An elevator voice call elevator sound source positioning method and system

Abstract

The application relates to an elevator voice call elevator sound source positioning method and system, and relates to the field of sound-controlled elevators, and comprises the following steps: collecting sound signals by adopting a double-microphone array, calculating the sound pressure level, horizontal azimuth angle and horizontal distance of a sound source; realizing real-time sensing of acoustic environment parameters including a reverberation time and a reflected sound proportion; dynamically adjusting the weight coefficients of three positioning parameters according to the environment parameters, and performing weighted summation on the standardized parameters to calculate a comprehensive score; and activating voice recognition and performing a call elevator operation when the comprehensive score meets a preset condition. The scheme provided by the application can effectively inhibit multi-floor mis-triggering caused by sound wave cross-layer propagation under an open stair through structure, and improve the positioning accuracy and call elevator reliability of a voice call elevator system.

Description

Technical Field

[0001] This application relates to the field of voice-controlled elevator technology, and in particular to a method and system for locating the sound source of an elevator voice call. Background Technology

[0002] Voice recognition elevator calling technology has been applied in elevator systems. Users simply need to say a voice command such as "I want to go upstairs" outside the elevator hall door, and the voice recognition elevator calling module can recognize the command and respond to the elevator call request, controlling the elevator to go to the corresponding floor. In elevators of public buildings, each floor usually has an independent waiting hall, and there are physical barriers between floors, which can effectively suppress the cross-floor propagation of sound. Therefore, the voice elevator calling system can work independently on each floor and is not prone to mutual interference.

[0003] However, in scenarios such as home elevators or villa elevators, building structures often employ open staircase designs, with each floor's entrance directly facing a shared stairwell. Due to the lack of effective physical barriers, sound waves can easily propagate upwards or downwards along the stairwell. For example, when a user issues a voice command "I want to go upstairs" in front of the third-floor entrance, this sound signal may be simultaneously picked up by voice acquisition modules on the second, fourth, or even more distant floors. This causes multiple floors' voice-activated elevator systems to simultaneously recognize the command and send separate call signals, leading to problems such as invalid elevator stops, response delays, and increased energy consumption. In existing technologies, some solutions rely solely on sound pressure level thresholds to determine whether voice recognition is triggered. However, the attenuation of sound pressure level with distance is insufficient to reliably distinguish the floor where the sound source is located, especially in stairwell structures where the difference in sound pressure levels collected from different floors may be small, making it difficult to avoid cross-floor false triggering.

[0004] Therefore, there is an urgent need for a method to accurately identify the floor where the sound source is located and distinguish the cross-floor propagation of sound waves in elevator voice call systems, so as to improve the positioning reliability and call accuracy of elevator voice call systems. Summary of the Invention

[0005] To address or partially address the problems existing in related technologies, this application provides a method and system for locating the sound source of elevator voice-activated elevator calls, aiming to solve the problem of simultaneous false triggering of elevator voice-activated elevator calls on multiple floors due to sound wave propagation across floors in open staircase structures.

[0006] The first aspect of this application provides a method for locating the sound source of an elevator voice-activated call, including: A dual-microphone array is used to collect sound signals. Based on the sound signals, the sound pressure level, horizontal azimuth, and horizontal distance of the sound source are calculated. The sound pressure level of the sound source is determined to be within the standard value for entering the sound source localization process. If it is not, the sound signal is ignored. Real-time perception of acoustic environment parameters, including at least reverberation time and the proportion of reflected sound; Based on the reverberation time and the proportion of reflected sound, the weighting coefficients of sound pressure level, horizontal azimuth angle and horizontal distance in sound source localization are dynamically adjusted. The standardized sound pressure level, horizontal azimuth, and horizontal distance are weighted and summed based on the adjusted weighting coefficients to calculate the comprehensive score. Determine whether the overall score meets the preset conditions. If so, activate voice recognition and execute the elevator call operation.

[0007] Optionally, the weighting coefficients of sound pressure level, horizontal azimuth, and horizontal distance in sound source localization can be dynamically adjusted, including: Environmental conditions are classified into three levels: high, medium, and low reverberation, based on the reverberation time. The weight of the horizontal azimuth decreases as the reverberation time increases, and is determined based on the functional relationship between the coefficient of the reverberation time interval and the reverberation time. The weight of horizontal distance decreases as the proportion of reflected sound increases, and is determined based on the functional relationship between the coefficient of the reverberation time interval and the proportion of reflected sound. The weight of the sound pressure level is obtained by subtracting the sum of the azimuth weight and the distance weight from 1.

[0008] Optionally, calculate a comprehensive score, including: The three positioning parameters—sound pressure level, horizontal azimuth, and horizontal distance—are standardized to dimensionless values ​​between 0 and 1: The sound pressure level standardization value is determined based on the ratio of the difference between the current sound pressure level and the preset minimum and maximum sound pressure levels; The standardized azimuth value is determined based on the ratio of the absolute value of the azimuth to 15 degrees; The distance standardization value is determined by the ratio of the difference between the preset upper distance limit and the current distance to the difference between the preset upper distance limit and the lower distance limit; The sound pressure level, azimuth, and distance are weighted and summed to obtain a comprehensive score.

[0009] Optionally, the sound pressure level, horizontal azimuth, and horizontal distance of the sound source are calculated based on the sound signal, including: The sound pressure level (SPL) is calculated using the following formula: In the formula, The current sound pressure level is obtained by collecting external sound sources through a microphone. The 20µPa reference sound pressure level is essentially an international standard set based on the physiological threshold of human hearing. The horizontal azimuth angle θ is calculated using the TDOA time difference algorithm: In the formula, The time difference between the arrival of sound at the two microphones. c is the speed of sound, taken as 343 m / s at room temperature; d is the microphone spacing, taken as 0.2 m to 0.4 m. The horizontal distance L is calculated based on the ratio of the reference sound intensity at the reference distance to the current sound intensity, using the following formula: In the formula, =1 meter In a quiet environment, at a distance from the microphone The sound is emitted from a location, and the sound intensity value measured by the microphone is I; I is the current sound intensity value measured by the microphone.

[0010] Optional, real-time sensing of acoustic environment parameters, including: Reverberation time was obtained by measuring the time required for the sound pressure level to decay from its initial value to -10 dB using a modified EDT method. The proportion of reflected sound is calculated based on the ratio of direct sound energy to the sum of direct sound energy and reflected sound energy: In the formula, To directly reach sound energy, This represents the reflected sound energy.

[0011] Optionally, determine whether the overall score meets preset conditions, including: Determine if the current overall score exceeds the preset threshold, and check if the current horizontal azimuth and horizontal distance meet the corresponding threshold requirements: When the reverberation time is greater than 1.2 seconds, the absolute value of the horizontal azimuth angle should not exceed 8 degrees and the horizontal distance should not exceed 1.2 meters. When the reverberation time is less than or equal to 0.8 seconds, the absolute value of the horizontal azimuth angle is required to be no greater than 20 degrees and the horizontal distance is required to be no greater than 2.46 meters. When the reverberation time is between 0.8 seconds and 1.2 seconds, the absolute value of the horizontal azimuth angle should not exceed 15 degrees and the horizontal distance should not exceed 1.8 meters. If all the above preset conditions are met, the speech recognition process will begin; otherwise, the sound signal will be ignored.

[0012] Optionally, the elevator voice call source localization method also includes: When sound events are detected simultaneously on adjacent floors, a consistency verification of the distance between floors is performed, and the difference in horizontal distance between the current floor and the adjacent floor is calculated. If the distance difference is less than 0.8 meters, the sound pressure level of the current floor must be at least 3 decibels higher than that of the adjacent floor; otherwise, the elevator call request of the current floor will be rejected. If the distance difference is greater than or equal to 0.8 meters, the absolute value of the azimuth angle of the current floor must not be greater than 10 degrees, and the absolute value of the azimuth angle of the adjacent floor must not be less than 30 degrees; otherwise, the elevator call request of the current floor will be rejected.

[0013] The second aspect of this application provides an elevator voice call sound source localization system, including: The system includes a voice acquisition module, an acoustic environment self-sensing module, a dynamic weight calculation module, a multi-condition joint scoring module, and a central processing unit module. The voice acquisition module is a dual-microphone array with a microphone spacing ranging from 0.2m to 0.4m, and a dual-microphone array is installed at each elevator landing door. The acoustic environment self-sensing module is used to collect acoustic environment parameters and preprocess them, calculate reverberation time and reflected sound ratio in real time, update environmental parameters every once in a while, and store historical environmental parameters for adaptive adjustment. The dynamic weight calculation module calculates the weights of parameters such as sound pressure level, horizontal azimuth, and horizontal distance based on reverberation time and the proportion of reflected sound, and updates and stores the weight coefficients in real time. The multi-condition joint scoring module converts sound pressure level, horizontal azimuth angle, and horizontal distance into standardized values ​​of 0 to 1, performs dynamic weighted calculation of comprehensive scoring, dynamically adjusts the trigger threshold according to environmental parameters, and conducts consistency verification of distance between floors. The central processing unit module is equipped with a CPU and memory to perform acoustic parameter calculations, weight adjustments, and scoring algorithm execution.

[0014] The technical solution provided in this application may include the following beneficial effects: By collecting sound signals through a dual-microphone array, the system comprehensively calculates three positioning parameters: sound pressure level, horizontal azimuth, and horizontal distance. It also dynamically adjusts the weighting coefficients of each parameter based on real-time reverberation time and the proportion of reflected sound. A weighted summation is used to calculate a comprehensive score to determine whether voice-activated elevator call is activated. This approach more accurately identifies the call intentions of users on the current floor and effectively suppresses sound wave propagation across floors and false triggering issues caused by open staircase structures. Simultaneously, by sensing acoustic environment parameters in real-time and dynamically adjusting the weighting coefficients, the system can adapt to acoustic environments under different reverberation and reflection conditions, maintaining high positioning stability and call reliability in complex sound fields, thus improving the overall performance and user experience of the elevator voice-activated elevator call system.

[0015] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0016] The above and other objects, features and advantages of this application will become more apparent from the more detailed description of exemplary embodiments thereof in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same components in the exemplary embodiments thereof.

[0017] Figure 1 This is a flowchart illustrating the elevator voice call sound source localization method according to an embodiment of this application; Figure 2 This is a schematic diagram of the structure of an elevator voice call sound source localization system shown in an embodiment of this application; Figure 3 This is a schematic diagram illustrating the use of the elevator voice call sound source localization system in an embodiment of this application. Detailed Implementation

[0018] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to make this application more thorough and complete, and to fully convey the scope of this application to those skilled in the art.

[0019] The technical solutions of the embodiments of this application are described in detail below with reference to the accompanying drawings.

[0020] In some implementations, see Figure 1 A method for locating the sound source of an elevator voice call, comprising: S101. Use a dual-microphone array to collect sound signals, calculate the sound pressure level, horizontal azimuth angle and horizontal distance of the sound source based on the sound signals, determine whether the sound pressure level of the sound source meets the standard value for entering the sound source localization process, and ignore the sound signal if it does not meet the standard value. Specifically, external sound signals are continuously collected by a dual-microphone array installed at the elevator lobby door on each floor; the spacing between the dual-microphone arrays is 0.2m to 0.4m. Based on the collected sound signals, three positioning parameters are calculated: sound pressure level, horizontal azimuth angle, and horizontal distance. Sound Pressure Level (SPL) Calculation: (1) In the formula, The current sound pressure level is obtained by collecting external sound sources through a microphone. The 20µPa reference sound pressure level is essentially an international standard set based on the physiological threshold of human hearing. The horizontal azimuth angle θ is calculated using the TDOA time difference algorithm: (2) In the formula, The time difference between the arrival of sound at the two microphones. c is the speed of sound, taken as 343 m / s at room temperature; d is the microphone spacing, taken as 0.2 m to 0.4 m. The horizontal distance L is calculated based on the ratio of the reference sound intensity at the reference distance to the current sound intensity, using the following formula: (3) In the formula, =1 meter In a quiet environment, at a distance from the microphone The sound is emitted from a location, and the sound intensity value measured by the microphone is I; I is the current sound intensity value measured by the microphone.

[0021] Determine if the sound pressure level (SPL) of the currently acquired sound signal is greater than a preset threshold. If SPL > 40 dB, it is considered a valid sound event and the sound source localization process begins; otherwise, continue listening.

[0022] S102. Real-time perception of acoustic environment parameters, including at least reverberation time and the proportion of reflected sound; Specifically, environmental features are extracted from the collected sound signals: Reverberation time Using the improved EDT method, the time required for the sound pressure level to decay from its initial value to -10 dB is obtained: (4) In the formula, EDT is the time required for the sound pressure level to decay from its initial value to -10 dB; The reflected sound ratio (DRR) is calculated based on the ratio of direct sound energy to the sum of direct sound energy and reflected sound energy. (5) In the formula, To directly reach sound energy, This represents the reflected sound energy.

[0023] S103. Based on the reverberation time and the proportion of reflected sound, dynamically adjust the weighting coefficients of sound pressure level, horizontal azimuth angle and horizontal distance in sound source localization; Specifically, environmental conditions are categorized based on reverberation time values, including high, medium, and low reverberation ranges: High reverberation environment: >1.2s; Low reverberation environment: ≤0.8s; Medium reverberation environment: 0.8s< ≤1.2s.

[0024] The initial weights are calculated as follows: Horizontal azimuth weight: (6) In the formula, This represents the azimuth weight base determined based on the reverberation time interval, and 0.1 represents the reverberation time influence factor, used to control the reverberation time. The degree of attenuation of the azimuth weight; Horizontal distance weight: (7) In the formula, This represents the distance weighting base determined based on the reverberation time interval. This represents a reference value for the proportion of reflected sound, used to normalize the proportion of reflected sound (DRR) to adjust the sensitivity of the distance weight as the reflected sound increases. Sound pressure level weighting: (8) The weights are dynamically adjusted based on the high, medium, and low reverberation ranges, as shown in the table below: The weight of the horizontal azimuth decreases with increasing reverberation time, and is determined based on the functional relationship between the reverberation time and the coefficient of the interval containing the proportion of reflected sound, and the reverberation time. The weight of the horizontal distance decreases with increasing proportion of reflected sound, and is determined based on the functional relationship between the reverberation time and the coefficient of the interval containing the proportion of reflected sound, and the proportion of reflected sound. The weight of the sound pressure level is obtained by subtracting the sum of the azimuth weight and the distance weight from 1.

[0025] S104. Based on the adjusted weighting coefficients, the standardized sound pressure level, horizontal azimuth angle, and horizontal distance are weighted and summed to calculate the comprehensive score. Specifically, the three positioning parameters—sound pressure level, horizontal azimuth, and horizontal distance—are standardized to dimensionless values ​​between 0 and 1: The sound pressure level normalization value is determined based on the ratio of the difference between the current sound pressure level and the preset minimum and maximum sound pressure levels: (9) In the formula, This represents the normalized value of the sound pressure level. This indicates the preset minimum sound pressure level. , This indicates the preset maximum sound pressure level. .

[0026] The standardized azimuth value is determined by the ratio of the absolute value of the azimuth to 15 degrees: (10) In the formula, This represents the standardized value of the azimuth angle.

[0027] The distance standardization value is determined by the ratio of the difference between the preset upper distance limit and the current distance to the difference between the preset upper distance limit and the lower distance limit: (11) In the formula, express, This indicates the preset distance limit. , This indicates the preset lower limit of distance, which is dynamically adjusted according to the environment.

[0028] Calculate the weighted sum of the sound pressure level normalized value, the azimuth normalized value, and the distance normalized value: (12) Receive overall score .

[0029] S105. Determine whether the comprehensive score meets the preset conditions. If so, activate voice recognition and execute the elevator call operation.

[0030] Determine if the current overall score exceeds the preset threshold, and check if the current horizontal azimuth and horizontal distance meet the corresponding threshold requirements: When the reverberation time is greater than 1.2 seconds, the absolute value of the horizontal azimuth angle should not exceed 8 degrees and the horizontal distance should not exceed 1.2 meters. When the reverberation time is less than or equal to 0.8 seconds, the absolute value of the horizontal azimuth angle is required to be no greater than 20 degrees and the horizontal distance is required to be no greater than 2.46 meters. When the reverberation time is between 0.8 seconds and 1.2 seconds, the absolute value of the horizontal azimuth angle should not exceed 15 degrees and the horizontal distance should not exceed 1.8 meters. If all the above preset conditions are met, the speech recognition process will begin; otherwise, the sound signal will be ignored.

[0031] When sound events are detected simultaneously on adjacent floors, a consistency verification of the distance between floors is performed, and the difference in horizontal distance between the current floor and the adjacent floor is calculated. If the distance difference is less than 0.8 meters, the sound pressure level of the current floor must be at least 3 decibels higher than that of the adjacent floor; otherwise, the elevator call request of the current floor will be rejected. If the distance difference is greater than or equal to 0.8 meters, the absolute value of the azimuth angle of the current floor must not be greater than 10 degrees, and the absolute value of the azimuth angle of the adjacent floor must not be less than 30 degrees; otherwise, the elevator call request of the current floor will be rejected.

[0032] Corresponding to the aforementioned application function implementation device embodiments, this application also provides an elevator voice call sound source localization system and corresponding embodiments.

[0033] See Figure 2 An elevator voice call sound source localization system includes: The system includes a voice acquisition module, an acoustic environment self-sensing module, a dynamic weight calculation module, a multi-condition joint scoring module, and a central processing unit module. The voice acquisition module is a dual-microphone array with a microphone spacing ranging from 0.2m to 0.4m, and a dual-microphone array is installed at each elevator landing door.

[0034] The acoustic environment self-sensing module is used to collect and preprocess environmental signals, and simultaneously calculate the reverberation time in real time. The system includes a DRR (Difference Reflection Rate) and an environmental parameter update triggered every 5 seconds, which uses a sliding window parameter fusion method. It also stores an acoustic environment feature database. Specifically, the acoustic environment self-sensing module includes a signal preprocessing unit and a reverberation time... The system includes a calculation unit, a DRR (Difference Reflection Rate) calculation unit, and an acoustic environment feature database; a signal preprocessing unit performs noise reduction and feature extraction on the acquired environmental signals; and a reverberation time unit. The computing unit is implemented using the improved EDT method. Real-time measurement, =6×EDT, where EDT is the time required for the sound pressure level to decay from its initial value to -10dB; the reflected sound ratio DRR calculation unit calculates DRR based on the ratio of direct sound to reflected sound energy. ,in To directly reach sound energy, This represents the reflected sound energy.

[0035] The dynamic weight calculation module is based on reverberation time. The weighting coefficients for the sound pressure level SPL, horizontal azimuth angle θ, and horizontal distance L are calculated based on the proportion of reflected sound (DRR), ensuring that the sum of the weighting coefficients is 1. Specifically, the dynamic weight calculation module includes a weight adjustment rule unit, a dynamic weight storage unit, and a weight coefficient verification unit; the weight adjustment rule unit is based on... The piecewise linear function of DRR adjusts the weights, supporting both lookup table method and piecewise function method; the dynamic weight storage unit updates and stores the weight coefficients in real time; the weight coefficient verification unit verifies that the sum of the weight coefficients is 1.

[0036] The multi-condition joint scoring module converts the sound pressure level (SPL), horizontal azimuth angle (θ), and horizontal distance (L) into standardized values ​​of 0 to 1 and performs dynamic weighted calculations. At the same time, it verifies the consistency of distances between floors and dynamically adjusts the trigger threshold based on environmental parameters. Specifically, the multi-condition joint scoring module includes a parameter standardization unit, a dynamic weighted calculation unit, a floor-to-floor distance consistency verification unit, and a scoring threshold dynamic adjustment unit. The parameter standardization unit performs 0-1 standardization conversion of sound pressure level (SPL), horizontal azimuth angle (θ), and horizontal distance (L). The dynamic weighted calculation unit calculates the weighted sum of the three parameters to obtain the comprehensive score. The floor-to-floor distance consistency verification unit suppresses simultaneous triggering of multiple floors. The scoring threshold dynamic adjustment unit adjusts the score based on… The environmental condition adjustment comprehensive score trigger threshold corresponding to DRR.

[0037] The central processing unit module is equipped with a CPU and memory to perform acoustic parameter calculations, weight adjustments, and scoring algorithm execution.

[0038] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for locating the sound source of an elevator voice-activated call, characterized in that, include: A dual-microphone array is used to collect sound signals. Based on the sound signals, the sound pressure level, horizontal azimuth angle and horizontal distance of the sound source are calculated. It is determined that the sound pressure level of the sound source meets the standard value for entering the sound source localization process. If it does not meet the standard value, the sound signal is ignored. Real-time sensing of acoustic environment parameters, wherein the acoustic environment parameters include at least reverberation time and the proportion of reflected sound; Based on the reverberation time and the proportion of reflected sound, the weighting coefficients of the sound pressure level, horizontal azimuth angle, and horizontal distance in sound source localization are dynamically adjusted. The standardized sound pressure level, horizontal azimuth angle, and horizontal distance are weighted and summed based on the adjusted weighting coefficients to calculate the comprehensive score. Determine whether the comprehensive score meets the preset conditions. If so, activate voice recognition and execute the elevator call operation.

2. The elevator voice call sound source localization method according to claim 1, characterized in that, The dynamic adjustment of the weighting coefficients of the sound pressure level, horizontal azimuth angle, and horizontal distance in sound source localization includes: Environmental conditions are classified into three levels: high, medium, and low reverberation, based on the reverberation time. The weight of the horizontal azimuth angle decreases as the reverberation time increases, and is determined based on the functional relationship between the coefficient of the reverberation time interval and the reverberation time. The weight of the horizontal distance decreases as the proportion of reflected sound increases, and is determined based on the functional relationship between the coefficient of the reverberation time interval and the proportion of reflected sound. The weight of the sound pressure level is obtained by subtracting the sum of the azimuth weight and the distance weight from 1.

3. The elevator voice call sound source localization method according to claim 1, characterized in that, The calculation of the comprehensive score includes: The three positioning parameters—sound pressure level, horizontal azimuth angle, and horizontal distance—are standardized to dimensionless values ​​between 0 and 1: The sound pressure level standardization value is determined based on the ratio of the difference between the current sound pressure level and the preset minimum and maximum sound pressure levels; The standardized azimuth value is determined based on the ratio of the absolute value of the azimuth to 15 degrees; The distance standardization value is determined by the ratio of the difference between the preset upper distance limit and the current distance to the difference between the preset upper distance limit and the lower distance limit; The sound pressure level, azimuth, and distance are weighted and summed to obtain a comprehensive score.

4. The elevator voice call sound source localization method according to claim 1, characterized in that, The calculation of the sound pressure level, horizontal azimuth, and horizontal distance of the sound source based on the sound signal includes: The sound pressure level (SPL) is calculated using the following formula: In the formula, The current sound pressure level is obtained by collecting external sound sources through a microphone. The 20µPa reference sound pressure level is essentially an international standard set based on the physiological threshold of human hearing. The horizontal azimuth angle θ is calculated using the TDOA time difference algorithm: In the formula, The time difference between the arrival of sound at the two microphones. c is the speed of sound, taken as 343 m / s at room temperature; d is the microphone spacing, taken as 0.2 m to 0.4 m. The horizontal distance L is calculated based on the ratio of the reference sound intensity at the reference distance to the current sound intensity, using the following formula: In the formula, =1 meter In a quiet environment, at a distance from the microphone The sound is emitted from a location, and the sound intensity value measured by the microphone is I; I is the current sound intensity value measured by the microphone.

5. The elevator voice call sound source localization method according to claim 1, characterized in that, The real-time sensed acoustic environment parameters include: Reverberation time was obtained by measuring the time required for the sound pressure level to decay from its initial value to -10 dB using a modified EDT method. The proportion of reflected sound is calculated based on the ratio of direct sound energy to the sum of direct sound energy and reflected sound energy: In the formula, To directly reach sound energy, This represents the reflected sound energy.

6. The elevator voice call sound source localization method according to claim 1, characterized in that, The determination of whether the comprehensive score meets the preset conditions includes: Determine if the current overall score exceeds the preset threshold, and check if the current horizontal azimuth and horizontal distance meet the corresponding threshold requirements: When the reverberation time is greater than 1.2 seconds, the absolute value of the horizontal azimuth angle should not exceed 8 degrees and the horizontal distance should not exceed 1.2 meters. When the reverberation time is less than or equal to 0.8 seconds, the absolute value of the horizontal azimuth angle is required to be no greater than 20 degrees and the horizontal distance is required to be no greater than 2.46 meters. When the reverberation time is between 0.8 seconds and 1.2 seconds, the absolute value of the horizontal azimuth angle should not exceed 15 degrees and the horizontal distance should not exceed 1.8 meters. If all the above preset conditions are met, the speech recognition process will begin; otherwise, the sound signal will be ignored.

7. The elevator voice call sound source localization method according to claim 1, characterized in that, Also includes: When sound events are detected simultaneously on adjacent floors, a consistency verification of the distance between floors is performed, and the difference in horizontal distance between the current floor and the adjacent floor is calculated. If the distance difference is less than 0.8 meters, the sound pressure level of the current floor must be at least 3 decibels higher than that of the adjacent floor; otherwise, the elevator call request of the current floor will be rejected. If the distance difference is greater than or equal to 0.8 meters, the absolute value of the azimuth angle of the current floor must not be greater than 10 degrees, and the absolute value of the azimuth angle of the adjacent floor must not be less than 30 degrees; otherwise, the elevator call request of the current floor will be rejected.

8. An elevator voice call sound source localization system, used to execute the elevator voice call sound source localization method according to any one of claims 1-7, characterized in that, include: The system includes a voice acquisition module, an acoustic environment self-sensing module, a dynamic weight calculation module, a multi-condition joint scoring module, and a central processing unit module. The voice acquisition module is a dual-microphone array with a microphone spacing ranging from 0.2m to 0.4m, and one set of the dual-microphone array is installed at each floor door of the elevator. The acoustic environment self-sensing module is used to collect acoustic environment parameters and preprocess them, calculate reverberation time and reflected sound ratio in real time, update environmental parameters every once in a certain period of time, and store historical environmental parameters for adaptive adjustment. The dynamic weight calculation module calculates the parameter weights of sound pressure level, horizontal azimuth angle, and horizontal distance based on reverberation time and the proportion of reflected sound, and updates and stores the weight coefficients in real time. The multi-condition joint scoring module converts sound pressure level, horizontal azimuth angle and horizontal distance into standardized values ​​of 0 to 1, performs dynamic weighted calculation of comprehensive score, dynamically adjusts trigger threshold according to environmental parameters, and simultaneously conducts consistency verification of distance between floors. The central processing unit module is equipped with a CPU and memory, and realizes acoustic parameter calculation, weight adjustment and scoring algorithm execution.

Images