalized elevator control parameters; The method dynamically optimizes the operation strategy of the elevator control system according to user feedback through a reinforcement learning algorithm.

Images