One-key call based multi-channel audio distribution processing system

Abstract

This invention discloses a multi-channel audio dispatching system based on one-click calling, relating to the field of audio dispatching technology. By performing multi-dimensional analysis of the audio signal of a user's one-click call, it accurately determines the call value, type, and level of the call event, thereby improving the accuracy of event assessment and ensuring timely response to emergencies. It can effectively identify the severity of events, which helps to optimize response strategies and resource allocation. By comprehensively considering the efficiency of police officers, route conditions, and the urgency of the call, it quickly and effectively assigns calls to the most suitable police officers and recommends the best route, achieving precise allocation, improving processing speed, and enhancing overall emergency response capabilities.

Description

Technical Field

[0001] This invention relates to the field of audio dialing technology, and in particular to a multi-channel audio dialing processing system based on one-button calling. Background Technology

[0002] "One-click multi-channel audio distribution" refers to the function of distributing audio signals to multiple different output channels or destinations through a simple operation (such as pressing a button or clicking an icon on an interface). For example, in the event of an emergency (such as a fire or natural disaster), a one-click audio distribution system can quickly transmit alarms or notifications to various areas, ensuring that all personnel can receive important information in a timely manner. Therefore, one-click multi-channel audio distribution is particularly important.

[0003] Current multi-channel audio processing for one-click calls typically involves directly transferring the call to the nearest police officer. However, this method is overly simplistic and fails to fully consider factors such as the urgency of the call, the officer's processing capacity, traffic conditions, resource allocation, and professional skills. This approach negatively impacts call processing efficiency and quality, thereby reducing the system's overall responsiveness. Summary of the Invention

[0004] Therefore, it is necessary to provide a multi-channel audio splitting processing system based on one-click calling to address the problems mentioned in the background technology above.

[0005] The objective of this invention can be achieved through the following technical solution: a multi-channel audio dispatch processing system based on one-click calling, comprising a multi-channel audio dispatch processing platform, a user terminal, and a police officer terminal; wherein the multi-channel audio dispatch processing platform includes a database, a call evaluation module, and an allocation processing module;

[0006] The database stores basic information for each type of police officer; the basic information includes name, contact information, and historical call events; historical call events include the number of calls and the corresponding call value, trip interval, call start time, call end time, and call completion time for each call;

[0007] The call evaluation module performs multi-dimensional analysis based on the user's one-click call to extract call feature information, which includes call type, call level, call address, and call value.

[0008] The allocation processing module performs allocation processing based on the call characteristic information of the user's one-click call, specifically as follows:

[0009] Step 1: Set a corresponding number of dispatchers for each call level, and compare the call level with all the set call levels to match the corresponding number of dispatchers;

[0010] Step 2: Multiply the call value by the set distance conversion factor to obtain the call distance. Draw a circle with the call location as the center and the call distance as the radius to obtain the call range. Record the police officers' terminals that are within the call range, in a leisure state, and of the same type as the call as the initial selection terminals.

[0011] Step 3: Retrieve the basic information corresponding to the initial selection terminal, and analyze the timeliness of call event processing based on this information to obtain the processing value;

[0012] Step 4: Send a location acquisition command to each initial selection terminal to obtain the location of each initial selection terminal, and perform path analysis and road value based on the location of the initial selection terminal;

[0013] Step 5: Substitute the processing value Cf and road value Lf corresponding to each initial selection endpoint into the set formula. The optimal value CL is calculated, where h1 and h2 are the set proportional coefficients. Each initial selection terminal is sorted in descending order of its corresponding optimal value. The initial selection terminal with the same number of dispatched personnel is selected as the target terminal for this call event in descending order. The target path corresponding to each target terminal and the call characteristic information of this call event are sent to the target terminal, and the sending time is recorded as the call start time. The target terminal with the largest optimal value is selected as the connection terminal, and the user's one-click call is transferred to the connection terminal.

[0014] Step 6: When the target terminal arrives at the call location according to the target path, record the arrival time and mark it as the call end time; when a completion instruction is received from each target terminal, the corresponding target terminal adds a historical call event and records the completion time, marking it as the call completion time; update the historical call events of each police officer terminal to the database.

[0015] In some embodiments, the specific process for analyzing the timeliness of call event processing is as follows:

[0016] The call duration is calculated by taking the time difference between the call start time and the call end time, and the call speed is obtained by dividing the trip interval by the call duration; thus, the call speed corresponding to each call can be obtained.

[0017] The processing time is calculated by taking the time difference between the end time and the completion time of the call. This gives the processing time for each call. The call efficiency value for each call is obtained by dividing the call value for each call by the processing time.

[0018] Substitute the call speed H1 and call efficiency value H2 into the set formula Hd = log2(e d1×H1 +e d2×H2+1) Calculate the timeliness value Hd for each call, where d1 and d2 are the set proportional coefficients;

[0019] The timeliness value is compared and analyzed with the set timeliness interval to classify the call events corresponding to the timeliness value into high-efficiency processing, medium-efficiency processing and low-efficiency processing. The cumulative number of high-efficiency processing, medium-efficiency processing and low-efficiency processing is counted respectively and recorded as C1, C2 and C3 respectively. The timeliness values ​​corresponding to high-efficiency processing, medium-efficiency processing and low-efficiency processing are summed to obtain the high-efficiency processing value, medium-efficiency processing value and low-efficiency processing value, and recorded as C4, C5 and C6 respectively.

[0020] Substitute C1, C2, C3, C4, C5, and C6 into the set formula. The processing value Cf is obtained by calculation, where f1 > f2 > f3 > 0, and f1 > f2 > f3 > 0. Thus, the processing value corresponding to each initial selection end can be obtained.

[0021] In some embodiments, the specific process of path analysis based on the location of the initial selection endpoint is as follows:

[0022] Choose any initial selection point, using its location as the starting point and the calling location as the destination. Utilize Dijkstra's algorithm to obtain several paths. Use a GPS navigation map to obtain path information for each path, including path length, traffic flow, and pedestrian flow, and denote them as L1, L2, and L3 respectively. Substitute L1, L2, and L3 into the set formula. The road value Lf is calculated, where f4, f5, and f6 are set proportional coefficients, and f4 > f5 > f6 > 1; thus, the target path of each initial selection end and the corresponding road value can be obtained.

[0023] In some embodiments, the specific process of performing multi-dimensional analysis based on a user's one-click call is as follows:

[0024] 401: Uses artificial intelligence's natural language processing algorithms to identify the user's audio signal to determine the call location;

[0025] 402: A continuous audio signal is divided into short time frames. Each frame is usually given a window function, and adjacent frames usually have some overlap. This allows the audio signal to be divided into several frames, and audio fluctuation analysis can be performed to obtain audio fluctuation values.

[0026] 403: The noise level for each frame is identified and calculated using the short-time energy calculation method, and then quantified to obtain the background noise value.

[0027] 404: There are several call categories, each call category corresponds to several keywords, and each keyword corresponds to a severity coefficient; the keywords in the audio signal are extracted using speech recognition technology and recorded as audio keywords, and the audio keywords are compared with the set categories and the keywords corresponding to each category to match the corresponding call type and the severity coefficient of each audio keyword; the number of audio keywords is counted, and the severity coefficients corresponding to each audio keyword are summed to obtain the severity value;

[0028] 405: Substitute the audio fluctuation value DF, background noise value Zb, number of audio keywords M1, and severity value M2 into the set formula. The call value Ma is calculated, where a1, a2, a3, and a4 are the set proportional coefficients.

[0029] 406: Compare and analyze the call value with the set call range. When the call value is greater than the maximum value in the set call range, the user's one-click call is recorded as a level 1 call event; when the call value is within the set call range, the user's one-click call is recorded as a level 2 call event; when the call value is less than the minimum value in the set call range, the user's one-click call is recorded as a level 3 call event.

[0030] 407: Record the call location, call level, call value, and call type of a user's one-click call as the call feature information of the user's one-click call.

[0031] In some embodiments, the specific process of performing audio fluctuation analysis on each frame is as follows:

[0032] Applying a Short-Time Fourier Transform (SFT) to each frame yields its spectrum. The SFT converts the time-domain signal to a frequency-domain representation, then maps the spectrum to the Mel frequency scale. A logarithmic transform is performed on the spectrum at the Mel frequency scale, followed by a Discrete Cosine Transform (DCT) to calculate the MFCC coefficients. This allows identification of a set of MFCC coefficients corresponding to each frame in the audio signal. The mean of these coefficients is calculated and denoted as Fn, where n = 1, 2, 3…N, where N is a positive integer representing the total number of frames in the audio signal and n represents the sequence number of any given frame. This is achieved through a predefined formula. The audio fluctuation value DF is obtained through calculation.

[0033] In some embodiments, the specific process of quantifying and analyzing the noise level corresponding to each frame is as follows:

[0034] The noise level is compared and analyzed with the set noise range to identify the frames corresponding to the noise level as high noise frames, medium noise frames, and low noise frames, respectively. The number of high noise frames, medium noise frames, and low noise frames in the audio signal is counted and identified as Z1, Z2, and Z3, respectively. The noise levels corresponding to the high noise frames, medium noise frames, and low noise frames are summed to obtain the high noise value, medium noise value, and low noise value, respectively, and identified as Z4, Z5, and Z6, respectively.

[0035] Substitute Z1, Z2, Z3, Z4, Z5, and Z6 into the set formula. The background noise value Zb of the audio signal is calculated, where b1 > b2 > b3 > 0.

[0036] Compared with the prior art, the beneficial effects of the present invention are:

[0037] 1. By performing multi-dimensional analysis of the audio signal of a user's one-click call, the call value, type, and level of the call event can be accurately determined, thereby improving the accuracy of event assessment and ensuring timely response to emergencies; it can effectively identify the severity of the event, which helps to optimize response strategies and resource allocation;

[0038] 2. By comprehensively considering the efficiency of police officers, route conditions, and the urgency of the call, the system can quickly and effectively assign calls to the most suitable officers and recommend the best routes, thereby achieving precise allocation, improving processing speed, and enhancing overall emergency response capabilities. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of this application or the conventional technology, the drawings used in the description of the embodiments or the conventional technology will be briefly introduced below. Obviously, the drawings described below 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.

[0040] Figure 1 This is a schematic diagram of the principle of the present invention. Detailed Implementation

[0041] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0042] like Figure 1As shown, the multi-channel audio dispatching system based on one-click calling includes: a multi-channel audio dispatching platform, a user terminal, and a police officer terminal; the multi-channel audio dispatching platform has a built-in call evaluation module and a dispatching processing module; the user terminal uses the one-click calling function to access the multi-channel audio dispatching platform and uploads audio signals (specifically, the audio signals refer to the raw audio data generated when the user initiates a call, including their voice and background noise, etc.); the multi-channel audio dispatching platform processes the one-click call from the user terminal to dispatch the call to the corresponding police officer terminal;

[0043] The database stores basic information from the police officers' devices;

[0044] The call assessment module analyzes call characteristics based on the user's one-click call to obtain call type, call location, and call level; specifically:

[0045] Artificial intelligence natural language algorithms are used to identify the user's audio signal to determine the call location;

[0046] Continuous audio signals are divided into short time frames (typically 20-40 milliseconds each). This is because audio signals are relatively stable within a short time, making them suitable for spectral analysis. Each frame typically uses a window function (such as a Hamming or Hanning window) to reduce signal distortion at frame boundaries (the window function smooths frame boundaries and reduces spectral leakage). Adjacent frames usually have some overlap (e.g., 40% or 50% overlap) to ensure that key features of the audio signal are not lost during frame segmentation. A Short-Time Fourier Transform (STFT) is applied to each frame to obtain its spectrum. The STFT transforms the time-domain information... The spectrum is converted to a frequency domain representation for easier feature extraction. Then, the spectrum is mapped to the Mel frequency scale, and a logarithmic transform is performed on the spectrum at the Mel frequency scale. Finally, a Discrete Cosine Transform (DCT) is applied to calculate the MFCC coefficients. This allows identification of a set of MFCC coefficients (typically 20 to 40 coefficients) corresponding to each frame of the audio signal. The mean of the MFCC coefficients for each frame is calculated and denoted as Fn, where n = 1, 2, 3…N, where N is a positive integer representing the total number of frames in the audio signal and n represents the sequence number of any given frame. This is achieved through a predefined formula. The audio fluctuation value DF is obtained through calculation;

[0047] It's important to note that MFCC (Multi-Frequency Cepstral Coefficient) is a technique for converting audio signals into frequency domain features. It involves performing a Short-Time Fourier Transform (STFT) on the audio signal, then converting its spectrum to the Mel frequency scale, and finally calculating its cepstral coefficients. A series of MFCC characteristic coefficients, typically 20 to 40, represent the spectral characteristics of the audio signal on the Mel frequency scale. The standard deviation of the MFCC coefficients can be used to measure the degree of fluctuation of a specific coefficient. A high standard deviation may indicate more variation and jitter in the audio signal. Calculating the standard deviation of the MFCC coefficients for each frame reveals the fluctuation of the features, thus reflecting the overall fluctuation of the audio signal. Generally, fluctuations in a user's audio signal often indicate an urgent or serious situation.

[0048] The noise level for each frame is identified and calculated using a short-time energy (short-time energy refers to the energy of the signal within a short time frame) calculation method. The noise level is compared with a set noise range. When the noise level is greater than the maximum value in the set noise range, the noise in that frame is particularly significant; this frame is then designated as a high-noise frame. When the noise level is within the set noise range, it is designated as a medium-noise frame. When the noise level is less than the minimum value in the set noise range, it is designated as a low-noise frame. The number of high-noise, medium-noise, and low-noise frames in the audio signal is counted and designated as Z1, Z2, and Z3, respectively. The noise levels corresponding to the high-noise, medium-noise, and low-noise frames are summed to obtain the high-noise value, medium-noise value, and low-noise value, respectively, and designated as Z4, Z5, and Z6. The set formula is then used to calculate the noise level. The background noise value Zb of the audio signal is calculated, where b1, b2, and b3 are set scaling factors, and b1 > b2 > b3 > 0.

[0049] Several call categories are defined, each corresponding to several keywords, and each keyword has a severity coefficient. For example, keywords under the traffic police category include "vehicle scraping" and "hit-and-run," with the severity coefficient for "vehicle scraping" being lower than that for "hit-and-run." Specific keywords are set by those skilled in the art based on actual needs. Typically, the traffic police category mainly involves traffic management and accident handling, while the police officer category involves daily security and policing activities. For example, keywords for the traffic police category include "hit-and-run," "traffic accident," "vehicle scraping," "violation," and "traffic congestion," while keywords for the police officer category include "fighting," "theft," "domestic violence," and "telecom fraud." Speech recognition technology is used to extract keywords from the audio signal, which are then recorded as audio keywords. These audio keywords are compared with the defined categories and their corresponding keywords to match the corresponding call type and the severity coefficient of each audio keyword.

[0050] The number of audio keywords is counted (this refers to different audio keywords, excluding repeated audio keywords), and denoted as M1; the severity coefficients corresponding to each audio keyword are summed to obtain the severity value, and denoted as M2; then, using the defined formula... The call value Ma is calculated, where a1, a2, a3, and a4 are set proportional coefficients. As can be seen from the formula, the more volatile the user's audio signal and the noisier the background, the more serious the call event may be, requiring more police officers to handle and maintain the scene compared to normal times, and the larger the call value will be. The more audio keywords the user's call audio corresponds to and the higher their severity coefficient, the larger the call value will be.

[0051] The call value is compared and analyzed with the set call range. When the call value is greater than the maximum value in the set call range, it indicates that the user's one-click call event is relatively serious, and the user's one-click call is recorded as a level 1 call event; when the call value is within the set call range, the user's one-click call is recorded as a level 2 call event; when the call value is less than the minimum value in the set call range, the user's one-click call is recorded as a level 3 call event.

[0052] The call location, call level (level 1, level 2, or level 3), call value, and call type of a user's one-click call are recorded as the call feature information of the user's one-click call;

[0053] By performing multi-dimensional analysis of the audio signals of users' one-click calls, the call value, type, and level of the call event can be accurately determined, thereby improving the accuracy of event assessment and ensuring timely response to emergencies; it can effectively identify the severity of the event, which helps to optimize response strategies and resource allocation.

[0054] The allocation and processing module performs call allocation and route recommendation based on the call characteristics information of the user's one-click call, so as to achieve rapid response of one-click call allocation and improve the speed of police officers' handling of call events; specifically:

[0055] Step 1: Retrieve the call feature information of the user's one-click call, which includes the call location, call level (level 1, 2, or 3), call value, and call type; set a corresponding number of dispatchers for each call level. It should be noted that the higher the call level (level 1 is the highest level), the more dispatchers are required; compare the call level with all the set call levels to match the corresponding number of dispatchers.

[0056] Step 2: Multiply the call value by the set distance conversion factor to obtain the call distance; it can be seen that the larger the call value, the more serious the event at the call location is, and more police officers are needed to handle it, and the larger the call distance is; draw a circle with the call location as the center and the call distance as the radius to obtain the call range; mark the police officers within the call range, in a leisure state, and of the same type as the call as the initial selection terminals;

[0057] Step 3: Retrieve the basic information corresponding to the initial selection terminal. The basic information includes name, contact information, and historical call events. Historical call events include the number of calls, the call value corresponding to each call, the trip interval, the call start time, the call end time, and the call completion time. Calculate the call duration by the time difference between the call start time and the call end time, and divide the trip interval by the call duration to obtain the call speed. The call speed corresponding to each call can be recorded as H1.

[0058] The processing time is then calculated by taking the time difference between the call end time and the call completion time. This gives the processing time for each call. The call efficiency value for each call is then divided by the processing time, and denoted as H2.

[0059] The call speed H1 and call efficiency value H2 are calculated using the set formula Hd = log2(e d1×H1 +e d2×H2 +1) Calculate the timeliness value Hd for each call, where d1 and d2 are the set proportional coefficients;

[0060] The timeliness value is compared and analyzed with the set timeliness interval. When the timeliness value is greater than the maximum value in the set timeliness interval, it indicates that the police officer's handling of this call is very efficient, and an efficient processing is recorded. When the timeliness value is within the set timeliness interval, a medium-efficiency processing is recorded. When the timeliness value is less than the minimum value in the set timeliness interval, an inefficient processing is recorded. The cumulative number of efficient, medium-efficiency, and inefficient processing are counted separately and recorded as C1, C2, and C3, respectively. The timeliness values ​​corresponding to efficient, medium-efficiency, and inefficient processing are summed to obtain the efficient processing value, medium-efficiency processing value, and inefficient processing value, and recorded as C4, C5, and C6, respectively. The set formula is then used to calculate the efficient processing value, medium-efficiency processing value, and inefficient processing value. The processing value Cf is obtained through calculation, where f1, f2, and f3 are the set proportional coefficients, and f1 > f2 > f3 > 0; thus, the processing value corresponding to each initial selection end can be obtained;

[0061] Step 4: Send location acquisition commands to each initial selection terminal to obtain their locations. Select one initial selection terminal, using its location as the starting point and the calling location as the destination. Utilize Dijkstra's algorithm to obtain several paths. Use GPS navigation maps (such as Google Maps API, OpenStreetMap, or other geographic information systems) to obtain path information for each path. This path information includes path length (based on actual road and traffic conditions), traffic volume, and pedestrian volume, denoted as L1, L2, and L3 respectively. It should be noted that higher traffic and pedestrian volumes indicate greater road congestion. Use the established formula... The road value Lf is calculated, where f4, f5, and f6 are set proportional coefficients, and f4 > f5 > f6 > 1. The formula shows that a longer path length indicates a longer travel time, resulting in a higher road value. Similarly, higher traffic and pedestrian volumes on the path indicate greater congestion and a longer travel time, also resulting in a higher road value. The path with the lowest road value is selected as the target path for this initial selection. This yields the target paths and their corresponding road values ​​for each initial selection.

[0062] Step 5: Calculate the processing value Cf and road value Lf corresponding to each initial selection endpoint using the set formula. The optimal value CL is calculated, where h1 and h2 are the set proportional coefficients. Each initial selection terminal is sorted in descending order of its corresponding optimal value. The initial selection terminal with the same number of dispatched personnel is selected as the target terminal for this call event in descending order. The target path corresponding to each target terminal and the call characteristic information of this call event are sent to the target terminal, and the sending time is recorded as the call start time. The target terminal with the largest optimal value is selected as the connection terminal, and the user's one-click call is transferred to the connection terminal.

[0063] Step Six: When the target terminal arrives at the call location according to the target path, the arrival time is recorded and marked as the call end time; when a completion instruction is received from each target terminal, the corresponding target terminal adds a historical call event, records the completion time, and marks it as the call completion time; the historical call events of each police officer terminal are updated to the database to ensure timely information updates and accuracy of processing results;

[0064] By comprehensively considering the efficiency of police officers, route conditions, and the urgency of the call, the system can quickly and effectively assign calls to the most suitable officers and recommend the best routes, thereby achieving precise allocation, improving processing speed, and enhancing overall emergency response capabilities.

[0065] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0066] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A multi-channel audio splitting processing system based on one-click calling, comprising a multi-channel audio splitting processing platform, a user terminal, and a police officer terminal; characterized in that, The multi-channel audio splitting processing platform includes a database, a call evaluation module, and an allocation processing module; The database stores basic information for each type of police officer. The basic information includes name, contact information, and historical call events; historical call events include the number of calls and the corresponding call value, trip interval, call start time, call end time, and call completion time for each call; The call evaluation module performs multi-dimensional analysis based on the user's one-click call to extract call feature information, which includes call type, call level, call location and call value; The allocation processing module performs allocation processing based on the call characteristic information of the user's one-click call, specifically as follows: Step 1: Set a corresponding number of dispatchers for each call level, and compare the call level with all the set call levels to match the corresponding number of dispatchers; Step 2: Multiply the call value by the set distance conversion factor to obtain the call distance. Draw a circle with the call location as the center and the call distance as the radius to obtain the call range. Record the police officers' terminals that are within the call range, in a leisure state, and of the same type as the call as the initial selection terminals. Step 3: Retrieve the basic information corresponding to the initial selection terminal, and analyze the timeliness of call event processing based on this information to obtain the processing value; Step 4: Send a location acquisition command to each initial selection terminal to obtain the location of each initial selection terminal, and perform path analysis based on the location of the initial selection terminal to obtain the road value; Step 5: Normalize the processing value and road value corresponding to each initial selection terminal and take their values. Analyze the values ​​to obtain the optimal value. Sort each initial selection terminal in descending order of its corresponding optimal value. Select the initial selection terminal with the same number of dispatched personnel in descending order as the target terminal for this call event. Send the target path corresponding to each target terminal and the call characteristic information of this call event to the target terminal, record the sending time, and mark it as the call start time. Select the target terminal with the largest optimal value from the target terminals as the receiving terminal and transfer the user's one-click call to the receiving terminal. Step Six: When the target terminal arrives at the call location according to the target path, record the arrival time and mark it as the call end time; when a completion instruction is received from each target terminal, the corresponding target terminal adds a historical call event and records the completion time, marking it as the call completion time; update the historical call events of each police officer's terminal to the database; The specific process of performing multi-dimensional analysis based on a user's one-click call is as follows: 401: Uses artificial intelligence's natural language processing algorithms to identify the user's audio signal to determine the call location; 402: Divide a continuous audio signal into short time frames, apply a window function to each frame, and adjacent frames usually have some overlap; in this way, the audio signal can be divided into several frames, and audio fluctuation analysis can be performed on it to obtain audio fluctuation values. 403: The noise level for each frame is identified and calculated using the short-time energy calculation method, and then quantified to obtain the background noise value. 404: There are several call categories, each call category corresponds to several keywords, and each keyword corresponds to a severity coefficient; the keywords in the audio signal are extracted using speech recognition technology and recorded as audio keywords, and the audio keywords are compared with the set categories and the keywords corresponding to each category to match the corresponding call type and the severity coefficient of each audio keyword; the number of audio keywords is counted, and the severity coefficients corresponding to each audio keyword are summed to obtain the severity value; 405: Normalize the audio fluctuation value, background noise value, number of audio keywords and severity value, and take their values. Analyze the values ​​to obtain the call value. 406: Compare and analyze the call value with the set call range. When the call value is greater than the maximum value in the set call range, the user's one-click call is recorded as a level 1 call event; when the call value is within the set call range, the user's one-click call is recorded as a level 2 call event; when the call value is less than the minimum value in the set call range, the user's one-click call is recorded as a level 3 call event. 407: Record the call location, call level, call value, and call type of a user's one-click call as the call feature information of the user's one-click call.

2. The multi-channel audio dialing system based on one-button calling according to claim 1, characterized in that, The specific process for analyzing the timeliness of call event handling is as follows: The call duration is calculated by taking the time difference between the call start time and the call end time, and the call speed is obtained by dividing the trip interval by the call duration; thus, the call speed corresponding to each call can be obtained. The processing time is calculated by taking the time difference between the call end time and the call completion time, and thus the processing time for each call can be obtained. Divide the call value corresponding to each call by the processing time to obtain the call efficiency value for each call; The call speed and call efficiency values ​​are normalized and their values ​​are taken. The numerical values ​​are then analyzed to obtain the timeliness value corresponding to each call. The timeliness value is compared and analyzed with the set timeliness interval to classify the call events corresponding to the timeliness value into high-efficiency processing, medium-efficiency processing and low-efficiency processing. The cumulative number of high-efficiency processing, medium-efficiency processing and low-efficiency processing is counted respectively. The timeliness values ​​corresponding to high-efficiency processing, medium-efficiency processing and low-efficiency processing are summed to obtain the high-efficiency processing value, medium-efficiency processing value and low-efficiency processing value respectively. The processing value is obtained by formulaically calculating and analyzing the cumulative number of efficient processing, the cumulative number of medium-efficiency processing, the cumulative number of inefficient processing, the efficient processing value, the medium-efficiency processing value, and the inefficient processing value; thus, the processing value corresponding to each initial selection end can be obtained.

3. The multi-channel audio dialing system based on one-button calling according to claim 1, characterized in that, The specific process of path analysis based on the location of the initial selection point is as follows: Choose one of the initial selection terminals, take the location of the initial selection terminal as the starting point and the call location as the destination, and use Dijkstra's algorithm to obtain several paths; use GPS navigation map to obtain the path information of each path, including path length, traffic flow and pedestrian flow, and normalize them and take their values, and analyze the values ​​to obtain the road value; from this, we can know the target path of each initial selection terminal and the corresponding road value.

4. The multi-channel audio dialing system based on one-button calling according to claim 1, characterized in that, The specific process of performing audio fluctuation analysis on each frame is as follows: Applying a Short-Time Fourier Transform (SFT) to each frame yields its spectrum. The SFT converts the time-domain signal to a frequency-domain representation, then maps the spectrum to the Mel frequency scale. A logarithmic transform is performed on the spectrum at the Mel frequency scale, followed by a Discrete Cosine Transform (DCT) to calculate the MFCC coefficients. This allows identification of a set of MFCC coefficients corresponding to each frame in the audio signal. The mean of these coefficients is calculated and denoted as Fn, where n = 1, 2, 3…N, where N is a positive integer representing the total number of frames in the audio signal and n represents the sequence number of any given frame. This is achieved through a predefined formula. The audio fluctuation value DF is obtained through calculation.

5. The multi-channel audio dialing system based on one-button calling according to claim 4, characterized in that, The specific process of quantifying and analyzing the noise level for each frame is as follows: The noise level is compared and analyzed with the set noise range to record the frames corresponding to the noise level as high noise frame, medium noise frame and low noise frame respectively. The number of high noise frame, medium noise frame and low noise frame in the audio signal is counted respectively. The noise level corresponding to the high noise frame, medium noise frame and low noise frame is summed to obtain the high noise value, medium noise value and low noise value respectively. The background noise value of the audio signal is obtained by formulaically calculating and analyzing the number of high-noise frames, medium-noise frames, low-noise frames, high noise value, medium noise value, and low noise value.

Images