Intelligent dialogue generation method oriented to elevator emergency rescue and based on InterHT model

Abstract

The embodiment of the invention relates to the technical field of intelligent dialogue generation, in particular to an intelligent dialogue generation method oriented to elevator emergency rescue and based on an InterHT model, and the method comprises the steps: building a plurality of triples oriented to elevator emergency rescue based on collected big data related to elevator emergency rescue and analyzed business characteristics; based on an InterHT model, adding self-interaction between the head entity and the tail entity of each triad, then adding interaction between the head entity and the tail entity of the associated triad, constructing an artificial intelligence fusion knowledge base, and deploying the artificial intelligence fusion knowledge base to an emergency rescue command center; when the emergency rescue command center receives an emergency rescue request, the Internet of Things emergency call-for-help device corresponding to the emergency rescue request is connected, and voice data in an elevator car where the Internet of Things emergency call-for-help device is located are collected; and based on the collected voice data, performing intelligent dialogue generation by using a voice recognition model and an artificial intelligence fusion knowledge base, and performing intelligent dialogue with trapped passengers.

Description

Technical Field

[0001] The embodiments of this application relate to the field of intelligent dialogue generation technology, and in particular to an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue. Background Technology

[0002] Intelligent dialogue generation is an interactive technology built on natural language processing (NLP) and deep learning (DL) that can simulate human dialogue. Its core goal is to understand user intent and generate reasonable and coherent responses.

[0003] In elevator emergency rescue scenarios, intelligent dialogue generation technology is typically deployed as a "task-based dialogue system," primarily including pipeline systems and end-to-end systems. Currently deployed intelligent dialogue generation systems for elevator emergency rescue can achieve functions such as proactive distress call recognition, emotional reassurance guidance, and multimodal data collaborative analysis, making a significant contribution to improving the success rate of elevator emergency rescues.

[0004] Although intelligent dialogue generation technology has been widely used in elevator emergency rescue scenarios, it still faces four major technical bottlenecks. First, it suffers from poor robustness in intent recognition under complex environments. Elevator shaft signal interference, differences in trapped passengers' accents, and environmental noise (such as knocking on doors) can all lead to speech recognition errors. Furthermore, the unclear expressions of trapped passengers when emotionally agitated (such as crying) may be misjudged as non-emergency events. Second, it lacks coverage of passive distress scenarios, failing to recognize unconscious passengers, deaf individuals, etc. Third, it faces the complexity of multi-source data fusion and real-time decision-making. Intelligent dialogue generation requires simultaneous processing of voice, video, and sensor data, but the accuracy of cross-modal alignment algorithms is insufficient, easily leading to latency. Fourth, it has limited generalization ability in extreme scenarios. Training data for rare fault cases (such as broken steel cables or multiple overlapping faults) is extremely scarce, which may result in incorrect guidance generated during intelligent dialogue generation, potentially endangering trapped passengers.

[0005] A knowledge graph (KG), also known as knowledge visualization or knowledge mapping map, can display a series of different graphs representing data generation, service processes, and structural relationships. Knowledge graphs utilize visualization techniques to describe knowledge resources and their carriers, mining, analyzing, constructing, drawing, and displaying knowledge and the interrelationships between them. A knowledge graph is a structured semantic knowledge base used to describe concepts and their relationships in the physical world in symbolic form. Its basic building blocks are the "entity, relation, entity" triples, as well as entities and their associated attribute value pairs. Entities are interconnected through relations, forming a network-like knowledge structure.

[0006] The knowledge graph-based intelligent dialogue generation technology has made significant breakthroughs in overcoming the aforementioned technical bottlenecks. It can achieve multi-source data fusion, accurately diagnose complex elevator malfunctions, reduce the probability of incorrect guidance, and improve the success rate of elevator emergency rescue.

[0007] However, the primary requirement for elevator emergency rescue is speed. The traditional knowledge graph construction and search process is cumbersome, complex, and time-consuming, and cannot be perfectly adapted to elevator emergency rescue scenarios. Summary of the Invention

[0008] In view of this, embodiments of this application propose an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue. The method aims to utilize the InterHT model to identify massive amounts of data, complex logical relationships, and various entity scenarios, effectively uncover hidden relationships and representations, achieve accurate speech understanding, logical analysis, and intelligent dialogue generation, and effectively improve the success rate of elevator emergency rescue.

[0009] Firstly, embodiments of this application propose an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue, applicable to emergency rescue command centers. This method includes: collecting big data related to elevator emergency rescue and performing business characteristic analysis; constructing several triples composed of entities, relationships, and entities based on the collected big data and analyzed business characteristics; adding self-interaction between the head and tail entities of each triple based on the InterHT model, and adding mutual interaction between the head and tail entities of related triples, thus constructing an AI-integrated knowledge base and deploying it to the emergency rescue command center; when the emergency rescue command center receives an emergency rescue request, connecting the corresponding IoT emergency call device and collecting voice data from the elevator car; based on the collected voice data, generating intelligent dialogue using a speech recognition model and the AI-integrated knowledge base, engaging in intelligent dialogue with trapped passengers to both soothe their emotions and determine the elevator malfunction type based on the content of the intelligent dialogue, thereby dispatching the corresponding rescue team to the scene.

[0010] This application proposes an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue. Starting from the analysis of big data and business characteristics related to elevator emergency rescue, it further develops knowledge graph technology based on the InterHT model. The InterHT model overcomes the bottleneck of large-scale complex graph prediction technology with multiple relationships and modes. Through self-interaction and mutual interaction between head and tail entities, it strengthens entity semantics and relation representation, focuses on different entity roles, deeply mines hidden information, and improves the effect of knowledge graph prediction. The AI-powered knowledge base constructed using the InterHT model can identify massive amounts of data information and complex logical relationships and entity scenarios, achieve new triplet prediction and verification, and effectively mine hidden relationships and representations. This enables speech understanding, logical analysis, knowledge engine search, and intelligent dialogue generation, achieving deep thinking logic similar to human judgment. The intelligent dialogue generated based on the AI-powered knowledge base can soothe the emotions of trapped passengers and quickly dispatch the corresponding rescue team to the scene, effectively improving the success rate of elevator emergency rescue.

[0011] Optionally, the collected big data related to elevator emergency rescue includes: basic information, basic operation data and historical fault data of elevators within the jurisdiction of the emergency rescue command center. The basic information of the elevator includes elevator brand, service life, elevator size and working scenario. The basic operation data of the elevator includes rated load, rated speed, running time, car size, number of floors and drive mode. The historical fault data of the elevator includes fault type, fault occurrence time, fault root cause, economic loss and emergency rescue records. The collected big data related to elevator emergency rescue was analyzed for business characteristics, specifically as follows: Determine the relationship between the elevator's basic information and basic operating data and historical fault data.

[0012] Optionally, based on the InterHT model, self-interactions are added between the head and tail entities of each triple, and mutual interactions are added between the head and tail entities of related triples, thus constructing an AI-powered knowledge base and deploying it to the emergency rescue command center, including: For the Three pairs , Based on its tail entity Generate an auxiliary tail entity vector and based on its head entity Generate an auxiliary head entity vector At the same time, a unit moment is determined for it. ;in, , and They represent the first The head entity, relation, and tail entity of a triple; Based on the following formula, and right Add self-interaction, based on and right Add self-interaction, and get the result after adding self-interaction. Three pairs : ; ; ; in, This indicates that a composite function operation is being performed. Indicates the first Add a head entity to each triple after self-interaction. Indicates the first Add a tail entity to each triple after self-interaction; Determine all The associated triples after adding self-interactions are calculated using the following formula, based on all triples related to... The associated auxiliary tail entity vector and unit moment of the triple after adding self-interaction, for Add interaction, based on all and The associated auxiliary head entity vector and unit moment of the triple after adding self-interaction, for Add interaction, get the result after adding interaction Three pairs : ; ; ; in, , and They represent the first Individual and The associated auxiliary tail entity vector, auxiliary head entity vector, and unit moment of the triple after adding self-interaction. Indicates all with The total number of associated triples after adding self-interactions. Indicates the first The head entity after adding interaction to a triple. Indicates the first Add an interactive tail entity to a triple; By integrating all the triples with added interactions, an AI-powered knowledge base is constructed and deployed to the emergency rescue command center.

[0013] Optionally, in the process of integrating all the triples after adding interactions to construct the AI-integrated knowledge base, the InterHT model is optimized using the following loss function: ; in, Represents the loss function. This represents the total number of triples. This represents the sigmoid function. This represents the preset fixed boundary parameters. This represents a function that calculates distance.

[0014] Optionally, based on all the triples with added interactions, an AI-powered knowledge base is constructed and deployed to the emergency rescue command center, including: By integrating all the triples with added interactions, an AI-integrated knowledge base is constructed. The constructed AI-integrated knowledge base is then used for this purpose. index, Indicators and The calculation of the indicators, in index, Indicators and If all indicators meet the preset design requirements, the constructed AI-integrated knowledge base will be deployed to the emergency rescue command center; otherwise, a new AI-integrated knowledge base will be constructed by re-integrating all the triples with added interactions.

[0015] Optionally, the IoT emergency call device is installed on the side wall of the elevator car, and the IoT emergency call device is equipped with a one-button emergency call button, a display screen, a speaker, a sound receiver and a camera; When a trapped passenger triggers the one-click emergency call button, the IoT emergency call device immediately sends an emergency rescue request to the emergency rescue command center. When the elevator's self-checking device confirms that a person is trapped in the elevator, and the one-button emergency call device of the Internet of Things (IoT) is not triggered within the preset reaction time, the IoT emergency call device automatically sends an emergency rescue request to the emergency rescue command center.

[0016] Optionally, the emergency rescue command center is equipped with a DFCNN-based speech recognition model and emotion recognition model. Based on the collected speech data, it uses the speech recognition model and an AI-integrated knowledge base to generate intelligent dialogue and engage in intelligent dialogue with the stranded passengers, including: Using a DFCNN-based speech recognition model, speech recognition is performed on the collected speech data, and the collected speech data is converted into text data. Using an emotion recognition model, emotion recognition is performed based on the collected voice data and the converted text data to obtain the current emotion of the trapped passengers; Based on the converted text data and the current emotions of the trapped passengers, the system mines hidden relationships and searches for and matches knowledge in the AI-powered knowledge base to generate dialogue audio data corresponding to the converted text data. This audio data is then sent to the corresponding IoT distress call device to enable intelligent dialogue with the trapped passengers.

[0017] Secondly, embodiments of this application propose an intelligent dialogue generation system based on the InterHT model for elevator emergency rescue, suitable for emergency rescue command centers. The system includes: an analysis module, a triplet construction module, a knowledge base construction module, a rescue positioning module, and a dialogue generation module. The analysis module collects big data related to elevator emergency rescue and performs business characteristic analysis. The triplet construction module constructs several triples composed of entities, relations, and entities for elevator emergency rescue based on the collected big data and analyzed business characteristics. The knowledge base construction module, based on the InterHT model, assigns a head entity to each triple. Self-interaction is added between the head and tail entities, and mutual interaction is added between the head and tail entities of the associated triples to build an AI-powered knowledge base, which is then deployed to the emergency rescue command center. The rescue positioning module is used to connect to the IoT emergency call device corresponding to the emergency rescue request received by the emergency rescue command center and collect voice data from the elevator car where it is located. The dialogue generation module is used to generate intelligent dialogue based on the collected voice data, using a voice recognition model and the AI-powered knowledge base, to conduct intelligent dialogue with the trapped passengers. On the one hand, it soothes the trapped passengers' emotions, and on the other hand, it determines the type of elevator malfunction based on the content of the intelligent dialogue so as to dispatch the corresponding rescue team to the scene for rescue.

[0018] Thirdly, embodiments of this application propose an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to execute a smart dialogue generation method based on the InterHT model for elevator emergency rescue as described above.

[0019] Fourthly, embodiments of this application propose a computer-readable storage medium storing a computer program that, when executed by a processor, enables a smart dialogue generation method based on the InterHT model for elevator emergency rescue as described above.

[0020] It is understood that the beneficial effects of the second to fourth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies of this application will be briefly introduced below. Obviously, the following drawings are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. The drawings described herein are only used to explain this application and are not intended to limit this application.

[0022] Figure 1 This is a flowchart of an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue provided in one embodiment of this application; Figure 2 This is a schematic diagram of an InterHT model provided in one embodiment of this application; Figure 3 This is a schematic diagram of an IoT emergency call device provided in one embodiment of this application; Figure 4 This is a schematic diagram of the structure of an intelligent dialogue generation system based on the InterHT model for elevator emergency rescue provided in another embodiment of this application; Figure 5 This is a schematic diagram of the structure of an electronic device provided in another embodiment of this application. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the various embodiments of this application will be described in detail below with reference to the accompanying drawings. Those skilled in the art will understand that many technical details have been presented in the embodiments of this application to facilitate better understanding. However, the technical solutions claimed in this application can be implemented even without these technical details and various variations and modifications based on the following embodiments. The division of the following embodiments is for ease of description and should not constitute any limitation on the specific implementation of this application. The following embodiments can be combined with and referenced by each other without contradiction.

[0024] One embodiment of this application proposes an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue, applicable to emergency rescue command centers. The implementation details of this embodiment's intelligent dialogue generation method based on the InterHT model are described below. These details are provided for ease of understanding and are not essential for implementing this solution.

[0025] The specific process of the intelligent dialogue generation method based on the InterHT model for elevator emergency rescue proposed in this embodiment can be described as follows: Figure 1 As shown, it includes: Step 11: Collect big data related to elevator emergency rescue and conduct business characteristic analysis.

[0026] In practical implementation, building an AI-integrated knowledge base is the foundation for generating intelligent dialogues for elevator emergency rescue. The first step in building the AI-integrated knowledge base is to collect massive amounts of big data related to elevator emergency rescue and to analyze the business characteristics of the collected big data.

[0027] In one example, the collected big data related to elevator emergency rescue includes: basic information, basic operational data, and historical fault data of elevators within the jurisdiction of the emergency rescue command center. The basic elevator information includes elevator brand, service life, elevator size, and operating environment. The basic operational data includes rated load capacity, rated speed, operating time, car size, number of floors, and drive method. The historical fault data includes fault type, fault occurrence time, root cause, economic loss, and emergency rescue records.

[0028] In one example, basic elevator information can be used for secondary classification. For instance, based on service life, elevators can be classified as new elevators, elevators with a healthy service life, and old elevators. Elevators less than one year old are classified as new, those between one and ten years old as elevators with a healthy service life, and those over ten years old as old elevators. Another example is based on elevator size, classifying them as small, medium, and large elevators. Elevators with a load capacity less than 600 kg are classified as small, those between 600 kg and 1300 kg as medium, and those over 1300 kg as large. Yet another example is based on the work environment, classifying elevators as commercial elevators, industrial elevators, sightseeing elevators, residential elevators, office elevators, and medical elevators. Additionally, each elevator brand can also correspond to a secondary classification.

[0029] In one example, business characteristic analysis is performed on the collected big data related to elevator emergency rescue, specifically to determine the relationship between the elevator's basic information and basic operating data and historical fault data.

[0030] Step 12: Based on the collected big data related to elevator emergency rescue and the analyzed business characteristics, construct several triples consisting of entities, relationships, and entities for elevator emergency rescue.

[0031] In practical implementation, after analyzing the business characteristics of elevator emergency rescue, several triples consisting of entities, relationships, and entities can be constructed based on the collected big data related to elevator emergency rescue and the analyzed business characteristics.

[0032] In one example, each triple consists of a head entity, a relation, and a tail entity, assuming a total of [number] triples are constructed. The triplet, the first A triplet can be represented as , , , and They represent the first The triple consists of a head entity, a relation, and a tail entity.

[0033] Step 13: Based on the InterHT model, add self-interaction between the head and tail entities of each triple, and add mutual interaction between the head and tail entities of related triples to build an AI-integrated knowledge base and deploy it to the emergency rescue command center.

[0034] In the specific implementation, after the construction of all triples is completed, self-interactions can be added between the head and tail entities of each triple based on the InterHT model, and mutual interactions can be added between the head and tail entities of related triples, thereby constructing an AI-integrated knowledge base, which is then deployed to the emergency rescue command center.

[0035] In one example, the process of adding self-interactions and mutual interactions to triples based on the InterHT model can be as follows: Figure 2 As shown. For the first Three pairs In contrast, the InterHT model is based on its tail entity. Generate an auxiliary tail entity vector Based on its head entity Generate an auxiliary head entity vector At the same time, a unit moment is determined for it. .

[0036] Next, based on the following formula, and right Add self-interaction, based on and right Add self-interaction, and get the result after adding self-interaction. Three pairs : ; ; ; in, This indicates that a composite function operation is being performed. Indicates the first Add a head entity to each triple after self-interaction. Indicates the first Add a tail entity to each triple after self-interaction.

[0037] After adding self-interactions, you can add mutual interactions. The InterHT model determines all interactions with... The associated triples, after self-interaction, are then processed using the following formula, based on all... The associated auxiliary tail entity vector and unit moment of the triple after adding self-interaction, for Add interaction, based on all and The associated auxiliary head entity vector and unit moment of the triple after adding self-interaction, for Add interaction, get the result after adding interaction Three pairs : ; ; ; in, , and They represent the first Individual and The associated auxiliary tail entity vector, auxiliary head entity vector, and unit moment of the triple after adding self-interaction. Indicates all with The total number of associated triples after adding self-interactions. Indicates the first The head entity after adding interaction to a triple. Indicates the first The tail entity after adding interactions to the triples. The score function of the triples after adding interactions can be expressed by the formula: The InterHT model enhances entity semantics and relational representation through self-interaction and mutual interaction between head and tail entities, focuses on different entity roles, can deeply mine hidden information, and effectively improve the prediction effect of knowledge graphs.

[0038] Finally, based on the integration of all the triples with added interactions, an AI-powered knowledge base was constructed and deployed to the emergency rescue command center.

[0039] In one example, the server can use NodePiece to learn a fixed-size entity vocabulary to reduce the number of parameters and thus improve the speed of building an AI-powered knowledge base.

[0040] In one example, during the process of constructing the AI-integrated knowledge base by integrating all the triples after adding interactions, the InterHT model is optimized using the following loss function: ; in, Represents the loss function. This represents the total number of triples. This represents the sigmoid function. This represents the preset fixed boundary parameters. This indicates calculating the distance. express and The distance between them Then it means and The distance between them.

[0041] In one example, after adding all the interactions to the triples, the server can integrate all the triples with added interactions to build an AI-integrated knowledge base. This knowledge base can then be used for... index, Indicators and The calculation of the indicators, in index, Indicators and If all indicators meet the preset design requirements, the constructed AI-integrated knowledge base will be deployed to the emergency rescue command center; otherwise, a new AI-integrated knowledge base will be constructed by re-integrating all the triples with added interactions.

[0042] In one example The formula for calculating the indicator is: ; in, Indicates the first Link prediction ranking for each triplet express index, The higher the value of the indicator, the better.

[0043] In one example The formula for calculating the indicator is:

[0044] in, express index, The smaller the value of the indicator, the better.

[0045] In one example The formula for calculating the indicator is: ; in, This refers to the indicator function. express The preset in value, express index, The higher the value of the indicator, the better.

[0046] Step 14: When the emergency rescue command center receives an emergency rescue request, it connects the IoT emergency call device corresponding to the emergency rescue request and collects the voice data inside the elevator car where the request is located.

[0047] In practice, when the emergency rescue command center receives an emergency rescue request, it can immediately connect to the IoT emergency call device corresponding to the emergency rescue request and collect the voice data inside the elevator car where the request is located.

[0048] In one example, the structure of an IoT emergency call device is as follows: Figure 3As shown, the IoT emergency call device is specifically installed on the side wall of the elevator car. It includes a one-button emergency call, a display screen, a speaker, a microphone, and a camera. The one-button emergency call is located in the center below the display screen, with the speaker and microphone positioned to its left and right, respectively. The camera is located in the center above the display screen. When no malfunction occurs, the display screen and speaker work together to play pre-set elevator safety rules and advertising videos, while the camera remains off. In other words, in daily use, the IoT emergency call device can function as an advertising board within the elevator. When a trapped passenger triggers the one-button emergency call, the IoT emergency call device immediately sends an emergency rescue request to the emergency rescue command center. When the elevator's self-checking device (which establishes a communication connection with the IoT emergency call device) confirms a passenger entrapment malfunction, and the one-button emergency call is not triggered within a preset reaction time, the IoT emergency call device automatically sends an emergency rescue request to the emergency rescue command center.

[0049] Step 15: Based on the collected voice data, intelligent dialogue is generated using a voice recognition model and an AI-powered knowledge base. This dialogue is then used to reassure the trapped passengers and determine the type of elevator malfunction based on the content of the dialogue, so that the appropriate rescue team can be dispatched to the scene.

[0050] In practice, after collecting voice data, the emergency rescue command center can generate intelligent dialogue based on the collected voice data using a voice recognition model and an AI-integrated knowledge base. This allows them to engage in intelligent dialogue with trapped passengers, both to calm them down and to determine the type of elevator malfunction based on the content of the intelligent dialogue, so that the appropriate rescue team can be dispatched to the scene for rescue.

[0051] In one example, the emergency rescue command center is equipped with a DFCNN-based speech recognition model and an emotion recognition model. After collecting speech data, the command center uses the DFCNN-based speech recognition model to perform speech recognition and convert the collected speech data into text data. Then, using the emotion recognition model, it performs emotion recognition based on the collected speech data and the converted text data to determine the current emotion of the trapped passenger. Finally, based on the converted text data and the trapped passenger's current emotion, it mines latent relationships and performs knowledge search and matching in an AI-powered knowledge base to generate dialogue audio data corresponding to the converted text data, which is then sent to the corresponding IoT distress call device to conduct intelligent dialogue with the trapped passenger.

[0052] This embodiment proposes an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue. Starting from the analysis of big data and business characteristics related to elevator emergency rescue, it further develops knowledge graph technology based on the InterHT model. The InterHT model overcomes the bottleneck of large-scale complex graph prediction technology with multiple relationships and modes. Through self-interaction and mutual interaction between head and tail entities, it strengthens entity semantics and relation representation, focuses on different entity roles, deeply mines hidden information, and improves the effect of knowledge graph prediction. The AI-integrated knowledge base constructed using the InterHT model can identify massive amounts of data information and complex logical relationships and entity scenarios, achieve new triplet prediction and verification, and effectively mine hidden relationships and representations. This enables speech understanding, logical analysis, knowledge engine search, and intelligent dialogue generation, achieving deep thinking logic similar to human judgment. The intelligent dialogue generated based on the AI-integrated knowledge base can soothe the emotions of trapped passengers and quickly dispatch the corresponding rescue team to the scene, effectively improving the success rate of elevator emergency rescue.

[0053] The steps described above are for clarity only. In implementation, they can be combined into one step, or some steps can be broken down into multiple steps, as long as they involve the same logical relationship, they are all within the scope of protection of this application. Adding insignificant modifications or introducing insignificant designs to the algorithm or process, without changing the core design of the algorithm and process, are also within the scope of protection of this application.

[0054] Another embodiment of this application proposes an intelligent dialogue generation system based on the InterHT model for elevator emergency rescue, which is suitable for emergency rescue command centers. The details of the intelligent dialogue generation system based on the InterHT model for elevator emergency rescue proposed in this embodiment are described below. The following content is only for the convenience of understanding and is not necessary for implementing this example.

[0055] Figure 4 This is a schematic diagram of the structure of an intelligent dialogue generation system based on the InterHT model for elevator emergency rescue proposed in this embodiment, including: analysis module 21, triplet construction module 22, knowledge base construction module 23, rescue positioning module 24, and dialogue generation module 25.

[0056] Analysis module 21 is used to collect big data related to elevator emergency rescue and conduct business characteristic analysis.

[0057] The triplet construction module 22 is used to construct several triplets composed of entities, relationships, and entities for elevator emergency rescue based on the collected big data related to elevator emergency rescue and the analyzed business characteristics.

[0058] The knowledge base construction module 23 is used to add self-interaction between the head and tail entities of each triple based on the InterHT model, and to add mutual interaction between the head and tail entities of related triples, thereby constructing an AI-integrated knowledge base and deploying it to the emergency rescue command center.

[0059] The rescue positioning module 24 is used to connect to the IoT emergency call device corresponding to the emergency rescue request received by the emergency rescue command center and to collect voice data inside the elevator car where it is located.

[0060] The dialogue generation module 25 is used to generate intelligent dialogue based on the collected voice data, using a voice recognition model and an artificial intelligence knowledge base. This allows for intelligent dialogue with trapped passengers, both to soothe their emotions and to determine the type of elevator malfunction based on the content of the intelligent dialogue, so as to dispatch the appropriate rescue team to the scene.

[0061] It is not difficult to see that this embodiment is a system embodiment corresponding to the above method embodiments, and this embodiment can be implemented in conjunction with the above method embodiments. The relevant technical details and technical effects mentioned in the above method embodiments are still valid in this embodiment, and will not be repeated here to reduce repetition. Accordingly, the relevant technical details mentioned in this embodiment can also be applied to the above method embodiments.

[0062] It is worth mentioning that all modules and units involved in this embodiment are logical modules. In practical applications, a logical unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. Furthermore, to highlight the innovative aspects of this application, this embodiment does not introduce units that are not closely related to solving the technical problems proposed in this application; however, this does not mean that other units do not exist in this embodiment.

[0063] Another embodiment of this application provides an electronic device, such as Figure 5 As shown, it includes: at least one processor 31; and a memory 32 communicatively connected to the at least one processor 31; wherein the memory 32 stores instructions executable by the at least one processor 31, the instructions being executed by the at least one processor 31 to enable the at least one processor 31 to execute an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue as described in the above method embodiment.

[0064] The memory and processor are connected via a bus, which includes any number of interconnecting buses and bridges, connecting various circuits of one or more processors and the memory. The bus can also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and will not be described further herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be a single component or multiple components, such as multiple receivers and transmitters, providing a unit for communicating with various other devices over a transmission medium. Data processed by the processor is transmitted over the wireless medium via an antenna, which further receives data and transmits it to the processor.

[0065] The processor manages the bus and general processing, and also provides various functions, including timing, peripheral interfaces, voltage regulation, power management, and other control functions. Memory is used to store data used by the processor during operation.

[0066] Another embodiment of this application proposes a computer-readable storage medium storing a computer program that, when executed by a processor, enables an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue, as described in the above method embodiments.

[0067] That is, those skilled in the art will understand that all or part of the steps in the above method embodiments can be implemented by a program instructing related hardware. The program is stored in a storage medium and includes several instructions to cause a device (such as a microcontroller, chip, etc.) or processor to execute all or part of the steps of the method described in the method embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0068] Those skilled in the art will understand that the above embodiments are specific implementations of this application, and in practical applications, various changes can be made in form and detail without departing from the spirit and scope of this application. For those skilled in the art, several improvements and modifications can be made without departing from the principles of this application, and these improvements and modifications are also considered to be within the scope of protection of this application.

Claims

1. A method for generating intelligent dialogue based on the InterHT model for elevator emergency rescue, applicable to emergency rescue command centers, characterized in that, The method includes: Collect big data related to elevator emergency rescue and conduct business characteristic analysis; Based on the collected big data related to elevator emergency rescue and the analyzed business characteristics, several triplets consisting of entities, relationships, and entities are constructed for elevator emergency rescue. Based on the InterHT model, self-interaction is added between the head and tail entities of each triple, and mutual interaction is added between the head and tail entities of related triples to build an AI-integrated knowledge base and deploy it to the emergency rescue command center. When the emergency rescue command center receives an emergency rescue request, it connects the corresponding IoT emergency call device and collects voice data from inside the elevator car. Based on the collected voice data, intelligent dialogue is generated using a voice recognition model and an AI-powered knowledge base. This allows for intelligent dialogue with trapped passengers, both to calm them down and to determine the type of elevator malfunction based on the content of the dialogue, so that the appropriate rescue team can be dispatched to the scene.

2. The intelligent dialogue generation method based on the InterHT model for elevator emergency rescue according to claim 1, characterized in that, The collected big data related to elevator emergency rescue includes: basic information, basic operation data and historical fault data of elevators within the jurisdiction of the emergency rescue command center. The basic information of elevators includes elevator brand, service life, elevator size and working scenario. The basic operation data of elevators includes rated load capacity, rated speed, running time, car size, number of floors and drive mode. The historical fault data of elevators includes fault type, fault occurrence time, fault root cause, economic loss and emergency rescue records. The collected big data related to elevator emergency rescue was analyzed for business characteristics, specifically: Determine the relationship between the elevator's basic information and basic operating data and historical fault data.

3. The intelligent dialogue generation method based on the InterHT model for elevator emergency rescue according to claim 1, characterized in that, Based on the InterHT model, self-interactions are added between the head and tail entities of each triple, and mutual interactions are added between the head and tail entities of related triples. This constructs an AI-powered knowledge base, which is then deployed to the emergency rescue command center, including: For the Three pairs , Based on its tail entity Generate an auxiliary tail entity vector and based on its head entity Generate an auxiliary head entity vector At the same time, a unit moment is determined for it. ;in, , and They represent the first The head entity, relation, and tail entity of a triple; Based on the following formula, and right Add self-interaction, based on and right Add self-interaction, and get the result after adding self-interaction. Three pairs : ； ； ； in, This indicates that a composite function operation is being performed. Indicates the first Add a self-interacting head entity to each triple. Indicates the first Add a tail entity to each triple after self-interaction; Determine all The associated triples after adding self-interactions are calculated using the following formula, based on all triples related to... The associated auxiliary tail entity vector and unit moment of the triple after adding self-interaction, for Add interaction, based on all and The associated auxiliary head entity vector and unit moment of the triple after adding self-interaction, for Add interaction, get the result after adding interaction Three pairs : ； ； ； in, , and They represent the first Individual and The associated auxiliary tail entity vector, auxiliary head entity vector, and unit moment of the triple after adding self-interaction. Indicates all with The total number of associated triples after adding self-interactions. Indicates the first The head entity after adding interaction to a triple. Indicates the first Add an interactive tail entity to each triple; By integrating all the triples with added interactions, an AI-powered knowledge base is constructed and deployed to the emergency rescue command center.

4. The intelligent dialogue generation method based on the InterHT model for elevator emergency rescue according to claim 3, characterized in that, In the process of integrating all the triples after adding interactions to construct the AI-integrated knowledge base, the InterHT model is optimized using the following loss function: ； in, Represents the loss function. This represents the total number of triples. This represents the sigmoid function. This represents the preset fixed boundary parameters. This represents a function that calculates distance.

5. A method for generating intelligent dialogue based on the InterHT model for elevator emergency rescue according to claim 3, characterized in that, Based on the integration of all the triples with added interactions, an AI-powered knowledge base is constructed and deployed to the emergency rescue command center, including: By integrating all the triples with added interactions, an AI-integrated knowledge base is constructed. The constructed AI-integrated knowledge base is then used for this purpose. index, Indicators and The calculation of the indicators, in index, Indicators and If all indicators meet the preset design requirements, the constructed AI-integrated knowledge base will be deployed to the emergency rescue command center; otherwise, a new AI-integrated knowledge base will be constructed by re-integrating all the triples with added interactions.

6. The intelligent dialogue generation method based on the InterHT model for elevator emergency rescue according to claim 1, characterized in that, The IoT emergency call device is installed on the side wall of the elevator car. The IoT emergency call device is equipped with a one-button emergency call button, a display screen, a speaker, a sound receiver and a camera. When a trapped passenger triggers the one-click emergency call button, the IoT emergency call device immediately sends an emergency rescue request to the emergency rescue command center. When the elevator's self-checking device confirms that a person is trapped in the elevator, and the one-button emergency call device of the Internet of Things (IoT) is not triggered within the preset reaction time, the IoT emergency call device automatically sends an emergency rescue request to the emergency rescue command center.

7. A method for generating intelligent dialogue based on the InterHT model for elevator emergency rescue according to any one of claims 1 to 6, characterized in that, The emergency rescue command center is equipped with a DFCNN-based speech recognition model and emotion recognition model. Based on the collected speech data, it uses the speech recognition model and an AI-integrated knowledge base to generate intelligent dialogues and engage in intelligent conversations with stranded passengers, including: Using a DFCNN-based speech recognition model, speech recognition is performed on the collected speech data, and the collected speech data is converted into text data. Using an emotion recognition model, emotion recognition is performed based on the collected voice data and the converted text data to obtain the current emotion of the trapped passengers; Based on the converted text data and the current emotions of the trapped passengers, the system mines hidden relationships and searches for and matches knowledge in the AI-powered knowledge base to generate dialogue audio data corresponding to the converted text data. This audio data is then sent to the corresponding IoT distress call device to enable intelligent dialogue with the trapped passengers.

8. A smart dialogue generation system based on the InterHT model for elevator emergency rescue, suitable for emergency rescue command centers, characterized in that, The system includes: The analysis module is used to collect big data related to elevator emergency rescue and perform business characteristic analysis; The triplet construction module is used to construct several triplets composed of entities, relationships, and entities for elevator emergency rescue based on the collected big data related to elevator emergency rescue and the analyzed business characteristics. The knowledge base construction module is used to add self-interaction between the head and tail entities of each triple based on the InterHT model, and then add mutual interaction between the head and tail entities of related triples to build an AI-integrated knowledge base and deploy it to the emergency rescue command center. The rescue positioning module is used to connect to the IoT emergency call device corresponding to the emergency rescue request received by the emergency rescue command center and to collect voice data inside the elevator car where it is located. The dialogue generation module is used to generate intelligent dialogues based on the collected voice data, using a speech recognition model and an AI-powered knowledge base. This allows for intelligent dialogue with trapped passengers, both to calm them down and to determine the type of elevator malfunction based on the content of the dialogue, so that the appropriate rescue team can be dispatched to the scene.

9. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it can implement an intelligent dialogue generation method based on the InterHT model for elevator emergency rescue as described in any one of claims 1 to 7.

Images